E-SQL: Direct Schema Linking via Question Enrichment in Text-to-SQL

E-SQL is a novel pipeline for translating natural language queries into SQL (Text-to-SQL), addressing challenges such as complex schema and ambiguous queries.E-SQL enriches natural language questions by incorporating relevant database elements—such as tables, columns, and potential conditions—directly into the query to enhance SQL generation accuracy. Our pipeline also introduces Candidate Predicate Generation to further reduce SQL errors caused by incomplete or incorrect predicates.

Image Segmentation with R2U-Net

This project implements the Recurrent Residual Convolutional Neural Network based on U-Net (R2U-Net) from a research paper. Unlike the original study, we trained and tested our model on the Oxford-IIIT Pet dataset for the image segmentation task, a key area in computer vision.

A Cloud-Native Application: GelGit Travel

Gelgit Travel is a cloud-native application which showcases transportation ticket management and purchase platform. It is developed to learn and utilize fundamental AWS cloud services. In this post, I will walk you through our project and its cloud-based innovations.

Expanded DBTagger: Keyword Mapping for Linking Text to Database

A research extends the work initiated by DBTagger, an instrument that performs keyword mapping—a bottleneck step in the text-to-SQL translation for natural language queries. DBTagger relies on the labor-intensive process of manually annotating extensive data. To address this challenge, we present a novel system that incorporates weak supervision, significantly reducing the need for manual annotation. Our method leverages existing annotated data from DBTagger and extends it to a broader range of database schemas, including those in the comprehensive Spider benchmark. Additionally, we fine-tuned pre-trained models like BERT and T5 to enhance the system's ability to map keywords more effectively. This expanded dataset and improved methodology enables query interpretation across diverse domains without extensive manual labeling.

Updating DistilBERT Stance Detection Model for Streaming Data

I would like to give a little bit of detail about our basic stance detection project which is conducted for the CS533: Information Retrieval System course which is one of the course I have taken for this semester.

From Mechanical Engineering to Computer Science

In this pots, I am sharing my experinces and thoughts during transition from mechanical engineering to the computer science. What are the contributions of academic preparation program to me?