研究内容
ユーザー情報に基づいたタスク指向型対話システム基盤について研究しています．具体的には:
- ユーザの性格と対話タスク達成の影響を明確
- 対話ログからユーザの性格や情報を推定
- ユーザの性格に基づき適切な対話ができる対話システムを研究と開発
経歴
- 2021年法政大学情報科学研究科博士後期課程修了
- 2016年華中科技大学ソフトウェア工学博士前期課程修了
- 2015年法政大学情報科学研究科DDP修了
発表
2024
平井龍, 郭傲, 東中竜一郎
タスク指向型対話システムへの項目反応理論の適用によるユーザのタスク達成能力の推定 Proceedings Article
In: 言語処理学会 第30回年次大会 発表論文集, 2024.
@inproceedings{平井龍2024,
title = {タスク指向型対話システムへの項目反応理論の適用によるユーザのタスク達成能力の推定},
author = {平井龍, 郭傲, 東中竜一郎},
url = {https://www.anlp.jp/proceedings/annual_meeting/2024/pdf_dir/B9-4.pdf},
year = {2024},
date = {2024-03-11},
urldate = {2024-03-11},
booktitle = {言語処理学会 第30回年次大会 発表論文集},
journal = {言語処理学会 第30回年次大会 発表論文集},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
山下紗苗, 井上昂治, 郭傲, 望月翔太, 河原達也, 東中竜一郎
RealPersonaChat: 話者本人のペルソナと性格特性を含んだ雑談対話コーパス Proceedings Article
In: 言語処理学会 第30回年次大会 発表論文集, 2024.
@inproceedings{山下紗苗2024,
title = {RealPersonaChat: 話者本人のペルソナと性格特性を含んだ雑談対話コーパス},
author = {山下紗苗, 井上昂治, 郭傲, 望月翔太, 河原達也, 東中竜一郎},
url = {https://www.anlp.jp/proceedings/annual_meeting/2024/pdf_dir/B10-4.pdf},
year = {2024},
date = {2024-03-11},
urldate = {2024-03-11},
booktitle = {言語処理学会 第30回年次大会 発表論文集},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Ao Guo, Ryu Hirai, Atsumoto Ohashi, Yuya Chiba, Yuiko Tsunomori, Ryuichiro Higashinaka
Personality Prediction from Task-oriented and Open-domain Human–machine Dialogues Journal Article
In: Scientific Reports, vol. 14, iss. 3868, 2024.
@article{guo_sr2024,
title = {Personality Prediction from Task-oriented and Open-domain Human–machine Dialogues},
author = {Ao Guo and Ryu Hirai and Atsumoto Ohashi and Yuya Chiba and Yuiko Tsunomori and Ryuichiro Higashinaka},
doi = {https://doi.org/10.1038/s41598-024-53989-y},
year = {2024},
date = {2024-02-16},
urldate = {2024-02-16},
journal = {Scientific Reports},
volume = {14},
issue = {3868},
abstract = {If a dialogue system can predict the personality of a user from dialogue, it will enable the system to adapt to the user's personality, leading to better task success and user satisfaction. In a recent study, personality prediction was performed using the Myers–Briggs Type Indicator (MBTI) personality traits with a task-oriented human-machine dialogue using an end-to-end (neural-based) system. However, it is still not clear whether such prediction is generally possible for other types of systems and user personality traits. To clarify this, we recruited 378 participants, asked them to fill out four personality questionnaires covering 25 personality traits, and had them perform three rounds of human-machine dialogue with a pipeline task-oriented dialogue system or an end-to-end task-oriented dialogue system. We also had another 186 participants do the same with an open-domain dialogue system. We then constructed BERT-based models to predict the personality traits of the participants from the dialogues. The results showed that prediction accuracy was generally better with open-domain dialogue than with task-oriented dialogue, although Extraversion (one of the Big Five personality traits) could be predicted equally well for both open-domain dialogue and pipeline task-oriented dialogue. We also examined the effect of utilizing different types of dialogue on personality prediction by conducting a cross-comparison of the models trained from the task-oriented and open-domain dialogues. As a result, we clarified that the open-domain dialogue cannot be used to predict personality traits from task-oriented dialogue, and vice versa. We further analyzed the effects of system utterances, task performance, and the round of dialogue with regard to the prediction accuracy.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
If a dialogue system can predict the personality of a user from dialogue, it will enable the system to adapt to the user's personality, leading to better task success and user satisfaction. In a recent study, personality prediction was performed using the Myers–Briggs Type Indicator (MBTI) personality traits with a task-oriented human-machine dialogue using an end-to-end (neural-based) system. However, it is still not clear whether such prediction is generally possible for other types of systems and user personality traits. To clarify this, we recruited 378 participants, asked them to fill out four personality questionnaires covering 25 personality traits, and had them perform three rounds of human-machine dialogue with a pipeline task-oriented dialogue system or an end-to-end task-oriented dialogue system. We also had another 186 participants do the same with an open-domain dialogue system. We then constructed BERT-based models to predict the personality traits of the participants from the dialogues. The results showed that prediction accuracy was generally better with open-domain dialogue than with task-oriented dialogue, although Extraversion (one of the Big Five personality traits) could be predicted equally well for both open-domain dialogue and pipeline task-oriented dialogue. We also examined the effect of utilizing different types of dialogue on personality prediction by conducting a cross-comparison of the models trained from the task-oriented and open-domain dialogues. As a result, we clarified that the open-domain dialogue cannot be used to predict personality traits from task-oriented dialogue, and vice versa. We further analyzed the effects of system utterances, task performance, and the round of dialogue with regard to the prediction accuracy.
2023
Ryu Hirai, Shinya Iizuka, Haruhisa Iseno, Ao Guo, Jingjing Jiang, Atsumoto Ohashi, Ryuichiro Higashinaka
Team Flow at DRC2023: Building Common Ground and Text-based Turn-taking in a Travel Agent Spoken Dialogue System Proceedings Article
In: 2023.
@inproceedings{team-flow-drc2023,
title = {Team Flow at DRC2023: Building Common Ground and Text-based Turn-taking in a Travel Agent Spoken Dialogue System},
author = {Ryu Hirai and Shinya Iizuka and Haruhisa Iseno and Ao Guo and Jingjing Jiang and Atsumoto Ohashi and Ryuichiro Higashinaka},
url = {https://arxiv.org/abs/2312.13816},
year = {2023},
date = {2023-12-23},
urldate = {2023-12-23},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Kentaro Taki, Jianhua Ma, Ao Guo, Muxin Ma, Alex Qi
Multilevel Classification of Drowsiness States Using ECG With Optimized Convolutional Neural Network Proceedings Article Forthcoming
In: Proceedings of the 16th IEEE Cyber, Physical and Social Computing (CPSCom), Forthcoming.
@inproceedings{taki2023,
title = {Multilevel Classification of Drowsiness States Using ECG With Optimized Convolutional Neural Network },
author = {Kentaro Taki and Jianhua Ma and Ao Guo and Muxin Ma and Alex Qi },
year = {2023},
date = {2023-12-17},
urldate = {2023-12-17},
booktitle = {Proceedings of the 16th IEEE Cyber, Physical and Social Computing (CPSCom)},
abstract = {As drowsiness is recognized as one of the primary factors that threaten road safety, achieving a high-accuracy detection of drowsiness with finer granularity is crucial to ensuring safe driving. Current research has mainly focused on detecting drowsiness using EEG and ECG, typically classifying it into two states: alert and sleepy. However, given the multitude of possible parameters, the key factors that influence classification performance are still ambiguous. Furthermore, while a limited number of studies have gone beyond a binary classification of drowsiness, it is still not clear how well multilevel classification performs. To address these issues, we performed classification of drowsiness states with different granularities using ECG signals from the DROZY dataset. We first performed a binary classification of drowsiness by comparing combinations of parameters related to data preprocessing and model configuration. During binary classification, we achieved the state-of-the-art (SOTA) result with an accuracy of 99.65% by optimizing these parameters. These results also suggest that selecting a suitable window size along with the appropriate kernel size of the CNN model is essential for superior performance. We then performed a multilevel classification, categorizing drowsiness states into three, four, and eight levels, respectively. By adjusting parameters, the multilevel classification reached accuracy levels comparable to binary classification: 98.72% for three-level classification, 98.54% for four-level classification and 98.33% for eight-level classification.},
keywords = {},
pubstate = {forthcoming},
tppubtype = {inproceedings}
}
As drowsiness is recognized as one of the primary factors that threaten road safety, achieving a high-accuracy detection of drowsiness with finer granularity is crucial to ensuring safe driving. Current research has mainly focused on detecting drowsiness using EEG and ECG, typically classifying it into two states: alert and sleepy. However, given the multitude of possible parameters, the key factors that influence classification performance are still ambiguous. Furthermore, while a limited number of studies have gone beyond a binary classification of drowsiness, it is still not clear how well multilevel classification performs. To address these issues, we performed classification of drowsiness states with different granularities using ECG signals from the DROZY dataset. We first performed a binary classification of drowsiness by comparing combinations of parameters related to data preprocessing and model configuration. During binary classification, we achieved the state-of-the-art (SOTA) result with an accuracy of 99.65% by optimizing these parameters. These results also suggest that selecting a suitable window size along with the appropriate kernel size of the CNN model is essential for superior performance. We then performed a multilevel classification, categorizing drowsiness states into three, four, and eight levels, respectively. By adjusting parameters, the multilevel classification reached accuracy levels comparable to binary classification: 98.72% for three-level classification, 98.54% for four-level classification and 98.33% for eight-level classification.
Sanae Yamashita, Koji Inoue, Ao Guo, Shota Mochizuki, Tatsuya Kawahara, Ryuichiro Higashinaka
RealPersonaChat: A Realistic Persona Chat Corpus with Interlocutors’ Own Personalities Proceedings Article
In: Proceedings of the 37th Pacific Asia Conference on Language, Information and Computation, 2023.
