Enhanced Model to Improve Memory Based Learning Algorithm
Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2014, Vol 16, Issue 6
Abstract
Abstract: Opinion mining/Sentiment analysis is a field of research which focuses on tracking human opinionswritten in natural language. Many companies, Now-a-days, extract opinions from various internet sources (suchas review sites, blogs, twitter) and try to predict customer's reviews on their upcoming products. Parts-OfSpeechTagging or POS Tagging, being a crucial step in the process of opinion mining, is the assigning of eachword with its POS tag, according to its definition and context in the corpus. In this paper, we will first havedetailed study of a POS tagging algorithm i.e. Memory Based Learning Algorithm, along with other POStagging techniques and then, dealing with its limitations, we will propose an enhanced model in order tosignificantly improve the efficiency of Memory-Based Learning Algorithm. The Memory Based LearningAlgorithm, maintains a word-tag-feature repository, storing word with all its possible tags, along with its fixedwidth context called feature. During tagging, the algorithm retrieves/computes an appropriate tag, deduced byanalyzing feature value of the word. Thus, this algorithm is ineffective, when dealing with sparse data. To beable to deal with this limitation, we have devised an enhanced Memory Based Learning model. In addition tothis, we have compared the proposed model with other POS tagging algorithm, in terms of its performance.
Authors and Affiliations
Aastha Gupta
Keywords
Related Articles
- EP ID EP142511
- DOI -
- Views 131
- Downloads 0
Remote Android Assistant with Global Positioning System Tracking
The paper describes a method which binds a web application and an Android mobile application, to,provide remote exploration services. A remote exploration system has been developed using web server,database, and an Andro...
Classification of Mammogram Images for Detection of BreastCancer
Abstract: Breast cancer is the most commonly observed cancer in women both in the developing and thedeveloped countries of the world. The survival rate in it has improved over the past few years with thedevelopment...
Implementation of Hybrid Wavelet Transform for Adaptive Lossless Image Compression
Abstract: In digital image processing, Image compression is a type of data compression causing reduction in image size but maintain the image quality. It reduce cost and time in image storage and transmission and thuscan...
A Survey On Data Mining Techniques To Find Out Type Of Heart Attack
Heart disease is a major cause of morbidity and mortality in present society. Medicinal identification is extremely important but complicated task that should be performed precisely and proficiently. Although su...
The Cyberspace: Redefining A New World
Abstract: The cyberspace driven by information systems and the Internet is transforming our environment inextraordinary ways by enabling economic growth and providing new means by which people connect, interactand collab...