Abstract

This Capstone project explores the potential of Natural Language Processing (NLP) techniques in optimizing decision-making and organizational workflows across various domains. The Capstone project uses three case studies to demonstrate how sentiment analysis, frequency analysis, and topic modeling can analyze unstructured textual data to provide insightful findings. The first case study evaluates employee satisfaction using sentiment analysis, uncovering trends across departments and roles to guide targeted organizational interventions. The second case study focuses on student feedback at a higher education institution, using sentiment and frequency analyses to identify key areas for improvement in academic programs and services. The third case study leverages advanced topic modeling techniques to analyze thematic trends in artificial intelligence (AI) research over a decade, providing strategic insights into emerging innovations and priorities. The findings highlight the efficiency and scalability of NLP techniques, with automated processes completing tasks in seconds or hours that would otherwise take weeks or months manually. The research emphasizes the importance of selecting models and techniques, including embedding models, clustering methods, and preprocessing approaches that are specifically tailored to the task and organizational needs to ensure detailed, relevant, and easily interpretable outcomes. While limitations remain, including the need for end-to-end pipelines to make these techniques accessible to non-technical users, this Capstone demonstrates the transformative role of NLP in enabling organizations to harness the power of unstructured data for strategic planning, resource allocation, and innovation.

Publication Date

7-2025

Document Type

Senior Project

Student Type

Undergraduate

Advisor

Venera Demukaj

Advisor/Committee Member

Mimoza Polloshka

Advisor/Committee Member

Sara Baxley

Campus

RIT Kosovo

Download

Share

COinS