Year: 2018
Type: article
Abstract: The huge challenge in discovering relevance feature is to determine the quality of user searched documents. The user wants relevant features to search the text, document, image, etc. approximately. The techniques earlier used where term based and pattern based. Now days clustering methods like partition based, density based and hierarchical is used along with different feature selection method. The term-based approach is extracting terms from the training set for describing relevant features. Partitioned text mining solves the low-level support problem, but it suffers from a large number of noise patterns. Information content in documents is identified by frequent sequential patterns and sequential patterns in the text documents and the useful features for text mining are extracted from this. Extracted terms are classified into three type’s positive terms, general terms and negative terms. In order to deploy advanced features on low-level features, this article finds positive and negative patterns in text documents.
Authors: R. R. Kamble, D. V. Kodavade
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W2887367262
