Research
Multimodal Restaurant Review Sentiment classification
CUET NLP LAB, July 2024 -- Ongoing, Chittagong, Bangladesh.
Research area: NLP
- Developing a multimodal restaurant review dataset(containing text, image, emoji) for accurately express the sentiment.
Bsc. Thesis
Chittagong University of Engineering and Technology, Aug 2023 - June 2024, Chittagong, Bangladesh.
Title: Sentiment Analysis of Restaurant Reviews
- Developed a Restaurant review dataset(BRRD) containing 11004 Bengali restaurant reviews.
- Investigated performance of various ML, DL, and transformer based models for sentiment analysis of restaurant reviews from Bengali Text.
- Achieved remarkable results with 0.90 F1 score surpassing previous multiclass sentiment analysis restaurant review detecting models.
Experience
Research Fellow
Chittagong University of Engineering and Technology, July 2024 -- Ongoing, Chittagong, Bangladesh.
- Research Fellow at CUET NLP Lab.
- Developing a multimodal restaurant review dataset for accurately express the sentiment
- Developing a generalized Bengali dataset for sentiment classification including movie reviews, Airplane service.
Undergraduate Student Researcher
Chittagong University of Engineering and Technology, Jan 2023 - June 2024, Chittagong, Bangladesh.
- Collaborated in datasets tailored for low-resource languages:
- BEmoLex(Bangla Emotion Lexicon Dataset) for emotion detection
- Restaurant review dataset (BRRD) containing 11004 Bengali restaurant reviews of three types positive (5622), negative (4402), and neutral (980).
Software Engineer, R&D
Frontier Semiconductor Metrology(FSM), June 2023 - Present, Milpitas, CA.
- Developing complex software interfaces for the tools and machinery used for semiconductor manufacturing
- Design and develop machine control high-performance, multithreaded software for automated semiconductor equipment in C++.
- â—¦ Implemented OCR(Optical Character Recognition) to track the wafer ID from wafer Image to trace the wafer more accurately.
- Developing image processing algorithms for wafers using Python and C++ .
Software Engineer Intern
SELISE Digital Platforms, Oct 2022 – Nov 2022, Dhaka, Bangladesh.
- Designed and developed an E-commerce website using Angular, NodeJS, and MongoDB
Publications
- Md. Ariful Islam, Shawly Ahasan and Mohammed Moshiul Hoque. "TranSenA: A Transformer-based Framework for Sentiment Analysis of Restaurant Reviews". Accepted in International Conference on Signal Processing, Information, Communication and Systems(SPICSCON’2024).
Projects
HPC and image processing
Frontier Semiconductor Metrology(FSM), Sep 2023 - Jylu 2024, Milpitas, CA.
- Worked on a novel semiconductor metrology tool. It is capable of 2D & 3D measurement for flatness, roughness, thin film thickness, bump height, and warp.
- It is designed with a tilting mechanical stage, and different probes capable of auto focus, high speed nano scale measurement of silicon wafers.
- Besides, it has high speed machine vision cameras for massive amounts of image processing in real time with CUDA C++ and high-performance computing (HPC) capability with complex embedded systems for measurement.
- My contribution was to develop accelerated image processing algorithms, camera interfacing, and implement HPC capability.
Rock Paper Scissor
Chittagong University of Engineering and Technology, May 2021, Chittagong, Bangladesh.
- Created a real-time gesture recognition game using OpenCV and MediaPipe to classify hand gestures (rock, paper, scissors) for interactive gameplay. Optimized for smooth, real-time user interaction and dynamic responses.
CUET Bus Seat Booking System
Chittagong University of Engineering and Technology, Jan 2023 - April 2023, Chittagong, Bangladesh.
- Developed a web application for daily seat booking tailored for CUET students. Utilized HTML, CSS, and JavaScript for the front-end, while the back-end was implemented with PHP and MySQL
Emergency-Point
Chittagong University of Engineering and Technology, Aug 2022 - Nov 2022, Chittagong, Bangladesh.
- Multivendor website that helps users find all necessary medical equipment sorted according to their location. Google Map API is used besides HTML, Bootstrap, and JavaScript.
