1.1. Overview
Importance of Search Engine is irreplaceable in Web arena due to the ever increasing e-population and associated services. Searching and sorting most feasible and accurate result based on user requirement is highly desirable. Hence researchers all over the world are working and introducing more accurate and feasible searching and ranking mechanism in search engine environment. The main function of search engine is to process the users’ query, look for a match in its own database and present the matching results in a ranked manner. Different Web pages are directed to and from other Web pages in Web sphere. Rank calculation depends largely on the weightage of these inbounded and outbounded Web pages. Assessment of inbounded and outbounded Web pages is reflected in rank calculation by various methodologies.
Large data set is considered in effective ranking system. In many real time applications large amount of data is not available (Wangy et al., 2009). Implementation of probability estimation trees is required to introduce probability based ranking. An innovative unsupervised method for automatic sentence extraction using graph-based ranking algorithms is introduced in (Mihalcea, 2004). The idea of automatically using the concepts of a thesaurus to improve search results is implemented in (Silveira and Ribeiro-Neto, 2004). A new technique is proposed to personalize the results of a generic search engine (Freyne et al., 2004). Graph-based ranking method is proposed in Dom et al. (2003) to rank email correspondents according to their degree of expertise on subjects of interest.
A regularization-based algorithm called ranking adaptation SVM (RA-SVM) is proposed in Geng et al. as a unique ranking mode (2012). Different applications are also made based on the domain based ranking algorithm (Greeshma et al., 2013). Location based search procedure is introduced in Li et al. (2009). A framework is proposed in Klementiev et al. (2009) to learn the aggregate votes of constituent rankers with domain specific expertise without supervision.
Feature based ranking method is proposed in Guha et al. (2013). Different parameters are introduced and rank is calculated by some equations (Guha et al., 2013. Session is considered as an important parameter to evaluate the importance of a Web page (Guha et al., 2013, 2012). Relevancy of session is examined to reduce the probable noise (Guha et al., 2012).
Web page ranking method by analysing rank of different incoming and outgoing Web pages is proposed (Saxena et al., 2010). In this paper linked Web pages are considered as an important parameter to measure the rank of a Web page (Saxena et al., 2010).
Operational procedures of PageRank and HITS algorithms are compared in Devi et al. (2014). It is concluded in Devi et al. (2014) that PageRank algorithm gives importance on measured rank of a Web page and HITS algorithm concentrates on authority and hubness of a page. Mouse movement is used to detect user session and relevance of a Web page in Agarwal et al. (2016). Different Web page ranking algorithm works at the time of indexing or user query (Kumar, 2015).
Different page ranking methods are proposed based on different weighted factors, importance of connected links, mining techniques (Garg and Jain, 2015; Gupta et al., 2016).
In this section, necessary background for the rest of the paper is discussed.