Using of Text Mining in Online Public Access Catalog

Abstract

One of the products of the information society is an excessive amount of information generated by mankind, and consequently the ways to search, use, process, present and store this information. We can consider entering a keyword to search engine/search tool as a most common way of searching for information. The user consequently expects simple decision-making process when selecting an object from a found set of records. The main factor influencing the popularity and reusability of particular tool or service is interpretation of content of search results as well as form of their presentation. Libraries are still key sources of information as they make the database objects available (printed and electronic) through website and mobile applications. To search for objects they use online public access catalog (OPAC) that allows users to search and order objects from any computer or mobile device connected to the Internet. This study focuses on obtaining the most desirable bibliographic elements of users and subsequently on identifying actually provided number of keywords in the Slovak library. As this number was completely insufficient in decision-making process when selecting an object by user we design and test the possibility of using text mining to obtain keywords appropriate for filling the database in online public access catalog. Consequently, we compare these generated keywords with the real metadata elements in the Slovak Library database and identify the percentage of their conformity. Based on the results of the study we recommend creating algorithm by the specialists in the field of library and information science. This algorithm will assign a weight to individual words, for example according to their location within a document (title of the publication, chapter title, abstract, document text) and thus create keywords (the meta-data elements) in the database of online public access catalog.