@inproceedings{Yamashita2023b,
title = {RealPersonaChat: A Realistic Persona Chat Corpus with Interlocutors’ Own Personalities},
author = {Sanae Yamashita, Koji Inoue, Ao Guo, Shota Mochizuki, Tatsuya Kawahara, Ryuichiro Higashinaka},
url = {https://aclanthology.org/2023.paclic-1.85/},
year = {2023},
date = {2023-12-02},
urldate = {2023-12-02},
booktitle = {Proceedings of the 37th Pacific Asia Conference on Language, Information and Computation},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Ao Guo, Walid Brahim, Jianhua Ma
Influences of Mental Stress Level on Individual Identification using Wearable Biosensors Proceedings Article
In: Proceedings of the 8th IEEE Cyber Science and Technology Congress, pp. 1058-1063, IEEE 2023.
@inproceedings{guo_cyberscitech2023,
title = {Influences of Mental Stress Level on Individual Identification using Wearable Biosensors},
author = {Ao Guo and Walid Brahim and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/10361426/},
year = {2023},
date = {2023-11-17},
urldate = {2023-11-17},
booktitle = {Proceedings of the 8th IEEE Cyber Science and Technology Congress},
pages = {1058-1063},
organization = {IEEE},
abstract = {As wearable biosensors become increasingly popular, personalized services are enabled in various scenarios by identifying individuals through their biometric signals. Due to the inherent characteristics that are unique to each individual, biometric signals are widely used for identification. However, since an individual may experience varying levels of mental stress in different scenarios (e.g., work and relaxation), it remains unclear whether such stress influences the identification process. To address this, we built CNN-based models to identify 12 subjects under four levels of mental stress using three common biometric signals: R-R interval, galvanic skin reaction, and respiration. Our analysis revealed that individuals can be more easily identified when they are at certain levels of mental stress. We also examined the effect of mental stress on the identification of different subjects. We found that the inherent rhythm differences in biometric signals among individuals influence the accuracy of identification. We further showed how stress levels impact individual identification over different time periods.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
As wearable biosensors become increasingly popular, personalized services are enabled in various scenarios by identifying individuals through their biometric signals. Due to the inherent characteristics that are unique to each individual, biometric signals are widely used for identification. However, since an individual may experience varying levels of mental stress in different scenarios (e.g., work and relaxation), it remains unclear whether such stress influences the identification process. To address this, we built CNN-based models to identify 12 subjects under four levels of mental stress using three common biometric signals: R-R interval, galvanic skin reaction, and respiration. Our analysis revealed that individuals can be more easily identified when they are at certain levels of mental stress. We also examined the effect of mental stress on the identification of different subjects. We found that the inherent rhythm differences in biometric signals among individuals influence the accuracy of identification. We further showed how stress levels impact individual identification over different time periods.
Shinya Iizuka, Shota Mochizuki, Atsumoto Ohashi, Sanae Yamashita, Ao Guo, Ryuichiro Higashinaka
Clarifying the Dialogue-Level Performance of GPT-3.5 and GPT-4 in Task-Oriented and Non-Task-Oriented Dialogue Systems Proceedings Article
In: Proceedings of the AI-HRI Symposium at AAAI-FSS 2023, 2023.
@inproceedings{Iizuka2023,
title = {Clarifying the Dialogue-Level Performance of GPT-3.5 and GPT-4 in Task-Oriented and Non-Task-Oriented Dialogue Systems},
author = {Shinya Iizuka, Shota Mochizuki, Atsumoto Ohashi, Sanae Yamashita, Ao Guo and Ryuichiro Higashinaka},
url = {https://ai-hri.github.io/2023/papers/FSS-23_paper_632_cr.pdf},
year = {2023},
date = {2023-10-25},
urldate = {2023-10-25},
booktitle = {Proceedings of the AI-HRI Symposium at AAAI-FSS 2023},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Ryu Hirai, Ao Guo, Ryuichiro Higashinaka
Applying Item Response Theory to Task-oriented Dialogue Systems for Accurately Determining User's Task Success Ability Proceedings Article
In: Proceedings of the 24th Meeting of the Special Interest Group on Discourse and Dialogue, pp. 421–427, Prague, Czech Republic, 2023.
@inproceedings{hirai-etal-2023-applying,
title = {Applying Item Response Theory to Task-oriented Dialogue Systems for Accurately Determining User's Task Success Ability},
author = {Ryu Hirai and Ao Guo and Ryuichiro Higashinaka},
url = {https://aclanthology.org/2023.sigdial-1.39/},
year = {2023},
date = {2023-09-00},
urldate = {2023-09-00},
booktitle = {Proceedings of the 24th Meeting of the Special Interest Group on Discourse and Dialogue},
pages = {421--427},
address = {Prague, Czech Republic},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Ao Guo, Atsumoto Ohashi, Yuya Chiba, Yuiko Tsunomori, Ryu Hirai, and Ryuichiro Higashinaka
Personality-aware Natural Language Generation for Task-oriented Dialogue using Reinforcement Learning Proceedings Article
In: Proceedings of the 32st IEEE International Conference on Robot and Human Interactive Communication (RO-MAN) , pp. 1823-1828, Busan, Korea, 2023.
@inproceedings{guo_roman2023,
title = {Personality-aware Natural Language Generation for Task-oriented Dialogue using Reinforcement Learning},
author = {Ao Guo and Atsumoto Ohashi and Yuya Chiba and Yuiko Tsunomori and Ryu Hirai and and Ryuichiro Higashinaka},
url = {https://ieeexplore.ieee.org/document/10309654},
year = {2023},
date = {2023-00-00},
urldate = {2023-00-00},
booktitle = {Proceedings of the 32st IEEE International Conference on Robot and Human Interactive Communication (RO-MAN) },
pages = {1823-1828},
address = {Busan, Korea},
abstract = {A task-oriented dialogue system capable of expressing personality can improve user engagement and satisfaction. To realize such a system, this paper presents a method of building a personality-aware natural language generation (NLG) module in task-oriented dialogue using reinforcement learning (RL). This method handles both the expression of personality and system intent. During the RL process, a positive reward is given when the generated utterance correctly expresses the assigned personality and conveys its system intent simultaneously. In our experiments on the MultiWOZ dataset, we fine-tuned a personality-aware NLG module for two personality traits (extraversion and neural perception sensitivity). We experimented with data from MultiWOZ and a user simulator to confirm its effectiveness in terms of the ability to express personality and task performance.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A task-oriented dialogue system capable of expressing personality can improve user engagement and satisfaction. To realize such a system, this paper presents a method of building a personality-aware natural language generation (NLG) module in task-oriented dialogue using reinforcement learning (RL). This method handles both the expression of personality and system intent. During the RL process, a positive reward is given when the generated utterance correctly expresses the assigned personality and conveys its system intent simultaneously. In our experiments on the MultiWOZ dataset, we fine-tuned a personality-aware NLG module for two personality traits (extraversion and neural perception sensitivity). We experimented with data from MultiWOZ and a user simulator to confirm its effectiveness in terms of the ability to express personality and task performance.
2022
Ryu Hirai, Atsumoto Ohashi, Ao Guo, Hideki Shiroma, Xulin Zhou, Yukihiko Tone, Shinya Iizuka, Ryuichiro Higashinaka
Team Flow at DRC2022: Pipeline System for Travel Destination Recommendation Task in Spoken Dialogue Proceedings Article
In: 2022.
@inproceedings{team-flow-drc2022,
title = {Team Flow at DRC2022: Pipeline System for Travel Destination Recommendation Task in Spoken Dialogue},
author = {Ryu Hirai and Atsumoto Ohashi and Ao Guo and Hideki Shiroma and Xulin Zhou and Yukihiko Tone and Shinya Iizuka and Ryuichiro Higashinaka },
url = {https://arxiv.org/abs/2210.09518},
year = {2022},
date = {2022-10-25},
urldate = {2022-10-25},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
平井龍, 大橋厚元, 郭傲, 東中竜一郎
タスク指向型対話システムにおけるチュートリアルを用いた発話理解の改善 Proceedings Article
In: 人工知能学会全国大会論文集 第 36 回全国大会 (2022), pp. 2F4GS903–2F4GS903, 2022.
@inproceedings{平井龍2022,
title = {タスク指向型対話システムにおけるチュートリアルを用いた発話理解の改善},
author = {平井龍, 大橋厚元, 郭傲, 東中竜一郎},
url = {https://www.jstage.jst.go.jp/article/pjsai/JSAI2022/0/JSAI2022_2F4GS903/_article/-char/ja/},
year = {2022},
date = {2022-06-14},
urldate = {2022-06-14},
booktitle = {人工知能学会全国大会論文集 第 36 回全国大会 (2022)},
journal = {人工知能学会全国大会論文集 第 36 回全国大会 (2022)},
pages = {2F4GS903--2F4GS903},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
2021
Ao Guo, Atsumoto Ohashi, Ryu Hirai, Yuya Chiba, Yuiko Tsunomori, Ryuichiro Higashinaka
Influence of user personality on dialogue task performance: A case study using a rule-based dialogue system Proceedings Article
In: 3rd Workshop on NLP for ConvAI, pp. 263–270, 2021.
