Title: Context-based information retrieval from large heterogeneous data sources using semantics and polarity-based ranking

Authors: S. Subitha; S. Sujatha

Addresses: Department of Science and Humanities, Anna University, Chennai, India ' Department of Computer Applications, Anna University of Technology, BIT Campus, Trichy, India

Abstract: Information retrieval, in its naïve form is not suitable for the current information age, where all possible documents are available in digital formats. This paper presents an effective semantic and polarity-based information retrieval strategy for heterogeneous data sets. Context of the input query is identified and all the documents that satisfy the polarity and context of the input query are retrieved from the data source. Another major advantage of this method is that it can work on any data format, utilising the tags associated with the multimedia documents (image, audio and video) to identify the context. Experiments show effective retrieval and filtering rates, hence making this method a reliable and generic architecture for information retrieval.

Keywords: information retrieval; semantics; context; polarity; Big Data; data heterogeneity; text processing; tokenisation; sentiment analysis; Hadoop.

DOI: 10.1504/IJCAET.2017.086957

International Journal of Computer Aided Engineering and Technology, 2017 Vol.9 No.4, pp.497 - 507

Received: 22 Apr 2016
Accepted: 27 Aug 2016
Published online: 03 Oct 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article