Database-Specific Keyword Frequency Analysis in Merged Web Log Data: A Preprocessing Method

This study investigates the complex intricacies of web log data within the Electronic Resources module of the Perpustakaan Sultanah Bahiyah (PSB) website at Universiti Utara Malaysia (UUM). Serving as a cornerstone of academic infrastructure, the Electronic Resources module acts as a vital gateway, seamlessly connecting the UUM academic community to a vast repository of scholarly information. To tackle challenges posed by the size and complexity of web log data, the research employs a meticulous preprocessing method, involving the restructuring of raw data, outlier cleaning, and user session identification, laying the foundation for a comprehensive analysis. The study further explores the identification of search keywords embedded in the log file, employing a systematic process that transforms data into a structured format. The subsequent extraction of databases and keywords yields intriguing findings, prominently highlighting IEEE and Serial Solution databases. The analysis of 19,146 keywords associated with 11 databases offers valuable insights into user behavior, preferences, and the overall effectiveness of the Electronic Resources module. The identification of frequent keyword