Search for a command to run...
Abstract This paper introduces a smart recommendation system that extracts and analyzes technical papers from OnePetro to deliver targeted oil and gas knowledge via LINE messaging application. It uses Natural Language Processing (NLP) to surface insights on focused categories such as digitalization, well integrity, production optimization, sustainability, etc. The goal is to reduce research time, improve access to technical knowledge, and deliver new technology news to the audience. More than 2,000 abstracts from OnePetro are collected and processed through a Python-based pipeline utilizing Term Frequency-Inverse Document Frequency (TF-IDF) vectorization and K-means clustering. User profiles are built based on declared interests, click behavior, and category-matching using keywords and the user’s interaction history. A hybrid recommendation engine ranks abstracts for relevance based on content similarity, user preferences, and category-matching using Jaccard similarity. The system utilizes Google Apps Script as a database Application Programming Interface (API) to automatically connect to the LINE messaging application to send weekly summaries and links. Recommendations are filtered and delivered in a user-friendly chatbot format. Clustered modelling results revealed clear topic clusters and strong recommendation accuracy, supported by a category-matching mechanism using Jaccard similarity. Manual validation showed 82% precision at N=10 for top recommendations, and LINE notifications achieved a 59.6% average click-through rate (CTR), reflecting high user engagement and effective content filtering. The system reduces the time engineers spend manually searching for literature, particularly in design and brownfield operations, by delivering relevant content directly to their phones. It also helps surface emerging research themes, enabling early adaptation of new technologies and design practices. The architecture is designed for flexibility, with future enhancements planned for other database integration such as white papers published by oil and gas vendors, Society of Petroleum Engineers (SPE) podcast transcripts, and news/articles from public domain. The code, which is published as open-source on GitHub, creates opportunities for rapid, community-driven development. By combining technical insight delivery with automation and open collaboration, this project demonstrates a scalable, low-cost model for advancing digital transformation in offshore operations. This paper demonstrates a use case to integrate OnePetro data collection, NLP, and mobile messaging for personalized engineering literature delivery. The work offers a replicable model for engineers and developers to automate technical insight delivery, streamline literature review workflows, keep updated on oil and gas technology and news, and support decision-making. It contributes a practical, lightweight approach to digital transformation in the petroleum industry.