@inproceedings{2021guoinfluence,
title = {Influence of user personality on dialogue task performance: A case study using a rule-based dialogue system},
author = {Ao Guo and Atsumoto Ohashi and Ryu Hirai and Yuya Chiba and Yuiko Tsunomori and Ryuichiro Higashinaka},
url = {https://aclanthology.org/2021.nlp4convai-1.25.pdf},
year = {2021},
date = {2021-11-10},
urldate = {2021-11-05},
booktitle = {3rd Workshop on NLP for ConvAI},
pages = {263–270},
abstract = {Endowing a task-oriented dialogue system with adaptiveness to user personality can greatly help improve the performance of a dialogue task. However, such a dialogue system can be practically challenging to implement, because it is unclear how user personality influences dialogue task performance. To explore the relationship between user personality and dialogue task performance, we enrolled participants via crowdsourcing to first answer specified personality questionnaires and then chat with a dialogue system to accomplish assigned tasks. A rule-based dialogue system on the prevalent Multi-Domain Wizard-of-Oz (MultiWOZ) task was used. A total of 211 participants’ personalities and their 633 dialogues were collected and analyzed. The results revealed that sociable and extroverted people tended to fail the task, whereas neurotic people were more likely to succeed. We extracted features related to user dialogue behaviors and performed further analysis to determine which kind of behavior influences task performance. As a result, we identified that average utterance length and slots per utterance are the key features of dialogue behavior that are highly correlated with both task performance and user personality.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Endowing a task-oriented dialogue system with adaptiveness to user personality can greatly help improve the performance of a dialogue task. However, such a dialogue system can be practically challenging to implement, because it is unclear how user personality influences dialogue task performance. To explore the relationship between user personality and dialogue task performance, we enrolled participants via crowdsourcing to first answer specified personality questionnaires and then chat with a dialogue system to accomplish assigned tasks. A rule-based dialogue system on the prevalent Multi-Domain Wizard-of-Oz (MultiWOZ) task was used. A total of 211 participants’ personalities and their 633 dialogues were collected and analyzed. The results revealed that sociable and extroverted people tended to fail the task, whereas neurotic people were more likely to succeed. We extracted features related to user dialogue behaviors and performed further analysis to determine which kind of behavior influences task performance. As a result, we identified that average utterance length and slots per utterance are the key features of dialogue behavior that are highly correlated with both task performance and user personality.
Jiaoman Du, Jiandong Zhou, Xiang Li, Lei Li, Ao Guo
Integrated self-driving travel scheme planning Journal Article
In: International Journal of Production Economics, vol. 232, pp. 107963, 2021.
@article{du2021integrated,
title = {Integrated self-driving travel scheme planning},
author = {Jiaoman Du and Jiandong Zhou and Xiang Li and Lei Li and Ao Guo},
url = {https://www.sciencedirect.com/science/article/abs/pii/S0925527320303133},
year = {2021},
date = {2021-02-01},
journal = {International Journal of Production Economics},
volume = {232},
pages = {107963},
abstract = {Travel scheme planning is a crucial operational-level decision to be made in travel supply chain management. We investigate an integrated self-driving travel scheme planning (ISTSP) problem to optimize routing, hotel selection, and time scheduling under several streams of personalized considerations: best site-viewing time windows, rest requirements, and preference for site visiting sequences. The travel scheme planning problem is formulated in two models: (i) total cost minimization, and (ii) bi-objective optimization with total cost minimization and tourists’ utility maximization. A heuristic solution framework integrating multi-categorical attribute K-means clustering, dynamic programming algorithm, and constraint satisfaction procedure is designed to solve these two models. Finally, we provide illustrative examples to demonstrate the effectiveness and validity of the proposed models and solution methods.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Travel scheme planning is a crucial operational-level decision to be made in travel supply chain management. We investigate an integrated self-driving travel scheme planning (ISTSP) problem to optimize routing, hotel selection, and time scheduling under several streams of personalized considerations: best site-viewing time windows, rest requirements, and preference for site visiting sequences. The travel scheme planning problem is formulated in two models: (i) total cost minimization, and (ii) bi-objective optimization with total cost minimization and tourists’ utility maximization. A heuristic solution framework integrating multi-categorical attribute K-means clustering, dynamic programming algorithm, and constraint satisfaction procedure is designed to solve these two models. Finally, we provide illustrative examples to demonstrate the effectiveness and validity of the proposed models and solution methods.
2020
Hongyu Jiang, Ao Guo, Jianhua Ma
Genre-based emoji usage analysis and prediction in video comments Proceedings Article
In: 6th IEEE Cyber Science and Technology Congress, pp. 290–299, IEEE 2020.
@inproceedings{jiang2020genre,
title = {Genre-based emoji usage analysis and prediction in video comments},
author = {Hongyu Jiang and Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/9251135},
year = {2020},
date = {2020-11-12},
booktitle = {6th IEEE Cyber Science and Technology Congress},
pages = {290--299},
organization = {IEEE},
abstract = {Emojis can be seen as a visual language inserted in texts to express emotions, attitudes, and situations. It is widely used in computer-mediated communication, e.g., video comments. The emojis can express more detailed information beyond text information, and their usage can improve interlocutors' communication efficiency and emotions. The latest advances in natural language processing and deep learning have made it possible for chatbots to automatically add emojis in their dialogue. Precisely predicting emojis to be added is very challenging, especially in the video comments, where the use of emojis is complex, subtle, and associated with the cultural characteristics of video genres, e.g., anime and dancing. In this article, we first construct a benchmark dataset Bilibili comments dataset with more than 3.9 million comments that contain emojis in the video-sharing website Bilibili and then statistically analyze features of emoji's usage of video genre, comment content, and where an emoji appears in a sentence. According to the analyzed results and the gated recurrent unit (GRU) neural network, we propose a novel model of genre-based multitask GRU (GM-GRU) and its attention-added edition (GM- GRU+) to predict an emoji's category and position in a video comment. Our experiment and evaluation show that the proposed method can significantly increase the accuracy of predicted emojis for video comments.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Emojis can be seen as a visual language inserted in texts to express emotions, attitudes, and situations. It is widely used in computer-mediated communication, e.g., video comments. The emojis can express more detailed information beyond text information, and their usage can improve interlocutors' communication efficiency and emotions. The latest advances in natural language processing and deep learning have made it possible for chatbots to automatically add emojis in their dialogue. Precisely predicting emojis to be added is very challenging, especially in the video comments, where the use of emojis is complex, subtle, and associated with the cultural characteristics of video genres, e.g., anime and dancing. In this article, we first construct a benchmark dataset Bilibili comments dataset with more than 3.9 million comments that contain emojis in the video-sharing website Bilibili and then statistically analyze features of emoji's usage of video genre, comment content, and where an emoji appears in a sentence. According to the analyzed results and the gated recurrent unit (GRU) neural network, we propose a novel model of genre-based multitask GRU (GM-GRU) and its attention-added edition (GM- GRU+) to predict an emoji's category and position in a video comment. Our experiment and evaluation show that the proposed method can significantly increase the accuracy of predicted emojis for video comments.
Ao Guo, Hongyu Jiang, Jianhua Ma
Multi-scenario fusion for more accurate classifications of personal characteristics Proceedings Article
In: 6th IEEE Cyber Science and Technology Congress, pp. 300-305, IEEE 2020.
@inproceedings{guo2020multi,
title = {Multi-scenario fusion for more accurate classifications of personal characteristics},
author = {Ao Guo and Hongyu Jiang and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/9251160},
year = {2020},
date = {2020-11-11},
urldate = {2020-11-11},
booktitle = {6th IEEE Cyber Science and Technology Congress},
pages = {300-305},
organization = {IEEE},
abstract = {Personal character is a stable and comprehensive description of individual human being. It consists of a multitude of characteristics, and a comprehensive personal character can provide better-personalized services. Multi-modal fusion is a popular method to classify personal characteristic. The data with different forms and structures are integrated to simulate accurate personal characteristics. However, current studies mainly focus on classifying one or few personal characteristics, whereas the majority of personal data are collected from single scenario. The problem with uni-scenario data is that it is insufficient to achieve comprehensive and accurate classification of the whole characteristics of personal character. Thusly, multi-scenario fusion of personal characteristic classification is proposed to make for such flaw. This research proposes a multi-scenario framework to illustrate the fusion process and fusion modules. The framework contains two fusion methods, namely multi-scenario feature-level fusion and multi-scenario decision-level fusion. A detailed explanation of multi-scenario fusion algorithms is provided. The objective of experiments is to verify the effect of multi-scenario fusion to realize more accurate classifications of personal characteristics as opposed to the use of uni-scenario data. Accordingly, three types of experiments were conducted, and the physiological data of 30 participants were collected for characteristic classifications. The experimental results indicate that the multi-scenario fusion overwhelmingly surpasses the use of uniscenario of data in the classification of personal characteristics.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Personal character is a stable and comprehensive description of individual human being. It consists of a multitude of characteristics, and a comprehensive personal character can provide better-personalized services. Multi-modal fusion is a popular method to classify personal characteristic. The data with different forms and structures are integrated to simulate accurate personal characteristics. However, current studies mainly focus on classifying one or few personal characteristics, whereas the majority of personal data are collected from single scenario. The problem with uni-scenario data is that it is insufficient to achieve comprehensive and accurate classification of the whole characteristics of personal character. Thusly, multi-scenario fusion of personal characteristic classification is proposed to make for such flaw. This research proposes a multi-scenario framework to illustrate the fusion process and fusion modules. The framework contains two fusion methods, namely multi-scenario feature-level fusion and multi-scenario decision-level fusion. A detailed explanation of multi-scenario fusion algorithms is provided. The objective of experiments is to verify the effect of multi-scenario fusion to realize more accurate classifications of personal characteristics as opposed to the use of uni-scenario data. Accordingly, three types of experiments were conducted, and the physiological data of 30 participants were collected for characteristic classifications. The experimental results indicate that the multi-scenario fusion overwhelmingly surpasses the use of uniscenario of data in the classification of personal characteristics.
Hongyu Jiang, Ao Guo, Jianhua Ma
Automatic prediction and insertion of multiple emojis in social media text Proceedings Article
In: 13th IEEE Cyber, Physical and Social Computing (CPSCom), pp. 505-512, IEEE 2020.
