CORE NLP MODULE - Me and teamates constructed the core NLP module including information extraction and topic detection module. Microsoft Text Analytics API and Spacy
were used to detect key phrases, and Google Knowledge Graph API was used for entity linking to
make sense of the potential knowledge entity types.
DATA ANALYSIS PIPELINE -Me and teamates built up data analysis pipeline that helps the team monitor and evaluates social bot performance via AWS S3, DynamoDB and Athena.
USER PROFILING - With millions of human-bot interaction scripts, I'm working on the user profiling process and ultinizing clustering to classify different user groups.