The availability of increasingly wider repositories of biomedical and biological texts requires effective techniques to manage the huge mass of unstructured information there contained. The availability of ad-hoc document summaries, targeted to specific topics, may assist researchers in inferring previously undisclosed knowledge and in performing the biological validation of the results of data mining analysis. This paper presents BioSumm, a flexible framework which analyzes large collections of unclassified biomedical texts and produces ad-hoc summaries oriented to inferring knowledge of gene/protein relationships. Summary generation is driven by a novel grading function, which biases sentence selection by means of an appropriate domain dictionary.
|Title of host publication||8th IEEE International Conference on BioInformatics and BioEngineering, BIBE 2008|
|Publication status||Published - 2008|
|Event||8th IEEE International Conference on BioInformatics and BioEngineering, BIBE 2008 - Athens, Greece|
Duration: Oct 8 2008 → Oct 10 2008
|Other||8th IEEE International Conference on BioInformatics and BioEngineering, BIBE 2008|
|Period||10/8/08 → 10/10/08|
ASJC Scopus subject areas