@inproceedings{jiang2020automatic,
title = {Automatic prediction and insertion of multiple emojis in social media text},
author = {Hongyu Jiang and Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/9291606},
year = {2020},
date = {2020-10-28},
urldate = {2020-10-28},
booktitle = {13th IEEE Cyber, Physical and Social Computing (CPSCom)},
pages = {505-512},
organization = {IEEE},
abstract = {With the development of social media, many users are attracted by social platforms such as Twitter, Youtube, and TikTok. Emojis can be seen as a visual language inserted in texts to express emotions, attitudes, and situations. It is also widely used in social media communication, e.g., chit-chat and status sharing. The emojis can express more detailed and lively information beyond text information and can help chatbot become more like human beings. Current studies have explored how to predict single emoji according to a set of texts and context information. Some studies stated that people often add multiple emojis in texts and the different inserted positions corresponding to emojis' different functions. However, there is no study on inserting multiple emojis in texts. The latest advances in natural language processing and neural approaches have made it possible for chatbots to automatically add multiple emojis in chatbot's dialogue. In this article, we first construct a benchmark dataset contains more than 3.9 million comments in the video-sharing website Bilibili, and then analyze the features of emoji's usage and the relationships between emojis. Finally, a neural-based model is proposed to predict and insert multiple emojis in social media texts. The experiments and evaluation show our model got significant performance on predict and insert multiple emojis according to given sentences.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
With the development of social media, many users are attracted by social platforms such as Twitter, Youtube, and TikTok. Emojis can be seen as a visual language inserted in texts to express emotions, attitudes, and situations. It is also widely used in social media communication, e.g., chit-chat and status sharing. The emojis can express more detailed and lively information beyond text information and can help chatbot become more like human beings. Current studies have explored how to predict single emoji according to a set of texts and context information. Some studies stated that people often add multiple emojis in texts and the different inserted positions corresponding to emojis' different functions. However, there is no study on inserting multiple emojis in texts. The latest advances in natural language processing and neural approaches have made it possible for chatbots to automatically add multiple emojis in chatbot's dialogue. In this article, we first construct a benchmark dataset contains more than 3.9 million comments in the video-sharing website Bilibili, and then analyze the features of emoji's usage and the relationships between emojis. Finally, a neural-based model is proposed to predict and insert multiple emojis in social media texts. The experiments and evaluation show our model got significant performance on predict and insert multiple emojis according to given sentences.
Ao Guo, Jianhua Ma, Guanqun Sun, Shunxiang Tan
A personal character model of affect, behavior and cognition for individual-like research Journal Article
In: Computers & Electrical Engineering, vol. 81, pp. 106544, 2020.
@article{guo2020personal,
title = {A personal character model of affect, behavior and cognition for individual-like research},
author = {Ao Guo and Jianhua Ma and Guanqun Sun and Shunxiang Tan},
url = {https://www.sciencedirect.com/science/article/abs/pii/S0045790618334852},
year = {2020},
date = {2020-01-01},
journal = {Computers & Electrical Engineering},
volume = {81},
pages = {106544},
abstract = {Humanoid robots, avatars, as well as some machines or tools possessing distinctive human features or characteristics, have been studied and developed in recent years. Alongside these developments, a new research area has emerged, known as individual-like research, the aim of which is the creation of physical or digital entities that resemble, to a certain extent, an existing human individual. Such individual-like entities could generate novel and as yet undreamed-of applications in fields such as lifestyle management. A general or comprehensive model of an individual's character is the key to individual-like research. Derived from the personality model in psychology, this paper proposes a structuralized and computable model, namely the Personal Character Model of affect, behavior and cognition (ABC). We first assign mathematical abstractions to the proposed personal character model, then present a general computing process of personal character in the model, and finally perform an experiment to collect the state data of twenty subjects and further analyze the results pertaining to personal emotional stability and attention ability, as well as the relational characteristic of each subject's affect and cognition.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Humanoid robots, avatars, as well as some machines or tools possessing distinctive human features or characteristics, have been studied and developed in recent years. Alongside these developments, a new research area has emerged, known as individual-like research, the aim of which is the creation of physical or digital entities that resemble, to a certain extent, an existing human individual. Such individual-like entities could generate novel and as yet undreamed-of applications in fields such as lifestyle management. A general or comprehensive model of an individual's character is the key to individual-like research. Derived from the personality model in psychology, this paper proposes a structuralized and computable model, namely the Personal Character Model of affect, behavior and cognition (ABC). We first assign mathematical abstractions to the proposed personal character model, then present a general computing process of personal character in the model, and finally perform an experiment to collect the state data of twenty subjects and further analyze the results pertaining to personal emotional stability and attention ability, as well as the relational characteristic of each subject's affect and cognition.
2019
Ao Guo, Jianhua Ma, Kevin I-Kai Wang
Towards integrative personal character modeling using multi-strategy fusion across scenarios and periods Proceedings Article
In: 5th IEEE Cyber Science and Technology Congress, pp. 185–192, IEEE 2019.
@inproceedings{guo2019towards,
title = {Towards integrative personal character modeling using multi-strategy fusion across scenarios and periods},
author = {Ao Guo and Jianhua Ma and Kevin I-Kai Wang },
url = {https://ieeexplore.ieee.org/abstract/document/8890403},
year = {2019},
date = {2019-11-04},
booktitle = {5th IEEE Cyber Science and Technology Congress},
pages = {185--192},
organization = {IEEE},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Ao Guo, Jianhua Ma, Kevin I-Kai Wang
Integrated modeling of personal character using personal big data Proceedings Article
In: 12th IEEE Cyber, Physical and Social Computing (CPSCom), pp. 58–65, IEEE 2019.
@inproceedings{guo2019integrated,
title = {Integrated modeling of personal character using personal big data},
author = {Ao Guo and Jianhua Ma and Kevin I-Kai Wang},
url = {https://ieeexplore.ieee.org/document/8875286},
year = {2019},
date = {2019-10-21},
urldate = {2019-10-21},
booktitle = {12th IEEE Cyber, Physical and Social Computing (CPSCom)},
pages = {58--65},
organization = {IEEE},
abstract = {An individual-like intelligent artifact is a special kind of humanoid, which resembles a real human being in aspects of its human counterpart's appearance, behaviors, and characteristics. Such individual-like intelligent artifacts have potential for some fantastic applications, including better-personalized services and digital immortality. Although some researches have been carried out on its appearance and behavior, the essential component making such intelligent artifact really like its human counterpart is still the inherent characteristics or the personality. Accordingly, the personal character model (PCM) is proposed in this paper to achieve a comprehensive description of different individuals' characteristics. The PCM is an integrated model, which consists of characteristics of affect, behavior, and cognition (ABC), personality (P), and their relational (R) characteristics. To build up such characteristic model, an integrated model is proposed, consisting integrated modeling process, and personal characteristic computing. Due to the personal characteristics would be represented by different behavior in different situations, the personal big data is selected as the data source. The personal big data featured in multi-source, multi-modal, multi-situational, and multi-temporal. To verify the feasibility of personal character modeling, 20 participants are recruited to take three experiments for data collection, and computation of selected personal characteristics.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
An individual-like intelligent artifact is a special kind of humanoid, which resembles a real human being in aspects of its human counterpart's appearance, behaviors, and characteristics. Such individual-like intelligent artifacts have potential for some fantastic applications, including better-personalized services and digital immortality. Although some researches have been carried out on its appearance and behavior, the essential component making such intelligent artifact really like its human counterpart is still the inherent characteristics or the personality. Accordingly, the personal character model (PCM) is proposed in this paper to achieve a comprehensive description of different individuals' characteristics. The PCM is an integrated model, which consists of characteristics of affect, behavior, and cognition (ABC), personality (P), and their relational (R) characteristics. To build up such characteristic model, an integrated model is proposed, consisting integrated modeling process, and personal characteristic computing. Due to the personal characteristics would be represented by different behavior in different situations, the personal big data is selected as the data source. The personal big data featured in multi-source, multi-modal, multi-situational, and multi-temporal. To verify the feasibility of personal character modeling, 20 participants are recruited to take three experiments for data collection, and computation of selected personal characteristics.
Shunxiang Tan, Ao Guo, Jianhua Ma, Shengbing Ren
Personal affective trait computing using multiple data sources Proceedings Article
In: Proceedings of the 12th IEEE Cyber, Physical and Social Computing (CPSCom), pp. 66-73, IEEE 2019.
@inproceedings{tan2019personal,
title = {Personal affective trait computing using multiple data sources},
author = {Shunxiang Tan and Ao Guo and Jianhua Ma and Shengbing Ren},
url = {https://ieeexplore.ieee.org/document/8875394},
year = {2019},
date = {2019-10-21},
urldate = {2019-10-21},
booktitle = {Proceedings of the 12th IEEE Cyber, Physical and Social Computing (CPSCom)},
pages = {66-73},
organization = {IEEE},
abstract = {The personal affective trait (PAT) is a relatively stable affective trait that reflects differences among individuals. PAT is crucial to better understanding individuals in various applications, such as human-computer interaction (HCI) and personalized services. PAT computing can calculate several individual affective traits based on personal data. Because the PAT contributes to personality, PAT computing is based predominantly on personality computing. Moreover, the PAT can be classified into labeled and unlabeled PAT according to whether corresponding results from a psychological questionnaire. A general PAT computing process is proposed in order to compute various PAT categories. Because PAT is expressed in different forms under different scenarios, such as speech, gesture and context, PAT computing uses multiple data sources to generate comprehensive and accurate results. In order to illustrate the feasibility of the proposed PAT computing, 13 participants are recruited and three experiments completed so that their personal data can be collected and analyzed. Labeled PAT computing measures the PAT by fitting, while unlabeled PAT computing uses clustering. Furthermore, two PAT types - affective intensity and emotional stability - are computed using the analyzed data features. In order to verify the calculated PAT, the correlation between PAT and personality is measured.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The personal affective trait (PAT) is a relatively stable affective trait that reflects differences among individuals. PAT is crucial to better understanding individuals in various applications, such as human-computer interaction (HCI) and personalized services. PAT computing can calculate several individual affective traits based on personal data. Because the PAT contributes to personality, PAT computing is based predominantly on personality computing. Moreover, the PAT can be classified into labeled and unlabeled PAT according to whether corresponding results from a psychological questionnaire. A general PAT computing process is proposed in order to compute various PAT categories. Because PAT is expressed in different forms under different scenarios, such as speech, gesture and context, PAT computing uses multiple data sources to generate comprehensive and accurate results. In order to illustrate the feasibility of the proposed PAT computing, 13 participants are recruited and three experiments completed so that their personal data can be collected and analyzed. Labeled PAT computing measures the PAT by fitting, while unlabeled PAT computing uses clustering. Furthermore, two PAT types - affective intensity and emotional stability - are computed using the analyzed data features. In order to verify the calculated PAT, the correlation between PAT and personality is measured.
