A Hybrid Representation Model for Multi-Technique Citation Analysis Using a Unified Dataset

This article presents a hybrid model designed to perform various citation analysis operations on a collection of research papers linked through citations and references. Citation analysis is essential for assessing the influence and significance of papers in a specific field and for exploring their connections with other key works. These techniques also enable the mapping of relationships between domains, keywords, and other factors, based on the criteria selected for analysis. Unlike traditional approaches that focus on developing new methods to improve paper evaluation or ranking, this paper introduces a unified platform that integrates multiple citation analysis techniques, including co-citation analysis, bibliographic coupling, and citation density analysis. The article details the computational processes involved in developing this model, emphasising the importance of organising research papers by publication year and citation network data. Additionally, an algorithm is proposed to automate the extraction of relevant data for performing these analyses. The model’s effectiveness is demonstrated using a publicly available dataset, validating its capabilities in streamlining citation analysis.