Web Mining for Social Network Analysis: A Review, Direction and Future Vision

Although web is rich in data, gathering this data and making sense of this data is extremely difficult due to its unorganised nature. Therefore existing Data Mining techniques can be applied toextract information from the web data. The knowledge thus extracted can also be used for Analysis of Social Networks and Online Communities. This paper gives a brief insight to Web Mining and Link Analysis used in Social Network Analysis and reveals the algorithms such as HITS, PAGERANK, SALSA, PHITS, CLEVER and INDEGREE which gives a measure to identify Online Communities over Social Networks. The most common amongst these algorithms are PageRank and HITS. PageRank measures the importance of a page efficiently with the help of inlinks in less time, while HITS uses both inlinks and outlinks to measure the importance of a web page and is sensitive to user query. Further various extensions to these algorithms also exist to refine the query based search results. It opens many doors for future researches to find undiscovered knowledge of existing online communities over various social networks. Keywords:Web Structure Mining, Link Analysis, Link Mining, Online Community Mining