Development of an API for EDU segmentation

• Main Author: Lim, Patria Yun Xuan
• Other Authors: Sun Aixin
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering::Software::Software engineering
• Online Access: https://hdl.handle.net/10356/171834
• Institution: Nanyang Technological University

This development initiative focused on creating an Application Programming Interface (API) to facilitate the utilization of some research models developed for Elementary Discourse Unit (EDU) segmentation. The existing codebase was enhanced to ensure optimization and increased flexibility through the incorporation of customizable features. Consequently, the outcome of this endeavour was the creation and distribution of a Python package via the Python Package Index (PyPI). This package offers a range of segmentation capabilities, thereby enabling public access and simplifying the integration of these functionalities into various development projects or downstream activities. Additionally, the practical application of the EDU segmentation Python library was showcased within the context of an aspect-based sentiment analysis task. This task involved the evaluation of restaurant reviews sourced through the Yelp Fusion API. By effectively integrating the EDU segmentation model, this project successfully augmented its usability by introducing customization options and facilitating its accessibility through package distribution.