Using IBM Watson Discovery service in searching and retrieving information in libraries

Abstract

The activity of searching and retrieving information is one of the routine operations and tasks that users frequently perform in libraries. Despite this, many research tools have many problems, the most popular of it is the lack of conformity of the retrieved documents to the needs of the users expressed in their inquiries, which was confirmed by many studies, so this study seeks to monitor and test one of the most popular applications of artificial intelligence in searching and retrieving information which is the IBM Watson Discovery service. This study depends on descriptive analytical approach to describe and display the IBM Watson Discovery service and how is it work. The study also relies on the experimental approach to measure and test its ability of searching and retrieving information by including a set of documents. To test their retrieval by submitting a number of different inquiries. The researcher reached a set of results, the most important of which is providing a detailed definition of the concept of artificial intelligence, its origins, objectives and techniques for rooting it in the field of libraries and information sciences. With an explanation of the most important features available in the IBM Watson Discovery service. The study also revealed the credibility of the service in supporting the search in unstructured data such as PDF, Jason, word, html, Powerpoint files. As well as providing a detailed presentation of the method of testing and evaluating the retrieved search results, and the method of training Watson Discovery service to improve search results using machine learning technique, which contributes to improving the performance of the retrieval system.