p.1129
p.1137
p.1141
p.1145
p.1149
p.1153
p.1157
p.1161
p.1166
Preprocessing in Biomedical Literature Mining Using Natural Language Processing
Abstract:
The number of biomedical literatures is growing rapidly, and biomedical literature mining is becoming essential. An approach for article processing in text preprocessing is proposed in order to improve the performance of biomedical literature mining. This approach combines the Web and corpus counts in order to eliminate the limitations of noise data of the Web. We experimentally showed that the performance of the combination models is the best comparing to the pure Web and corpus models. We achieve the best precision of 89.1% on all article forms and 88.7% article loss class.
Info:
Periodical:
Pages:
1149-1152
Citation:
Online since:
November 2014
Authors:
Keywords:
Price:
Сopyright:
© 2014 Trans Tech Publications Ltd. All Rights Reserved
Share:
Citation: