Word Sense Disambiguation using case based Approach with Minimal Features Set

Journal Title: Indian Journal of Computer Science and Engineering - Year 2011, Vol 2, Issue 4

Abstract

In this paper we presented a case based approach for word sense disambiguation using minimal features set. To make the disambiguation, we took only two features for two different methods, post-bigram (immediate left word with ambiguous word – l1w) and pre-bigram (ambiguous word with immediate right word of it – wr1). To classify the cases for disambiguation, we followed three steps: instance or case identification from the distributed dictionaries, instance filtering based on the PoS and finally, case selection based on similarity measuring methods. Here, we used three different distant measuring functions, Euclidean, city-block and cosine methods. Set of reduced form of cases are treated as input for disambiguation. For making disambiguation, we applied K-nearest neighboring algorithm and artificial Neural Network. Among these two, KNN produced better disambiguation accuracy of 81.75% from cosine cases using pre-bigram features.

Authors and Affiliations

Tamilselvi P , S. K. Srivatsa

Keywords

Related Articles

AN ADVANCED SCALE INVARIANT FEATURE TRANSFORM ALGORITHM FOR FACE RECOGNITION

In computer vision, Scale-invariant feature transform (SIFT) algorithm is widely used to describe and detect local features in images due to its excellent performance. But for face recognition, the implementation of SIFT...

AN IMPROVED DOMAIN CLASSIFICATION SCHEME BASED ON LOCAL FRACTAL DIMENSION

In fractal image compression, most of the time during encoding is spent for finding the best matching pair of range-domain blocks. Different techniques have been analyzed for decreasing the number of operations required...

3D MODELLING FROM UN CALIBRATED IMAGES – A COMPARATIVE STUDY

3D modeling is a demanding area of research. Creating a 3D world from sequence of images captured using different mobile cameras pose additional challenge in this field. We plan to explore this area of computer vision to...

MSA AGENT FOR MULTIMEDIA APPLICATIONS

The MSA Agent is the software developed to design rectangular and U slotted micro strip antenna. It is applied for the various applications such as satellite communications, UHF applications. Within the multimedia freque...

Countering the DDoS Attacks for a Secured Web Service

Web services are the most widely used technology today. SOAP messages are the building blocks of web services. Since web services mostly deal with confidential data they must be protected from intruders. Intruders can be...

Download PDF file
  • EP ID EP124479
  • DOI -
  • Views 144
  • Downloads 0

How To Cite

Tamilselvi P, S. K. Srivatsa (2011). Word Sense Disambiguation using case based Approach with Minimal Features Set. Indian Journal of Computer Science and Engineering, 2(4), 628-633. https://europub.co.uk/articles/-A-124479