Ao Guo, Jianhua Ma, Tan Shunxiang, Guanqun Sun
From affect, behavior, and cognition to personality: An integrated personal character model for individual-like intelligent artifacts Journal Article
In: World Wide Web, vol. 23, no. 2, pp. 1217-1239, 2019.
@article{guo2020affect,
title = {From affect, behavior, and cognition to personality: An integrated personal character model for individual-like intelligent artifacts},
author = {Ao Guo and Jianhua Ma and Tan Shunxiang and Guanqun Sun},
url = {https://link.springer.com/content/pdf/10.1007/s11280-019-00713-w.pdf},
year = {2019},
date = {2019-08-31},
journal = {World Wide Web},
volume = {23},
number = {2},
pages = {1217-1239},
abstract = {An individual-like intelligent artifact is a special kind of humanoid which resembles a human being in assimilating aspects of its real human counterpart’s cognition and neurological functions. Such an individual-like intelligent artifact could have a number of far-reaching applications, such as in creating a digital clone of an individual and bringing about forms of digital immortality. Although such intelligent artifacts have been created in various forms, such as physical robots or digital avatars, these creations are still far from modeling the inner cognitive and neurological mechanisms of an individual human. To imbue individual-like intelligent artifacts with the characteristics of individuals, we propose a Personal Character Model that consists of personality, the characteristics of affect, behavior, and cognition, and the relations between these characteristics. According to differential psychology and personality psychology, personality is the set of essential characteristics that make a person unique whereas characteristics in affect, behavior, and cognition explain a person’s stable and abstract personality in their diverse daily behavior. In addition, relations among these characteristics serve as a bridge from one characteristic to another. To illustrate the computing process of the personal character model, we first designed three experiments to collect physiological data and behavior data from twenty participants. Then we selected data features from the collected data using correlational analysis. Finally, we computed several representative characteristics from selected data features and represented the computed results.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
An individual-like intelligent artifact is a special kind of humanoid which resembles a human being in assimilating aspects of its real human counterpart’s cognition and neurological functions. Such an individual-like intelligent artifact could have a number of far-reaching applications, such as in creating a digital clone of an individual and bringing about forms of digital immortality. Although such intelligent artifacts have been created in various forms, such as physical robots or digital avatars, these creations are still far from modeling the inner cognitive and neurological mechanisms of an individual human. To imbue individual-like intelligent artifacts with the characteristics of individuals, we propose a Personal Character Model that consists of personality, the characteristics of affect, behavior, and cognition, and the relations between these characteristics. According to differential psychology and personality psychology, personality is the set of essential characteristics that make a person unique whereas characteristics in affect, behavior, and cognition explain a person’s stable and abstract personality in their diverse daily behavior. In addition, relations among these characteristics serve as a bridge from one characteristic to another. To illustrate the computing process of the personal character model, we first designed three experiments to collect physiological data and behavior data from twenty participants. Then we selected data features from the collected data using correlational analysis. Finally, we computed several representative characteristics from selected data features and represented the computed results.
Guanqun Sun, Ao Guo, Jianhua Ma, Jianguo Wei
Personal trait analysis using word2vec based on user-generated text Proceedings Article
In: 5th IEEE Smart World Congress, pp. 1131-1137, IEEE 2019.
@inproceedings{sun2019personal,
title = {Personal trait analysis using word2vec based on user-generated text},
author = {Guanqun Sun and Ao Guo and Jianhua Ma and Jianguo Wei},
url = {https://ieeexplore.ieee.org/abstract/document/9060181},
year = {2019},
date = {2019-08-19},
booktitle = {5th IEEE Smart World Congress},
pages = {1131-1137},
organization = {IEEE},
abstract = {Personal trait is to measure the habitual patterns of behavior, thought, and emotion. It differs over individuals and is comparatively stable over time, relatively consistent over situations. Personal trait is significant for it has a lot of applications, such as recommendation system, chatbot and human resource management. It is convenient to recognize personal trait through wearable devices, social media and so on. Traditionally, personal trait is measured in general categories such as Big Five, which contains five traits: extroversion, neuroticism, agreeableness, conscientiousness, and openness. However, it is too abstract to describe personal trait in five aspects. We need the personal trait measured in more specific aspects, such as trait of interest or affect. We can know a person better through the traits in specific aspects than in the traditional abstract ways. In this paper, we proposed a general method of measuring personal trait called Personal Trait Matrix including topic word extraction and the word representation by word2vec based on user-generated text. Then a case study is made with datasets called myPersonality. The diversity of affects and social interactions are measured. Next, the correlation between the trait and the personality of Big Five was analyzed and discussed. The results demonstrate that the proposed method can measure the personal trait in affect and social interactions.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Personal trait is to measure the habitual patterns of behavior, thought, and emotion. It differs over individuals and is comparatively stable over time, relatively consistent over situations. Personal trait is significant for it has a lot of applications, such as recommendation system, chatbot and human resource management. It is convenient to recognize personal trait through wearable devices, social media and so on. Traditionally, personal trait is measured in general categories such as Big Five, which contains five traits: extroversion, neuroticism, agreeableness, conscientiousness, and openness. However, it is too abstract to describe personal trait in five aspects. We need the personal trait measured in more specific aspects, such as trait of interest or affect. We can know a person better through the traits in specific aspects than in the traditional abstract ways. In this paper, we proposed a general method of measuring personal trait called Personal Trait Matrix including topic word extraction and the word representation by word2vec based on user-generated text. Then a case study is made with datasets called myPersonality. The diversity of affects and social interactions are measured. Next, the correlation between the trait and the personality of Big Five was analyzed and discussed. The results demonstrate that the proposed method can measure the personal trait in affect and social interactions.
2018
Ao Guo, Jianhua Ma, Kevin I-Kai Wang
From user models to the Cyber-I model: Approaches, progresses and issues Proceedings Article
In: 4th IEEE Cyber Science and Technology Congress, pp. 33-40, IEEE 2018.
@inproceedings{guo2018user,
title = {From user models to the Cyber-I model: Approaches, progresses and issues},
author = {Ao Guo and Jianhua Ma and Kevin I-Kai Wang},
url = {https://ieeexplore.ieee.org/abstract/document/8511864},
year = {2018},
date = {2018-10-29},
booktitle = {4th IEEE Cyber Science and Technology Congress},
pages = {33-40},
organization = {IEEE},
abstract = {A user model is a collection and categorization of personal data associated with a specific user. With the development of personal computers, the usage of the user model has been evolving in the past 40 years. In the early age of the computer, as a part of computer application, one user model could only be used for a single domain, such as the Human-Computer Interaction (HCI) and the education support. With the increase of many different applications, a user model was separated from its specific computer application to support the cross-domain usage for multiple applications. The emergence of the Internet made the user model to be the users' specific representation on a single social platform for the identification by other users. In the foreseeable future, the user model will become a common representation for its user via multiple social platforms. Theoretically, a unique, digital, comprehensive description of every human individual, namely Cyber Individual (Cyber-I) may exist on the Internet. In this paper, we first focus on approaches to build up the Cyber-I model in comparison to the user model, in terms of model representations, architectures of modeling system and modeling techniques. We then summarize our previous works related to Cyber-I modeling, e.g., the personal data collection, the Cyber-I modeling architecture, and the modeling process. Finally, we present major issues that should be solved for the future development of Cyber-I model and modeling.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A user model is a collection and categorization of personal data associated with a specific user. With the development of personal computers, the usage of the user model has been evolving in the past 40 years. In the early age of the computer, as a part of computer application, one user model could only be used for a single domain, such as the Human-Computer Interaction (HCI) and the education support. With the increase of many different applications, a user model was separated from its specific computer application to support the cross-domain usage for multiple applications. The emergence of the Internet made the user model to be the users' specific representation on a single social platform for the identification by other users. In the foreseeable future, the user model will become a common representation for its user via multiple social platforms. Theoretically, a unique, digital, comprehensive description of every human individual, namely Cyber Individual (Cyber-I) may exist on the Internet. In this paper, we first focus on approaches to build up the Cyber-I model in comparison to the user model, in terms of model representations, architectures of modeling system and modeling techniques. We then summarize our previous works related to Cyber-I modeling, e.g., the personal data collection, the Cyber-I modeling architecture, and the modeling process. Finally, we present major issues that should be solved for the future development of Cyber-I model and modeling.
Yen Tsan, Ao Guo, Jianhua Ma, Runhe Huang, Zhong Chen
Incremental user modeling of online activity for Cyber-I growth with successive browsing logs Proceedings Article
In: 4th IEEE Cyber Science and Technology Congress, pp. 26-32, IEEE 2018.
