Next-Generation Library Information Systems: Evaluating Native Multi-Model Database Technology

Abstract

Multi-model databases handles the structured, semi-structured, and highly connected data of various applications such as Healthcare, Library Information System and many more. This research paper put focus on two multi-model databases, one is native multi-model database, ArangoDB, and a Hybrid multi-model database integrating PostgreSQL, Couch DB, and Neo4j. The objective of this paper is to measure the performance of these multi-model using various evaluation criteria as such as execution time, throughput, indexing efficiency, and latency. Also, further it highlights on query designing and data retrieval efficiency showing better approach for library management environment. A real-world college library system which handle big data workloads was chosen to evaluate the performance of these multi-models. The results shows that hybrid multi-model can be adapted in cases where a stronger transactional reliability is required. In contrast, ArangoDB, performs more efficiently in cross-model queries, especially a single AQL query unified document, graph, and relational data retrieval, minimizing query orchestration and communication overhead. ArangoDB performs 47% better in execution time and 42% in throughput than Hybrid Model. Natural Language Processing (NLP) was used for query translation that enabled users to submit queries in plain English, which automatically transformed into structured database commands, improving accessibility and user experience. This research will help developers and researchers to design a better multi-model which is efficient for providing faster and more organized academic resources to students and faculties.