Extraction and Analysis of Social Network Data Using Text Mining Techniques

Abstract

Social network websites provide a means of communication between people who are located in different locations and this can be done by establishing a network in which the information such as text messages, pictures, audio and videos can be shared. The information that is stored in these websites will be in unstructured manner and hence it may lead to ambiguity such as lexical, semantic, syntax of data. Moreover the data set that is generated from the information pattern is more complex and difficult to analyze. Another problem with these kinds of websites is that there may be a chance of fraudulent activities carried out, such as creating fake profiles. The proposed technique implements pre-processing (tokenization) which removes irrelevant or redundant data which are stored in unstructured manner and then the hybrid classifier is used to detect the fraudulent activities carried out in social networking websites which in turn increases the performance of the system. Keyword: Text Mining, Classification, Social Network. Extraction I.INTRODUCTION Social networking websites provides an easy way for two way communication. Face book is one type of such network that is widely used by people for communication which is rich in texts that enable the user to create various text contents in the form of comments, wall posts, social media, and blogs. Due to the ubiquitous use of social networks in recent years, an enormous amount of data is available over the web. Text mining is used by most of the application in social networking websites that provides appropriate result for person-to-person interaction. Moreover, text mining techniques in conjunction with social networks can be used for finding a general opinion about any specific subject such as human thinking patterns and group identification in large-scale systems. In most of the social networking websites people use formal language for communication. Since informal communication is been carried out between different people and also it differs from person to person so there is a chance of different kinds of ambiguities are occurred such as lexical, syntactic, and semantic [1] of data. Therefore, forming a standard acceptable way for convenient communication is a critical task and can be solved by text mining.Text mining is an interactive technique that enhances computational intelligence which comprises of multidisciplinary fields. In information retrieval, text analysis, natural language processing the information is classification based on logical and non-trivial patterns from large data sets and many authors defines that text mining is also part of data mining technique. Data mining techniques are mainly used for the extraction of logical patterns from the structured database. Text mining techniques are very complex than data mining due to unstructured and fuzzy nature of natural language text. Most of researchers till today use decision trees and hierarchical clustering for group recommendations in facebook where the user can join the group based on simila...