Metalife Web Recognition Tool (WebRT)

Text mining is one of the most challenging modes in the field of Bioinformatics solutions. Going beyond the familiar systems that are dealing with information retrieval (which has the main goal of finding the query relative information) and data mining (which is about discovering patterns in structured data stored in databases), Metalife has developed sophisticated algorithmic procedures capable of information extraction, i.e. extraction of facts and structured information from unstructured text.

The Web Recognition Tool, being a light demo version of the Metalife text mining approaches, employs the assets of 24 ontology-based thesauri to identify semantically consistent and biologically meaningful entities in the text, to recognize synonyms, acronyms and spelling variations, classify recognized entities into domain-specific and scientifically significant categories, unify contextually equivalent terms and phrases based on elaborate normalization routines, at the same time distinguishing results for seemingly equal terms which have different categorization. Finally, the extracted information is presented in a structured and comprehensive manner, mapping results to the original text, thus providing the requisite input for more complex and knowledge generating semantic analysis.