Phrase Prioritization Algorithm with Supporting Data Structure and Its Implementation

Abstract

Search Engines are the most common data on the World Wide Web Although Search Engine Optimization deals with growing visibility of search results web pages, this study work focuses on the search algorithm that will provide the most appropriate, relevant and needed records with the anticipated search query. A comparative research of the intersection algorithm and its variants provided the requirement for the fresh algorithm. CRANTOP database is a conventional WWW database that not only provides a database of documents and queries, but also provides appropriate queries for real appropriate documents. The study was conducted on that database, available in the Appendix, and the analysis is based on the values of precision, recall and F-measure. The proposal algorithm, which preference of the search queries, is at the core of the study job. This algorithm indicates fetching phrases methodology and then looking for all possible sentences in order to increase the recall value. In this respect, the most significant problem is the use of such sparse matrix representation, which promotes the effective search of sentences into the papers. For doing it, sparse matrix representation was suggested, which contains Columns corresponding to the papers, making searching for any possible sentence effective. Rows Linked Representations preserve the vibrant nature of papers and insert and delete phrases from papers. In addition, emphasis was placed on general words and query words into the dictionary, resulting in precision search outcomes. Synonyms were also regarded for document retrieval, thereby enhancing comprehension.

Keywords: Phrase prioritization, Algorithm, Data structure, CRANTOP database, SEO, Search Engine, SPARSE