A Relational Genomics Search Engine

We report on the development of a relational genomic search engine that integrates search of structured biological data and biomedical literature. After identifying an optimal preprocessing strategy, we present techniques for gene/protein term normalization, acronym expansion, and compound word normalization. Several retrieval models are evaluated on a large biomedical corpus using a common baseline. The system delivers results that meet or exceed the current state-of-the-art depending on the model used.

• Type: Conference
• Year: 2006
• Where: BIOCOMP
• Authors: Jay Urbain, Nazli Goharian
• Keywords: BIOCOMP 2006, BIOCOMP 2007, Compound Word Normalization, Gene/protein Term Normalization, Relational Genomic Search