An Empirical Model for Information Retrieval System Evaluation: The User's perspective

Abstract

The aspect of time has become an issue since computers were introduced into management studies and thus highlights the need for efficient paradigm that will enable easy access to information, with much ease. But this is not the case yet, since finding the desired search engine and learning how to use them for this purpose is still time consuming. The continued dominance of measures for use in the system-centered aspect of IR evaluation, which is weak for use in user-centered IR system evaluation process, remains a challenge. Therefore the purpose of this study is to present usable measures through using an empirical from user's perspective for use in the evaluation of IR system. The survey method, a robust multivariate mathematical model and the factor analytic method were experimented upon. Results showed that the model presented is promising concerning the challenges highlighted. Therefore user's own knowledge, experience and searching abilities could be harnessed and implicated in IR design and evaluation. This study therefore serves as a test bed and guide to studies in this area, which results could contribute immensely to the overall improvement of the system. However, more data and a larger sample size are required to verify the proposed model in future, with other extraction techniques.