TopEx

TopEx performs natural language processing to group semantically similar sentences and extract topics and keywords from textual corpora to identify underlying themes for biomedical and social media research.

Key Features:

• Semantic sentence grouping: Groups semantically similar sentences into clusters for downstream analysis.
• Topic analysis of clusters: Performs detailed topic analysis on sentence clusters to identify recurring themes.
• Keyword identification: Extracts key topics and keywords within collections of texts.
• Python library backend: Implemented as a Python library backend for programmatic processing.
• Corpus demonstrations: Applied to corpora including COVID-19 discharge summaries and tweets.
• Community evaluation: Evaluated in the BioCreative VII Challenge Track 4.

Scientific Applications:

• Biomedical corpus analysis: Identification of themes and topics in clinical texts such as COVID-19 discharge summaries.
• Social media analytics: Topic and keyword extraction from tweets to characterize trends and discourse.
• Corpus-level text exploration: Discovery of dominant topics and thematic patterns across large textual datasets.

Methodology:

Performs semantic grouping of sentences followed by detailed topic analysis and keyword extraction; implemented as a Python library backend.