EVALUATION OF INFORMATION RETRIEVAL SYSTEMS USING VARIOUS MEASURES

Conference: Creative Trends in Engineering and Technology
Author(s): Princy Sisodiya, Vandana S Sardar Year: 2016
Grenze ID: 02.CTET.2016.1.14 Page: 9-15

Abstract

Our world revolves around technology and information. From a computer system present on desk to smart\nphone carried everywhere, the use of technology to aid human life has increased enormously. This leads to the\nproduction of massive amount of data; be it files belonging to an organization or a person\'s heartbeat rate. All data is\nstored. The main challenge is to retrieve information out of it. Additionally, a user specific information retrieval is also\nneeded. Information Retrieval Systems is one of the most used applications in today’s life, ranging from search engine\nsearching for a given query to intelligently analyzing and retrieving accurate details of a particular disease. Along with\npredefined retrieval items, a user can give a new query to the system and relevant information will be retrieved. Since,\nthe usage is wide; the need for evaluating such systems becomes a priority. Federated search is an information retrieval\ntechnology that allows the simultaneous search of multiple searchable resources and aggregates the results that are\nreceived from the search engines for presentation to the user. It has data for numerous queries and search engines. In this\npaper, various applications of Information Retrieval Systems are discussed, followed by different approaches used for\nthe evaluation. The dataset used is Federated Web Search track TREC 2014 of FedWeb Greatest Hits collection which\nallows combining results of multiple search engines. The methods used for evaluation along with the results are\nprovided.

<< BACK

CTET - 2016