Files in this item



application/pdfMao562.pdf (439kB)
(no description provided)PDF


application/octet-streamMao562.epub (322kB)
(no description provided)Unknown


Title:An information extraction tool for microbial characters
Author(s):Mao, Jin; Moore, Lisa; Blank, Carrine; Cui, Hong
Subject(s):Phenotypic Character Extraction
Information Extraction
Natural Language Processing
Text Classification
Abstract:Automated extraction of phenotypic and metabolic characters from microbial taxonomic descriptions will benefit biology research and study. In this poster, we describe a Microbial Phenomics Information Extractor (MicroPIE) system. MicroPIE takes taxonomic descriptions in XML files as input and can extract 58 types of microbial characters. The main extraction steps are :1) splitting paragraphs into sentences; 2)predicting the characters described in the sentences by using automated classifiers; 3)extracting character values from the sentences by applying a variety of methods, such as Regular Expression Rule, Term Matching, and Unsupervised Semantic Parsing. Parts of the system have been implemented and currently been optimized for better performance. Results on optimizing the sentence classifiers show that the SVMs (Support Vector Machines) achieved better performance over the Naive Bayes classifiers, in addition, resolving the problem of unbalanced training instances helped improve the performance of SVMs.
Issue Date:2016-03-15
Citation Info:NA
Series/Report:IConference 2016 Proceedings
Genre:Conference Poster
Rights Information:Copyright 2016 is held by the authors. Copyright permissions, when appropriate, must be obtained directly from the authors.
Date Available in IDEALS:2016-03-08

This item appears in the following Collection(s)

Item Statistics