KinderMiner web

KinderMiner web computes ranked associations between user-defined target terms and key phrases by text mining the PubMed literature to prioritize biomedical hypotheses and experimental targets.

Key Features:

• Automated Knowledge Extraction: Performs text mining of the PubMed corpus to extract co-occurrence and association signals from biomedical literature.
• Ranked Associations: Computes and ranks associations between a user-defined list of target terms and a key phrase using the KinderMiner algorithm.
• Flexibility Across Biomedical Topics: Accepts arbitrary biomedical topics and target lists, enabling application across diverse biomedical domains.
• Underlying Text Index: Operates on an indexed representation of biomedical literature to enable rapid association queries.
• Comparative Performance: Has demonstrated comparable or superior performance relative to other state-of-the-art text mining applications in flexibility and effectiveness.

Scientific Applications:

• Hypothesis Generation: Ranks literature-derived associations to prioritize candidate hypotheses for further investigation.
• Experimental Prioritization: Aids selection of experimental targets when lengthy wet-lab processes require prioritization based on literature evidence.
• Literature Summarization: Produces concise, ranked summaries of existing knowledge on specific biomedical topics via association scores.

Methodology:

Applies the KinderMiner algorithm to mine the PubMed corpus and compute ranked associations between user-defined target terms and key phrases using an underlying text index.