PUBLICATIONS

Ayman Asad Khan, Md. Toufique Hasan, Kai Kristian Kemell, Jussi Rasku, and Pekka Abrahamsson, "Developing Retrieval Augmented Generation (RAG) based LLM Systems from PDFs: An Experience Report", arXiv preprint, 2024.

This paper presents an experience report on developing Retrieval-Augmented Generation (RAG) systems using PDFs as primary data sources. It outlines the end-to-end pipeline, including data preprocessing, retrieval indexing, and response generation, while addressing technical challenges. The study provides practical insights for implementing RAG systems with OpenAI's API and open-source models like Llama, emphasizing accuracy, transparency, and dynamic information retrieval.
URL: https://arxiv.org/abs/2410.15944

Zeeshan Rasheed, Malik Abdul Sami, Jussi Rasku, Kai-Kristian Kemell, Zheying Zhang, Janne Harjamäki, Shahbaz Siddeeq, Sami Lahti, Tomas Herda, Mikko Nurminen, Niklas Lavesson, José Siqueira de Cerqueira, Md. Toufique Hasan, Ayman Khan, Mahade Hasan, Mika Saari, Petri Rantanen, Jari Soini, and Pekka Abrahamsson, "TimeLess: A Vision for the Next Generation of Software Development", arXiv preprint, 2024.

This paper presents TimeLess, a visionary system for next-generation software development, leveraging AI agents to transform traditional meetings into action-driven sessions. It aims to reduce development complexity, time, and cost by enabling real-time task execution during discussions, fostering efficiency, and enhancing collaboration in software engineering.

URL: https://arxiv.org/abs/2411.08507

Jinat Ara, Md. Toufique Hasan, Abdullah Al Omar and Hanif Bhuiyan, "Understanding Customer Sentiment: Lexical Analysis of Restaurant Reviews," EEE Region 10 Symposium (TENSYMP), 2020, pp. 295-299, DOI: 10.1109/TENSYMP50017.2020.9230712. 

Understanding customer's sentiment (satisfaction or dissatisfaction) is considered as valuable information for both the potential customers and restaurant authority. However, analyzing customer reviews (unstructured texts) one by one is a difficult task and also practically impossible when the number of reviews is enormous. Therefore, it seems conceivable to have a mechanism to analyze customer reviews automatically and provide the necessary information in a precise way. Here, we introduce a Natural Language Processing (NLP) based opinion mining methodology to analyze the customer opinion automatically. For that, first, a captive portal is used to collect customer's reviews. Then, the opinion mining technique is applied to analyze the reviews to explore customer sentiment about food quality, service, environment, etc. A data-driven experiment is conducted to evaluate the proposed methodology. The experiment result showed the effectiveness of the proposed method for retrieving and analyzing customer sentiment.

URL: https://ieeexplore.ieee.org/document/9230712