An Active Recommendation Approach to Improve Book-Acquisition Process

In the current book-acquisition recommendation process of libraries (e.g, university’slibrary), only a part of borrowers actively recommends purchasing books and the bookrecommendation mostly needs to be processed artificially; thus, most borrowers’requirements can not be satisfied and the book acquisition efficiency is unsound.Therefore, this paper attempts to develop a book-acquisition recommendation model andsystem based on text mining technology and Internet technology to provide librarians forsuggestions of book-acquisition. The proposed book-acquisition recommendation modelincludes three kernel modules namely Keyword Density Thesaurus (KDT), KeywordSequence Thesaurus (KST) and Keyword-Book Mapping (KBM) modules. The bookinquiry strings inquired from borrowers can be collected for keyword extraction via KDTand KST modules. After that, the extracted keywords are matching with the bookdatabase of bookseller to obtain the recommended books and the recommendation list forbook-acquisition can be generated via KBM module. In addition to the book-acquisitionrecommendation model, a Web-based book-acquisition recommendation system is alsodeveloped. Under the book-acquisition recommendation platform, the librarians canautomatically derive the book-acquisition recommendation list to fit borrowers’requirements and the complicated recommendation processes for borrowers can also besimplified. In brief, the book-acquisition recommendation process of this paper is ofsystem-based active recommendation and the book recommendation list doesn’t need to becollected artificially. Moreover, the generated book-acquisition recommendation list canmeet most borrowers’ requirements, and the efficiency and effectiveness of the library onbook-acquisition can be improved.