Search-Engine-Simulator

A search engine simulator using algorithms of PageRank and Page scoring throughout computing the pages' click-through rate I was asked to build a small search engine that takes a search query, processing its writing way: whether it includes quotations or not, “AND” and “OR” words, and returns the most related and useful results to the user. Such a ranking of which the most useful and the most related depend on two algorithms I was asked to implement, which are the PageRank algorithm and then computing the overall score of each page that depends on the computed PR, and the Click-Through-Rate (CTR) that is mainly computed using the number of impressions of each page, and its number of clicks. As the program begins, I need to initialize the files that the user’s search will be done through. Four files are initialized and work as the big data base that all the processes into my program mainly depend on them. The first file is the web graph file that include all the web pages I have and describes how they all are connected to each other: Exe. Hyperlinks. Initializing this file is the backbone to build my web graph [Vertices→The Web Pages –Edges→ The Connections between them]. Building this web graph is very important in computing the Page Rank of each web page. The algorithm actually is dependent of it. The second file I need to initialize is the key words file. This file is important in connecting each web page to a few key words that user mainly will search for. Using the pair and vector data structures to read all these data into is done to be able to use them in the searching process. The last two files I need to initialize are the impression and the click files that show the number of impressions and clicks of each page. Reading these files into two maps data structure is very useful in computing the CTR and the overall ranking score of each page.