Development of Book Analysis System: A Tool of Collection Assessment

Abstract

The study outlines the design, development and system evaluation of a rule-based Book Analysis System to support collection assessment in an educational library.  The was designed in response to the problems commonly encountered in manual assessment of library holdings, particularly in monitoring CHED and LEB standards concerning the adequacy, relevance and currency of library materials. Using a Modified Rapid Application Development (RAD) method, a web-based system was created with features such as outdated books identification, subject gap analysis, duplication detection feature and automated numerical report generation to assist the librarians in collection assessment and decision making process.
The system was pilot-tested across selected academic programs and evaluated through purposive sampling involving expert user. System performance was tested using a controlled validation dataset and assessed through the use of accuracy, precision, recall and F1 score. Study finding revealed full agreement between the system-generated results and expert validation within the pilot dataset, particularly in the identification of library material relevance and recency. Acceptability evaluation was also tested, it showed positive responses in terms of perceived usefulness, ease of use and overall user satisfaction.
Findings from the study show that the system has potential to assist academic libraries in conducting collection assessment more efficiently. Though the study is limited to a single institutional context and a bounded pilot dataset, it offers an initial framework that may be used to strengthen library compliance monitoring, improve efficiency in collection assessment, and can support evidence-based collection management practices in higher education libraries.