21-Oct-2016 Orges Koci “Identification & analyses of ontological terms.. "
Issued: Thu, 11 Feb 2016 21:35:00 GMT
Friday 21st October 12-1 pm
Graham Kerr Building, library
Orges Koci (Institute of Health and Wellbeing/School of Engineering) will be giving a talk, entitled "Identification and analyses of ontological terms in clinical literature” on the Friday the 21st of October at 12 pm in LT2.
The annotation of large amounts of data from literature with associated metadata can be utilized to reveal significant insights to researchers, however it can be extremely intensive when manually conducted. We have developed pyTag, an automated and time-efficient text-mining pipeline able to perform annotation of ontological terms in application area specific literature surveys along with statistical analysis for exploring associations within the data. PyTag integrates with EXTRACT , a text-mining engine that lists the environment type, organism name, tissue and disease mentions identified in a given piece of text and maps them to their corresponding ontology/entries. We have used the pipeline to annotate journal abstracts from clinical literature related to four different types of Inflammatory Bowel Diseases and have performed multivariate statistical analyses to investigate for patterns and relationships among them in both temporal and spatial terms.