SIEM: An Integrated Evaluation Metric for Measuring Search Engine's Performance
Sojdeh Lotfipour, Fatemeh Ahmadi-Abkenari · International Journal of Computer Applications · 2014
Search engines as Web-based information retrieval applications traverse a database consisting of millions of Web documents upon receiving a user issued query.In order to evaluate the accuracy and strength of a search engine regarding its robustness in finding relevant Web documents, a set of metrics have been proposed by researchers that each of them evaluates one aspect of a search engine's performance.One of the existing challenges in this area is the lack of one measurement that could state a search engines performance from different perspectives.Some of developed metrics so far are general information retrieval evaluation measures that are not designed as specialized tools for search engine's evaluation.Some other metrics measure the system ability in finding accurate data while other metrics measures the speed of the system for performing the search process.In this paper different evaluation metrics such as precision, recall, f-measure, MAP, MRR, DCG and NDCG will be discussed.Then according to the conducted experiment and an analytical solution, a hybrid evaluation metric is proposed that based on it the overall strength of a search engine could be measured.