A5073610035- Sentiment Analysis and Opinion Mining
- Topic Modeling
- Recommender Systems and Techniques
- Advanced Graph Neural Networks
- Emotion and Mood Recognition
- Text and Document Classification Technologies
- Advanced Text Analysis Techniques
- Speech and dialogue systems
- Mobile Agent-Based Network Management
- Semantic Web and Ontologies
- Advanced Computational Techniques and Applications
- Multi-Agent Systems and Negotiation
- Geoscience and Mining Technology
- Machine Learning in Bioinformatics
- Complex Network Analysis Techniques
- Advanced Bandit Algorithms Research
- Advanced Clustering Algorithms Research
- Technology and Data Analysis
- Natural Language Processing Techniques
Chongqing University of Technology
2022-2025
Conversation recommender system (CRS) aims to provide high-quality recommendations users in fewer conversation turns. Existing studies often rely on knowledge graphs enhance the representation of entity information. However, these methods tend overlook inherent incompleteness graphs, making it challenging for models fully capture users' true preferences. Additionally, they fail thoroughly explore emotional tendencies toward entities or effectively differentiate varying impacts different user...
Abstract Sequential recommendation (SR) predicts the user’s future preferences based on sequence of interactions. Recently, some methods for SR have utilized contrastive learning to incorporate self-supervised signals into alleviate data sparsity problem. Despite these achievements, they overlook fact that users’ multi-behavior interactions in real-world scenarios (e.g., page view, favorite, add cart, and purchase). Moreover, disregard temporal dependencies their influence attribute...
Abstract Conversational Recommender Systems (CRS) aim to provide high-quality items users in fewer conversation rounds using natural language. Despite various attempts that have been made, there are still some problems: Previous CRS only learned item representations a single knowledge graph and ignored tags; information gaps exist the same from different graphs popularity both affect user preferences; system generated responses lack descriptiveness diversity. To address these problems fully...
Emotion Recognition in Conversation (ERC) plays an important role driving the development of human-machine interaction. Emotions can exist multiple modalities, and multimodal ERC mainly faces two problems: (1) noise problem cross-modal information fusion process, (2) prediction less sample emotion labels that are semantically similar but different categories. To address these issues fully utilize features each modality, we adopted following strategies: first, deep cues extraction was...
The basic concept of clustering and its correlating research work is firstly present, a new algorithm based on least cell (LCC) proposed analyzed which concerns the advantages disadvantage k-means grid algorithm. This efficient in dealing with huge amounts data can make paralleled processing, proved to be correct, fast through application customer relationship management. It overcomes given value k dense Lastly analysis evaluation given.
Aiming at the shortcoming of current automated negotiation systems, this paper applied machine learning to bilateral negotiation. It mainly researched mechanism in e-commerce. improved traditional Q-learning and designed dynamic algorithm. This algorithm estimated Q value according environment state action both agents, furthermore, recency-based exploration bonus were embedded. The Bayesian strategy negotiation, belief based on Bayesian. Finally, did experiments results show that can improve...
Conversational Recommender System (CRS) captures user's preferences based on the description of item to recommend for them. Existing CRS make recommendations using items mentioned in dialogues and external information; introduction information does not take into account completeness item's attributes. Besides, existing studies have paid attention semantic users' dialogues, ignoring effect attributes diversity utterance generation. Moreover, generated utterances lack explainability. To...
Emotion Recognition in Conversation (ERC) plays an important role driving the development of human-machine interaction. Emotions can exist multiple modalities, and multimodal ERC mainly faces two problems: (1) noise problem cross-modal information fusion process, (2) prediction less sample emotion labels that are semantically similar but different categories. To address these issues fully utilize features each modality, we adopted following strategies: first, deep cues extraction was...