Systematic Analysis of Mendeley Documents with ChatGPT as a Coding Utility

The efficient management and organization of an extensive document library in the academic research reflects the ongoing development of knowledge and expertise. Since 2011, I have diligently accumulated a substantial collection of academic documents in Mendeley with diverse research interests. Several attempts have been made to extract meaningful insights from this extensive archive, with the most recent endeavor taking place in the summer of 2022. This study explores the use of ChatGPT as an assistant-style Chatbot to convert a vast collection of academic documents from Mendeley into insightful graphs, concise summaries, and other tasks tailored to user needs.
Moreover, this study can also be applied to alternative reference management tools such as Zotero. Additionally, it demonstrates practical scalability to handle the analysis of substantial number of publications in PDF format.

Flowchart for the proposed systematic analysis of Mendeley documents

Results and Discussion

The results of the analysis are presented through a series of visualizations, including bar plots depicting publication trends over the years, distribution of document types, and insights into prominent publishers and journals. Word clouds and interactive widgets offer an engaging visualization of user tags and enable detailed exploration of individual publications. Furthermore, the incorporation of ChatGPT for topic modeling provides a comprehensive understanding of the overarching themes within the document collection. The generated graphs and summaries offer a distilled representation of the knowledge landscape encapsulated in the Mendeley documents.

Conclusion

This study showcases the successful integration of ChatGPT as a utility for analyzing a sub- stantial collection of Mendeley documents. The use of advanced natural language processing techniques, coupled with dynamic visualizations, transforms the traditional approach to docu- ment analysis. As a result, researchers and academics can derive valuable insights from their document repositories, enhancing the efficiency and depth of knowledge extraction. In summary, the synergy between Mendeley and ChatGPT presents a compelling narrative of innovation and adaptability in the ever-evolving landscape of academic research. This project not only underscores the potential of AI-driven tools in document analysis but also serves as an inspiration for future endeavors aimed at unlocking the wealth of knowledge embedded in extensive document repositories. Future work may involve further optimization of the search algorithm, exploration of addi- tional natural language processing techniques, and extending the application to other domains. Moreover, integration of ChatGPT 4.0 plugins into the analysis pipeline could provide a dynamic and context-aware approach to understanding the nuances of the docume

References

1. D. Yang, J. Kleissl, C. A. Gueymard, H. T. Pedro, and C. F. Coimbra, “History and trends in solar irradiance and PV power forecasting: A preliminary assessment and review using text mining,” Solar Energy, vol. 168, pp. 60–101, 2018, Advances in Solar Resource Assessment and Forecasting. [Online]. Available: ScienceDirect

2. L. L. Wang, K. Lo, Y. Chandrasekhar, R. Reas, J. Yang, D. Burdick, D. Eide, K. Funk, Y. Katsis, R. Kinney et al., “Cord-19: The COVID-19 open research dataset,” ArXiv, 2020. [Online]. Available: ArXiv

3. Covid-19 open research dataset challenge (cord-19). [Online]. Available: Kaggle Dataset

4. ChatGPT. [Online]. Available: ChatGPT

5. I. E. Pratama. (2023) COVID EDA: Initial exploration tool. [Online]. Available: Kaggle

6. M. Ekin. (2023) COVID-19 literature clustering. [Online]. Available: Kaggle

7. Llama 2 is here - get it on Hugging Face. [Online]. Available: Hugging Face Blog

Citation

Abuella, M. (2024). Automating Academic Document Analysis with ChatGPT: A Mendeley Case. bioRxiv, 2024-03. Available at: https://www.biorxiv.org/content/10.1101/2024.03.18.585620v1