Mobile Information Retrieval using Topic-Sensitive PageRank and Page Freshness

Abstract: Mobile users are on a rise in the past five years. Using of internet can be done anywhere and anytime because of mobile phone. PageRank Algorithm is used for ranking or sorting large number of hypertext documents. It can be used to tote up abundant sub graphs of the web with high efficiency. When the PageRank algorithm uses the large link arrangement of the web to retrieve the similarity of the required webpages, only a single vector comes as the outcome of the search query results. To overcome this we can use a group of PageRank vectors favoring a typical topic or subject matter. Using Topic sensitive PageRank a bit more accuracy can be obtained for the search-query. If the search is based on any catchword or keyword, the topic of the catchword or keyword is used by the topic sensitive PageRank to retrieve information accordingly. If the search is based on context, the PageRank scores of the topic can be figured out using the subject of the context. By grouping the above two based on linear combination, a more accurate ranking can be created, which will be better than a simple, basic PageRank vector. In our crawler we used Page Freshness and Age that retrieved the latest altered web pages compared to the old ones found in the remote date sources i.e., websites. This paper labels the techniques or methods for effectively retrieving information from web using mobile.

Keywords: Mobile computing, web search, information extraction, PageRank, page freshness, page age.