ISSN : 2319-7323

INTERNATIONAL JOURNAL OF COMPUTER SCIENCE ENGINEERING

ABSTRACT

Title	:	A Scalable Feature Extraction Technique for Social Media Analysis
Authors	:	Tauqeer Ahmad, Muhammad Rizwan Rashid Rana, Muhammad Aun Akbar, Asif Nawaz
Keywords	:	component; Data Mining; Features; Pointwise Mutual Information; Ant Colong Optimization.
Issue Date	:	Jul-Aug 2018
Abstract	:	Social media can be considered as a tool for expressing opinions and allow people to comment on a different topic. People especially youth are interested to use the different type of social media sites, messenger, blogs, microblogs etc. People comment positive and negative on published tweets both locally and globally. In social media it was very difficult to find out about specific event accrued in any part of the world. Feature extraction is now becoming the very active area of research. The data comes from various types of systems in the enterprise. Feature extraction technique used to reduce noisy data and increase the accuracy of the system. In the Past, many researchers work on feature extraction to improve semantic similarity between words using feature extraction techniques. These techniques which researchers used in the past not enough for better result and for improvement of accuracy of the system. With improving the old techniques, results should be improved. Semantic similarity between words can be identified through different feature extraction techniques by using social media sites. Results are taken on the basis of semantical way .so, The Proposed technique will be accomplish using Candidate Terms (Natural Language Tool Kit) and the Refining these Terms through (Pointwise mutual information) for Most suitable features ( ACO) at the end Final Features created and system accuracy increased 82% using these techniques.
Page(s)	:	197-201
ISSN	:	2319-7323
Source	:	Vol. 7, No. 4