NEUDM/PICA-V1
Feature Extraction
•
Updated
•
8
•
3
Using social media as a platform, and multimodal resources and all types of users as the data source, we study the representation, mining, search, and recommendation of the social media resources, users and their relationship. The research contents include multi-modal data processing, text mining, sentiment analysis, relation learning and information retrieval based on social media. We will implement the applications such as personalized recommendation, network public opinion analysis, dialogue system based on human-machine sentiment interaction, and cross-media search.