@inproceedings{tsan2018incremental,
title = {Incremental user modeling of online activity for Cyber-I growth with successive browsing logs},
author = {Yen Tsan and Ao Guo and Jianhua Ma and Runhe Huang and Zhong Chen},
url = {https://ieeexplore.ieee.org/abstract/document/8511863},
year = {2018},
date = {2018-10-29},
booktitle = {4th IEEE Cyber Science and Technology Congress},
pages = {26-32},
organization = {IEEE},
abstract = {Cyber-Individual (Cyber-I) is a cyber-counterpart of a real individual (Real-I) in the physical world. It is a Real-I’s comprehensive description and digital existence in the cyber world. As a copy or digital clone of a Real-I, Cyber-I has a life cycle of birth, growth and death processes. One of basic characteristics in the growth of Cyber-I is to continuously process successive and life-long data for approximation to its Real-I. That is, a Cyber-I’s model should be built incrementally and refined successively. Therefore, this article is focused on studying an incremental mechanism for a growable Cyber-I. Users’ web browsing logs as a kind of personal data are collected daily, weekly and monthly for incremental modeling of a user’s online activity patterns. With incremental processing of the collected data, the Cyber-I’s model can be enhanced with more detailed and precise feature description of user online activities. A prototype system has been implemented to demonstrate the proposed incremental modeling for Cyber-I growth along with successive data coming and processing. The actual browsing data is also used to test and evaluate the developed system.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Cyber-Individual (Cyber-I) is a cyber-counterpart of a real individual (Real-I) in the physical world. It is a Real-I’s comprehensive description and digital existence in the cyber world. As a copy or digital clone of a Real-I, Cyber-I has a life cycle of birth, growth and death processes. One of basic characteristics in the growth of Cyber-I is to continuously process successive and life-long data for approximation to its Real-I. That is, a Cyber-I’s model should be built incrementally and refined successively. Therefore, this article is focused on studying an incremental mechanism for a growable Cyber-I. Users’ web browsing logs as a kind of personal data are collected daily, weekly and monthly for incremental modeling of a user’s online activity patterns. With incremental processing of the collected data, the Cyber-I’s model can be enhanced with more detailed and precise feature description of user online activities. A prototype system has been implemented to demonstrate the proposed incremental modeling for Cyber-I growth along with successive data coming and processing. The actual browsing data is also used to test and evaluate the developed system.
Ruiying Cai, Ao Guo, Jianhua Ma, Runhe Huang, Ruiyun Yu, Chen Yang
Correlation analyses between personality traits and personal behaviors under specific emotion states using physiological data from wearable devices Proceedings Article
In: 4th IEEE Cyber Science and Technology Congress, pp. 46-53, IEEE 2018.
@inproceedings{cai2018correlation,
title = {Correlation analyses between personality traits and personal behaviors under specific emotion states using physiological data from wearable devices},
author = {Ruiying Cai and Ao Guo and Jianhua Ma and Runhe Huang and Ruiyun Yu and Chen Yang},
url = {https://ieeexplore.ieee.org/abstract/document/8511866},
year = {2018},
date = {2018-10-29},
booktitle = {4th IEEE Cyber Science and Technology Congress},
pages = {46-53},
organization = {IEEE},
abstract = {In addition to the computational ability, human-like characteristics such as behavior, emotion, and personality can also be augmented to many personalized computers, applications robots and other systems. In recent years, there are many studies about human characteristics by using rich personal data collected from information systems and ubiquitous devices such as wearables. Besides separated studies on each aspect of human behavior, emotion and personality by using the personal data, it is also necessary to further study various relationships among these human characteristics. Therefore, this research is to examine how personality traits are associated with personal behaviors under specific emotional states based on physiological data collected from three wearable devices, Emotive Insight, Spire Stone and Huawei Fit Watch. Experimental data was gathered from 50 participants subjected to; a Big Five Inventory (BFI) questionnaire to get their personality traits, presenting before a crowd and/or watching a movie where physiological data measured by wearables. Attributes of personal behavior, e.g. blink, wink, surprise, furrow, smile and clench, are analyzed correlatively with the participants' personality traits under respective emotion states of excitement, relaxation, stress, engagement, interest and focus. Finally, we identify significant attribute correlations and find that correlations between the personality traits and the personal behaviors are greatly depended on the emotional states.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
In addition to the computational ability, human-like characteristics such as behavior, emotion, and personality can also be augmented to many personalized computers, applications robots and other systems. In recent years, there are many studies about human characteristics by using rich personal data collected from information systems and ubiquitous devices such as wearables. Besides separated studies on each aspect of human behavior, emotion and personality by using the personal data, it is also necessary to further study various relationships among these human characteristics. Therefore, this research is to examine how personality traits are associated with personal behaviors under specific emotional states based on physiological data collected from three wearable devices, Emotive Insight, Spire Stone and Huawei Fit Watch. Experimental data was gathered from 50 participants subjected to; a Big Five Inventory (BFI) questionnaire to get their personality traits, presenting before a crowd and/or watching a movie where physiological data measured by wearables. Attributes of personal behavior, e.g. blink, wink, surprise, furrow, smile and clench, are analyzed correlatively with the participants' personality traits under respective emotion states of excitement, relaxation, stress, engagement, interest and focus. Finally, we identify significant attribute correlations and find that correlations between the personality traits and the personal behaviors are greatly depended on the emotional states.
Ao Guo, Jianhua Ma
An integrative and precise approach in personality computing based on ontic personae modeling Proceedings Article
In: 3rd IEEE Cyber Science and Technology Congress, pp. 9-15, IEEE 2018.
@inproceedings{guo2017integrative,
title = {An integrative and precise approach in personality computing based on ontic personae modeling},
author = {Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/8328361},
year = {2018},
date = {2018-04-02},
booktitle = {3rd IEEE Cyber Science and Technology Congress},
pages = {9-15},
organization = {IEEE},
abstract = {A model describing the wide variety of human behaviours called personality, has become increasingly popular among researchers due to the widespread availability of personal data generated from the use of prevalent digital devices, e.g., smartphones and wearables. Plenty of personal data could be used to model an individual or even digitally clone a person. Such as a cyber-I (cyber individual), which aims at a unique and comprehensive description for an individual to mesh with various personalized services and applications. According to the high complexity and low efficiency in mining humans' mental states, one possible solution for Cyber-I modeling is to approach or compute an individual's personality. Although, an extensive research literature on or related to personality computing exists, i.e., into automatic personality computing, however, the precision and integrity of personality traits resulting in current personality computing are insufficient for the elaborate modeling in Cyber-I. Due to the heterogeneity of personal data, it is critical to organize the mass of personal data following a kind of pattern, and further establish a model involving a various of factors, i.e. location, time, person and behavior, to recreate the situation. In addition, it is also important to make a precise calculation of personality based on the model above. In our previous research, we have proposed a kind of model called the ontic personae to describe an individual's behavior with corresponding situation, or scenario. However, the lack of systematic implementation of ontic personae hindered further researches in personality computing. Therefore, this research mainly focuses on an integrative and precise approach in personality computing based on a well-designed ontic personae modeling. Especially, an ontic personae broker and personality computing engine are correspondingly, implemented for the integrative ontic personae generation and precise personality computing. The case stud...},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A model describing the wide variety of human behaviours called personality, has become increasingly popular among researchers due to the widespread availability of personal data generated from the use of prevalent digital devices, e.g., smartphones and wearables. Plenty of personal data could be used to model an individual or even digitally clone a person. Such as a cyber-I (cyber individual), which aims at a unique and comprehensive description for an individual to mesh with various personalized services and applications. According to the high complexity and low efficiency in mining humans' mental states, one possible solution for Cyber-I modeling is to approach or compute an individual's personality. Although, an extensive research literature on or related to personality computing exists, i.e., into automatic personality computing, however, the precision and integrity of personality traits resulting in current personality computing are insufficient for the elaborate modeling in Cyber-I. Due to the heterogeneity of personal data, it is critical to organize the mass of personal data following a kind of pattern, and further establish a model involving a various of factors, i.e. location, time, person and behavior, to recreate the situation. In addition, it is also important to make a precise calculation of personality based on the model above. In our previous research, we have proposed a kind of model called the ontic personae to describe an individual's behavior with corresponding situation, or scenario. However, the lack of systematic implementation of ontic personae hindered further researches in personality computing. Therefore, this research mainly focuses on an integrative and precise approach in personality computing based on a well-designed ontic personae modeling. Especially, an ontic personae broker and personality computing engine are correspondingly, implemented for the integrative ontic personae generation and precise personality computing. The case stud...
Ruiying Cai, Ao Guo, Jianhua Ma, Runhe Huang
Correlational analyses among personality traits emotional responses and behavioral states using physiological data from wearable sensors Proceedings Article
In: 10th International Conference on eHealth, Telemedicine, and Social Medicine (eTELEMED), pp. 51, 2018.
