This project focuses on developing a model that can measure the semantic similarity between pairs of text while also estimating how confident the model is in its predictions. Traditional similarity models give fixed scores without indicating how reliable those scores are. To improve this, the project combines state-of-the-art deep learning models like BERT or SBERT with uncertainty quantification methods such as Bayesian Neural Networks and Monte Carlo Dropout. The goal is to create a system that not only compares text meaningfully but also provides interpretable confidence scores—useful in applications like question answering, document analysis, and information retrieval.
Dilrangi Sankalpana
I’m a fourth-year Computer Science and Engineering student at the University of Moratuwa, with a strong interest in Data Science and NLP. Right now, I’m working on my final year project, which focuses on measuring how similar two pieces of text are while also figuring out how confident the model is about its answer, as I am contributing to quantifying uncertainty
Brief Description of Project
Recent News
- All
