Evidence Finder

Evidence Finder extracts and indexes evidential sentences from approximately 1.5 million papers in the UKPMC/UPMC corpus integrated within Europe PMC using linguistic analysis and Named Entity Recognition (NER) to support retrieval of experimental and factual statements.

Key Features:

• Corpus Indexing: Indexes sentences from approximately 1.5 million papers in the UKPMC/UPMC corpus.
• Linguistic Analysis and NER: Applies linguistic analysis and Named Entity Recognition to identify and categorize entities and relevant sentence content.
• Evidence Identification: Systematically identifies sentences that contain evidential statements within research articles.
• Question-Based Results: Produces initial search outcomes as questions representing frequent types of evidence found in the indexed content.
• RESTful Web Services: Exposes article and grant access via RESTful web services (integration with Europe PMC).
• Citation-Count Sorting: Supports sorting search results by citation-count.
• Data Citation Features: Incorporates data citation features for search and result handling.

Scientific Applications:

• Evidence Retrieval: Enables retrieval of specific evidential statements across life sciences literature.
• Systematic Reviews and Meta-Analyses: Facilitates locating and aggregating evidence to support comprehensive reviews and meta-analyses.

Methodology:

Processes and indexes sentences using linguistic analysis and Named Entity Recognition (NER); systematically identifies sentences containing evidence; represents frequent evidence types as question-form outputs.