@inproceedings{cai2018correlational,
title = {Correlational analyses among personality traits emotional responses and behavioral states using physiological data from wearable sensors},
author = {Ruiying Cai and Ao Guo and Jianhua Ma and Runhe Huang},
url = {https://www.researchgate.net/profile/Shada-Alsalamah-2/publication/324525255_The_Tenth_International_Conference_on_eHealth_Telemedicine_and_Social_Medicine_eTELEMED/links/5ad3069fa6fdcc29357e893e/The-Tenth-International-Conference-on-eHealth-Telemedicine-and-Social-Medicine-eTELEMED.pdf#page=52},
year = {2018},
date = {2018-03-25},
urldate = {2018-03-25},
booktitle = {10th International Conference on eHealth, Telemedicine, and Social Medicine (eTELEMED)},
pages = {51},
abstract = {Mental health is crucial to the overall wellbeing of individuals, societies, and countries all the time. What’s more, personality traits, emotional responses and behavioral states with acute stress have a significant influence on mental health. However, correlations among the personality traits, emotional responses, and behavioral states were analyzed only by manual reports in psychology. This will cause problems of correlations updating delay and low reliability of data. Therefore, the main purpose of this paper is to examine whether and how personality traits are associated with emotional responses and behavioral states using physiological data from wearable devices. In experiments, 38 male and female university graduates and undergraduates volunteered as participants, and each one completed a Big Five Inventory (BFI) questionnaire and made a presentation in class, to get personality traits, emotional responses and behavioral states, respectively. In presenting, three wearable devices are used for emotional response data collection, including an Emotive Insight detecting electroencephalogram (EEG) data, a Spire Stone collecting respiration data and a Huawei Fit Watch getting heart rate value. In detail, six attributes of emotional responses: focus, interest, relaxation, engagement, stress and excitement, and 8 attributes of behavioral states: smile, clench, blink, surprise, furrow, wink, breath, and heart rate will be analyzed with personality traits. As a result, correlational analyses have indicated associations among the personality traits, emotional responses, and behavioral states. },
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Mental health is crucial to the overall wellbeing of individuals, societies, and countries all the time. What’s more, personality traits, emotional responses and behavioral states with acute stress have a significant influence on mental health. However, correlations among the personality traits, emotional responses, and behavioral states were analyzed only by manual reports in psychology. This will cause problems of correlations updating delay and low reliability of data. Therefore, the main purpose of this paper is to examine whether and how personality traits are associated with emotional responses and behavioral states using physiological data from wearable devices. In experiments, 38 male and female university graduates and undergraduates volunteered as participants, and each one completed a Big Five Inventory (BFI) questionnaire and made a presentation in class, to get personality traits, emotional responses and behavioral states, respectively. In presenting, three wearable devices are used for emotional response data collection, including an Emotive Insight detecting electroencephalogram (EEG) data, a Spire Stone collecting respiration data and a Huawei Fit Watch getting heart rate value. In detail, six attributes of emotional responses: focus, interest, relaxation, engagement, stress and excitement, and 8 attributes of behavioral states: smile, clench, blink, surprise, furrow, wink, breath, and heart rate will be analyzed with personality traits. As a result, correlational analyses have indicated associations among the personality traits, emotional responses, and behavioral states.
Ao Guo, Jianhua Ma
Archetype-based modeling of persona for comprehensive personality computing from personal big data Journal Article
In: Sensors, vol. 18, no. 3, pp. 684, 2018.
@article{guo2018archetype,
title = {Archetype-based modeling of persona for comprehensive personality computing from personal big data},
author = {Ao Guo and Jianhua Ma},
url = {https://www.mdpi.com/1424-8220/18/3/684},
year = {2018},
date = {2018-02-25},
journal = {Sensors},
volume = {18},
number = {3},
pages = {684},
abstract = {A model describing the wide variety of human behaviours called personality, is becoming increasingly popular among researchers due to the widespread availability of personal big data generated from the use of prevalent digital devices, e.g., smartphones and wearables. Such an approach can be used to model an individual and even digitally clone a person, e.g., a Cyber-I (cyber individual). This work is aimed at establishing a unique and comprehensive description for an individual to mesh with various personalized services and applications. An extensive research literature on or related to psychological modelling exists, i.e., into automatic personality computing. However, the integrity and accuracy of the results from current automatic personality computing is insufficient for the elaborate modeling in Cyber-I due to an insufficient number of data sources. To reach a comprehensive psychological description of a person, it is critical to bring in heterogeneous data sources that could provide plenty of personal data, i.e., the physiological data, and the Internet data. In addition, instead of calculating personality traits from personal data directly, an approach to a personality model derived from the theories of Carl Gustav Jung is used to measure a human subject’s persona. Therefore, this research is focused on designing an archetype-based modeling of persona covering an individual’s facets in different situations to approach a comprehensive personality model. Using personal big data to measure a specific persona in a certain scenario, our research is designed to ensure the accuracy and integrity of the generated personality model. View Full-Text},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
A model describing the wide variety of human behaviours called personality, is becoming increasingly popular among researchers due to the widespread availability of personal big data generated from the use of prevalent digital devices, e.g., smartphones and wearables. Such an approach can be used to model an individual and even digitally clone a person, e.g., a Cyber-I (cyber individual). This work is aimed at establishing a unique and comprehensive description for an individual to mesh with various personalized services and applications. An extensive research literature on or related to psychological modelling exists, i.e., into automatic personality computing. However, the integrity and accuracy of the results from current automatic personality computing is insufficient for the elaborate modeling in Cyber-I due to an insufficient number of data sources. To reach a comprehensive psychological description of a person, it is critical to bring in heterogeneous data sources that could provide plenty of personal data, i.e., the physiological data, and the Internet data. In addition, instead of calculating personality traits from personal data directly, an approach to a personality model derived from the theories of Carl Gustav Jung is used to measure a human subject’s persona. Therefore, this research is focused on designing an archetype-based modeling of persona covering an individual’s facets in different situations to approach a comprehensive personality model. Using personal big data to measure a specific persona in a certain scenario, our research is designed to ensure the accuracy and integrity of the generated personality model. View Full-Text
2017
Tongtong Xu, Ao Guo, Jianhua Ma, Kevin I-Kai Wang
Feature-based temporal statistical modeling of data streams from multiple wearable devices Proceedings Article
In: 3rd IEEE Cyber Science and Technology Congress, pp. 119-126, IEEE 2017.
@inproceedings{xu2017feature,
title = {Feature-based temporal statistical modeling of data streams from multiple wearable devices},
author = {Tongtong Xu and Ao Guo and Jianhua Ma and Kevin I-Kai Wang},
url = {https://ieeexplore.ieee.org/abstract/document/8328377},
year = {2017},
date = {2017-11-06},
booktitle = {3rd IEEE Cyber Science and Technology Congress},
pages = {119-126},
organization = {IEEE},
abstract = {Time is a vitally important issue in the coordination of multiple wearable devices. Theoretically, wearable applications should require data streams to be synchronized with the necessary degree of precision. However, in the available applications, this critical issue has not been well considered. Actually, time discrepancies exist among data streams, resulting in less accurate data analysis and fusion. The study of time discrepancy is rarely found in the literature, and there is no specific model to describe its features. In this paper, we first analyze the effect of time discrepancy on data. Then, by taking into account temporal features, we propose two typical models, which provide statistical methods for describing time discrepancy and its distribution. Furthermore, the accuracy of the models is verified by a set of experiments. Finally, we demonstrate the usability of the proposed models through a case study, in which the adaptive frequency strategy is adopted. Experimental results show that the strategy can not only guarantee the completeness of the data, but also reduce redundancy compared with the static frequency method. Our models and experiments of time discrepancy can be a basis for further research on the time synchronization of data from multiple wearable devices.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Time is a vitally important issue in the coordination of multiple wearable devices. Theoretically, wearable applications should require data streams to be synchronized with the necessary degree of precision. However, in the available applications, this critical issue has not been well considered. Actually, time discrepancies exist among data streams, resulting in less accurate data analysis and fusion. The study of time discrepancy is rarely found in the literature, and there is no specific model to describe its features. In this paper, we first analyze the effect of time discrepancy on data. Then, by taking into account temporal features, we propose two typical models, which provide statistical methods for describing time discrepancy and its distribution. Furthermore, the accuracy of the models is verified by a set of experiments. Finally, we demonstrate the usability of the proposed models through a case study, in which the adaptive frequency strategy is adopted. Experimental results show that the strategy can not only guarantee the completeness of the data, but also reduce redundancy compared with the static frequency method. Our models and experiments of time discrepancy can be a basis for further research on the time synchronization of data from multiple wearable devices.
Ao Guo, Jianhua Ma
Scenario-based modeling of ontic personae for automatic personality perception Proceedings Article
In: 2017 IEEE Ubiquitous Intelligence and Computing, pp. 1-7, IEEE 2017.
@inproceedings{guo2017scenario,
title = {Scenario-based modeling of ontic personae for automatic personality perception},
author = {Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/8397520},
year = {2017},
date = {2017-06-28},
booktitle = {2017 IEEE Ubiquitous Intelligence and Computing},
pages = {1-7},
organization = {IEEE},
abstract = {A model describing the wide variety of human behaviours called personality, is becoming increasingly popular among researchers due to the widespread availability of personal data generated from the use of prevalent digital devices, e.g., smartphones and wearables. Such an approach can be used to model an individual and even digitally clone a person, e.g., a Cyber-I (cyber individual), which can establish a unique and comprehensive description for an individual to mesh with various personalized services and applications. An extensive research literature on or related to personality computing exists, i.e., into automatic personality perception. However, the precision and integrity of current automatic personality perception is insufficient for the elaborate modeling in Cyber-I. Due to the heterogeneity of personal data, it is critical to classify the mass of personal data to valuable knowledge for sufficient personality perception. In addition, the precise calculation of personality based on such knowledge is also important. Therefore, this research is focused on designing scenario-based ontic personae as various individuals' facets in different situations for automatic personality perception. Using personal data, special scenarios, and based on specific ontic personae, our research is designed to ensure accurate modeling of ontic personae and personality perception.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A model describing the wide variety of human behaviours called personality, is becoming increasingly popular among researchers due to the widespread availability of personal data generated from the use of prevalent digital devices, e.g., smartphones and wearables. Such an approach can be used to model an individual and even digitally clone a person, e.g., a Cyber-I (cyber individual), which can establish a unique and comprehensive description for an individual to mesh with various personalized services and applications. An extensive research literature on or related to personality computing exists, i.e., into automatic personality perception. However, the precision and integrity of current automatic personality perception is insufficient for the elaborate modeling in Cyber-I. Due to the heterogeneity of personal data, it is critical to classify the mass of personal data to valuable knowledge for sufficient personality perception. In addition, the precise calculation of personality based on such knowledge is also important. Therefore, this research is focused on designing scenario-based ontic personae as various individuals' facets in different situations for automatic personality perception. Using personal data, special scenarios, and based on specific ontic personae, our research is designed to ensure accurate modeling of ontic personae and personality perception.
Tongtong Xu, Ao Guo, Jianhua Ma
Analysis of temporal features in data streams from multiple wearable devices Proceedings Article
In: 3rd IEEE International Conference on Cybernetics (CYBCONF), pp. 1-6, IEEE 2017.
