Application of Latent Dirichlet Allocation (LDA) to Identify Research Topics in Journal Publications

Abstract

Undergraduate and applied graduate students often face the challenge of selecting an interesting and relevant topic for their thesis topic, culminating in a scientific research journal. The conventional approach to topic selection involves reviewing several journal articles, a time-consuming process. Data science techniques, specifically Latent Dirichlet Allocation (LDA) topic modeling, are used to speed this up. LDA is a versatile technique for identifying common topics across different types of documents. This research utilizes LDA to identify the most common issues in a collection of journal articles related to the thesis topic. Topics with the highest coherence scores will be prioritized. The selected topics are then grouped to provide students with a list of the newest and most relevant areas of study for their thesis topic. The study results show that "Data Science" has become the most used topic in the last three years. Some of the resulting Data Science topics: data mining, image processing, data clustering, and expert systems. Several Network and Security topics are developed, including network devices and malware protection. This research aims to simplify the topic selection process and enable students to choose topics that align with current research trends, thereby encouraging more efficient and relevant research results.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.