Ontology-based retrieval of bio-medical information based on microarray text corpora

Microarray technology is often used in gene expression exper-
iments. Information retrieval in the context of microarrays has mainly
been concerned with the analysis of the numeric data produced; how-
ever, the experiments are often annotated with textual metadata. Al-
though biomedical resources are exponentially growing, the text corpora
are sparse and inconsistent in spite of attempts to standardize the format.
Ordinary keyword search may in some cases be insucient to nd rele-
vant information and the potential benet of using a semantic approach
in this context has only been investigated to a limited degree. We explore
the possibilities of retrieving biomedical information from microarrays in
Gene Expression Omnibus (GEO), of which we have indexed a sample
semantically, as a rst step towards ontology based searches. Through an
example we argue that it is possible to improve the retrieval of biological
information.