@inproceedings{xu2017analysis,
title = {Analysis of temporal features in data streams from multiple wearable devices},
author = {Tongtong Xu and Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/7985758},
year = {2017},
date = {2017-06-20},
booktitle = {3rd IEEE International Conference on Cybernetics (CYBCONF)},
pages = {1-6},
organization = {IEEE},
abstract = {Extensive efforts have been made in both academia and industry for the research and development of intelligent wearable device. Especially, in recent years, increasing attention in healthcare monitoring systems and applications makes it particularly valuable to coordinate multiple wearable devices to provide diverse personal services. While these devices can collect a vast amount of data, one aspect which cannot be ignored is that given multitude of devices, combining data to derive valuable information requires them to work together without time difference. However, traditional applications directly deal with raw data, regardless of whether there are discrepancies, thus lead to inaccurate analysis results. In this paper, we first discuss the sources of time discrepancy as well as some temporal issues among multiple wearable devices based on Wear-I system. Then, several experiments are designed and the experimental results confirm the existence of time difference. We quantitatively analyze the time discrepancy in the system and find out the temporal features of data streams from the perspective of intra-stream and inter-stream, which provides the basis for further study of synchronization among wearables.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Extensive efforts have been made in both academia and industry for the research and development of intelligent wearable device. Especially, in recent years, increasing attention in healthcare monitoring systems and applications makes it particularly valuable to coordinate multiple wearable devices to provide diverse personal services. While these devices can collect a vast amount of data, one aspect which cannot be ignored is that given multitude of devices, combining data to derive valuable information requires them to work together without time difference. However, traditional applications directly deal with raw data, regardless of whether there are discrepancies, thus lead to inaccurate analysis results. In this paper, we first discuss the sources of time discrepancy as well as some temporal issues among multiple wearable devices based on Wear-I system. Then, several experiments are designed and the experimental results confirm the existence of time difference. We quantitatively analyze the time discrepancy in the system and find out the temporal features of data streams from the perspective of intra-stream and inter-stream, which provides the basis for further study of synchronization among wearables.
Ao Guo, Jianhua Ma
Context-aware scheduling in personal data collection from multiple wearable devices Journal Article
In: IEEE Access, vol. 5, pp. 2602-2614, 2017.
@article{guo2017context,
title = {Context-aware scheduling in personal data collection from multiple wearable devices},
author = {Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7847301},
year = {2017},
date = {2017-02-08},
journal = {IEEE Access},
volume = {5},
pages = {2602-2614},
abstract = {Due to the prevalence of smartphones and various wearable devices, the collection of rich personal data that can be used for human activity recognition, user modeling, and personalized services has become feasible. Because of its popularity and high accessibility, the smartphone has not only become an effective terminal in personal data collection, but also a gateway connecting wearable devices and gathering various types of personal data from these wearables. In most current applications, such wearables operate to collect data according to a fixed schedule, often preset manually by a user. The main problems in the data collection arising from such fixed scheduling are weak adaptiveness to wearables' state changes, a high level of redundancy in collected data, and possible mismatches in the dynamic precision requirements of data capture. Therefore, we propose a context-aware scheduler, that is able to dynamically adjust a data collection schedule based on contingent situations in the condition of wearables, system resource availability, and user behavior. This paper is focused on context data detection and data collection scheduling in a smartphone-based client-server system. The smartphone functions as not only a gateway gathering data from multiple wearables, but also a terminal for the performance of a context-aware scheduler. A context-aware engine is implemented to handle different contextual information. The data quality and system performance have been evaluated and verified in practical experimental tests.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Due to the prevalence of smartphones and various wearable devices, the collection of rich personal data that can be used for human activity recognition, user modeling, and personalized services has become feasible. Because of its popularity and high accessibility, the smartphone has not only become an effective terminal in personal data collection, but also a gateway connecting wearable devices and gathering various types of personal data from these wearables. In most current applications, such wearables operate to collect data according to a fixed schedule, often preset manually by a user. The main problems in the data collection arising from such fixed scheduling are weak adaptiveness to wearables' state changes, a high level of redundancy in collected data, and possible mismatches in the dynamic precision requirements of data capture. Therefore, we propose a context-aware scheduler, that is able to dynamically adjust a data collection schedule based on contingent situations in the condition of wearables, system resource availability, and user behavior. This paper is focused on context data detection and data collection scheduling in a smartphone-based client-server system. The smartphone functions as not only a gateway gathering data from multiple wearables, but also a terminal for the performance of a context-aware scheduler. A context-aware engine is implemented to handle different contextual information. The data quality and system performance have been evaluated and verified in practical experimental tests.
2016
Ao Guo, Jianhua Ma
A context-aware scheduling mechanism for smartphone-based personal data collection from multiple wearable devices Proceedings Article
In: 2016 IEEE Cyber, Physical and Social Computing, pp. 528-533, IEEE 2016.
@inproceedings{guo2016context,
title = {A context-aware scheduling mechanism for smartphone-based personal data collection from multiple wearable devices},
author = {Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/7917149},
year = {2016},
date = {2016-10-18},
booktitle = {2016 IEEE Cyber, Physical and Social Computing},
pages = {528-533},
organization = {IEEE},
abstract = {Due to rapid progresses of smartphone and various wearable devices, it becomes feasible to collect rich personal data that can be used for activity recognition, user modeling and better personalized services. Because of the popularity and high accessibility, a smartphone becomes not only an effective terminal in personal data collection but also a gateway to connect wearable devices and gather various kinds of personal data from these wearables. In the most of current applications, the wearables work for data collection according to a fixed schedule often preset manually by a user. The main problems in the data collection with following such fixed scheduling are weak adaptiveness to wearables' state change, big redundancy in collected data, and possible mismatch to dynamic precision requirements in data capture. Therefore, we propose a context-aware scheduling mechanism that is able to dynamically adjust the data collection schedule based on varying situations of wearable condition, network availability, computing resource and user state. This paper presents the details of this context-aware scheduling mechanism, and a corresponding smartphone-based system to collect personal data from multiple wearables and upload the gathered data to a server. The efficiency and effectiveness of the proposed scheduling mechanism have been verified by the actual data collection using the developed system.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Due to rapid progresses of smartphone and various wearable devices, it becomes feasible to collect rich personal data that can be used for activity recognition, user modeling and better personalized services. Because of the popularity and high accessibility, a smartphone becomes not only an effective terminal in personal data collection but also a gateway to connect wearable devices and gather various kinds of personal data from these wearables. In the most of current applications, the wearables work for data collection according to a fixed schedule often preset manually by a user. The main problems in the data collection with following such fixed scheduling are weak adaptiveness to wearables' state change, big redundancy in collected data, and possible mismatch to dynamic precision requirements in data capture. Therefore, we propose a context-aware scheduling mechanism that is able to dynamically adjust the data collection schedule based on varying situations of wearable condition, network availability, computing resource and user state. This paper presents the details of this context-aware scheduling mechanism, and a corresponding smartphone-based system to collect personal data from multiple wearables and upload the gathered data to a server. The efficiency and effectiveness of the proposed scheduling mechanism have been verified by the actual data collection using the developed system.
2015
Ao Guo, Jianhua Ma
A smartphone-based system for personal data management and personality analysis Proceedings Article
In: 2015 IEEE Pervasive Intelligence and Computing, pp. 2114-2122, IEEE 2015.
@inproceedings{guo2015smartphone,
title = {A smartphone-based system for personal data management and personality analysis},
author = {Ao Guo and Jianhua Ma},
url = {https://ieeexplore.ieee.org/abstract/document/7363360},
year = {2015},
date = {2015-11-28},
booktitle = {2015 IEEE Pervasive Intelligence and Computing},
pages = {2114-2122},
organization = {IEEE},
abstract = {The data from or about an individual, called personal data, is continuously increasing due to popularity of smart phones, wearables and other ubiquitous devices. Such personal data can be used to model a user and even digitally clone a person, e.g., Cyber-I (cyber individual) that aims at creating a unique and comprehensive description for every individual to support various personalized services and applications. Due to heterogeneity and sensitivity of personal data, one important issue is how to effectively collect and manage person data with sufficient security protection. Another important issue is how to figure out an individual's character, i.e., personality from personal data. Therefore, this research is focused on personal data management and personality analysis in a smartphone based client-server system. The smartphone functions as not only a source of personal data but also a gateway to manage other wearables and communicate with a server that keeps personal data in a larger amount and a longer period. A multi-security mechanism is implemented to ensure data security in collection, transmission and storage. Personality analysis is made from data normalization, feature extraction and clustering, to personality computation based on sociological personality theories.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The data from or about an individual, called personal data, is continuously increasing due to popularity of smart phones, wearables and other ubiquitous devices. Such personal data can be used to model a user and even digitally clone a person, e.g., Cyber-I (cyber individual) that aims at creating a unique and comprehensive description for every individual to support various personalized services and applications. Due to heterogeneity and sensitivity of personal data, one important issue is how to effectively collect and manage person data with sufficient security protection. Another important issue is how to figure out an individual's character, i.e., personality from personal data. Therefore, this research is focused on personal data management and personality analysis in a smartphone based client-server system. The smartphone functions as not only a source of personal data but also a gateway to manage other wearables and communicate with a server that keeps personal data in a larger amount and a longer period. A multi-security mechanism is implemented to ensure data security in collection, transmission and storage. Personality analysis is made from data normalization, feature extraction and clustering, to personality computation based on sociological personality theories.
表彰
- Outstanding Paper Award, 16th IEEE International Conference on Cyber, Physical and Social Computing, 2023.
- Outstanding PhD Dissertation Award, IEEE Technical Committee on Hyper-Intelligence, 2022.
- Best Student Paper Award, 5th IEEE International Conference on Cyber Science and Technology, 2020.
- Best Paper Award, 9th IEEE International Conference on Cyber Physical and Social Computing, 2016.