Projects

project thumbnail

Conversational Qestion Generation using Large Language Models

Large Language Models | In-context Learning | Parametric Statistical Analysis

Research Project, IEC Lab, NCSU

Oct 2023 - Feb 2024

  • Proposes a three-stacked LLM model that utilizes in-context learning capabilities to generate context-aware conversational follow-up questions.
  • A/B experiments with 33 middle school students showed a 13% learning improvement compared to students who did not answer any follow-up questions.
project thumbnail

Generating Educational Explanations using Assertion-Enhanced Few-Shot Learning

Research Project, IEC Lab, NCSU

July 2023 - Dec 2023

  • Designed an effective prompting technique for Large Language Models (LLMs) to generate accurate explanations in problem-solving domains reducing hallucinations by 15% in an Intelligent Tutoring System when compared witrh traditional chain-of-thought prompting.
  • Evaluated and compared the quality of educational explanations through survey data analysis from in-service middle school teachers.
project thumbnail

Metadata Extraction from Educational Videos using Audio Transcript

Summer School Research Project, LearnLab, CMU

July 20, 2023 - July 28, 2023

  • This project aims to extract metadata from videos using audio transcripts to predict their usefulness based on popularity and educational values.
project thumbnail

Feature Engineering for Knowledge-building vs Knowledge-telling Responses

Research Project, IEC Lab, NCSU

May 20, 2022 - Feb 10, 2023

  • Performed inductive coding of response features indicative of knowledge-building vs knowledge-telling.
  • Trained a supervised classifier using Support Vector Machine.
project thumbnail

Itinerary Summarization with Temporal Information from Travelers’ Blog

Natural Language Processing project, NCSU

Nov, 2020 - Dec, 2020

  • The project is designed to generate summarized itineraries based on places and events extracted from traveller's blogs.
  • We used named-entity and event extraction to identify places convered by the traveller in the particular blog.
  • To differentiate between locations the traveler actually visited and those merely mentioned as memories or comparisons (e.g., "this place reminded me of..."), we used location-based tokenization. This method helps pinpoint which locations were visited during the current trip.