FWPTM-ASIR: Frequent Word Pattern Taxonomy Model (FWPTM) based Aspect Semantic Information Retrieval System for Web Documents
The tourist were frequently besieged with lots of details and it identifies it as complex one to make use of the available information to make a decision regarding the tourist places to visit, this is because, huge opinions were available in the website. In order to recognize the positive and negative opinions, earlier various opinion mining methods were suggested. Different aspects were there in the opinion text, to focus that, aspect based opinion mining were brought-in. At present various aspects based opinion classification methods were available but only few work focus on the automatic aspect identification and extraction of implicit, infrequent, and co referential aspects. To solve this problem, Frequent Word Pattern Taxonomy Model (FWPTM) is proposed in this work. In the FWPTM larger patterns, in the taxonomy are usually more specific since they may be used only in positive documents. The semantic information will be used in the pattern taxonomy to improve the performance of using closed patterns in text mining. In terms of aspects based clustering, similar sentences of the users are clustered via the use of Fuzzy C Means with Particle Swarm Optimization (FCM-PSO) clustering algorithm which includes self-introductory lines of opinion holders. The proposed FWPTM with Aspect Semantic Information Retrieval System (ASIR) scheme is evaluated by using popular data sets RCV1 data collection and TREC topics implement. The experimental results show that the proposed FWPTM- ASIR attained better performance compared than existing information retrieval schemes.
Author's Name: P. Vijayakumar and S. Sukumaran
Volume: Volume 15
Issues: Issue 3
Keywords: Text Mining, Information Retrieval, Semantic Information, Pre processing, Feature Extraction, Frequent Word Pattern Taxonomy Model (FWPTM), Clustering.