Detecting temporal patterns of technical phrases by using importance indices in a research documents
Hidenao Abe, Shusaku Tsumoto · 2009
In text mining processes, temporal text mining have attracted considerable attention as an one of the important issues for finding remarkable terms with temporal patterns in temporal set of documents. Although importance indices of the technical terms play a key role in finding valuable patterns from various documents, temporal changes of them are not explicitly treated by conventional methods. Since those methods depend on particular index in each method, they are not robust in changes of terms. In order to detect remarkable temporal trends of technical terms in given textual datasets robustly, we propose a method based on temporal changes in several importance indices by assuming the importance indices of the terms to be a dataset. Our empirical study shows that two representative importance indices are applied to the documents from a research area. After detecting the temporal trends, we compared the emergent trend of the technical phrases to some emergent phrases given by a domain expert.