Exploring coversational agent for citizen's charter of cspc using pre-trained transformer model / Nestor B. Sayson Jr., Andrea Krystel T. Estadilla, and Jeland O. Albonial

By: Contributor(s): Material type: TextTextDescription: 132 leavesContent type:
  • text
Media type:
  • unmediated
Carrier type:
  • volume
Dissertation note: Undergraduate Thesis (BSCS) - Computer Science Camarines Sur Polytechnic Colleges, 2024 Abstract: This study explored the development of a conversational agent for the Citizen's Charter of Camarines Sur Polytechnic Colleges (CSPC), utilizing a pre-trained Transformer model. The agent aimed to simplify access to the charter's information, improving transparency and enhancing user interaction. The research focused on leveraging a pre-trained LLaMA-3 model to provide accurate, contextually relevant responses to queries related to CSPC's institutional processes, such as admissions, enrollment, and student services. The method-ology involved processing a dataset derived from the CSPC Citizen's Charter Handbook and official website. The evaluation of the agent's performance was carried out in three CSPC offices, with 33 respondents providing feedback on its usability. Results showed that the agent performed well in structured queries, with pass rates of 93.6% and 98.5% in the Institutional Planning and Development Unit and the Student Registration and Records office, respectively. However, the Office of Student Affairs and Services showed a lower pass rate of 86.9%, highlighting challenges with complex queries. The study concluded that the conversational agent provided a valuable tool for enhancing access to CSPC's in-stitutional information, with ongoing development needed to handle a wider range of user queries more effectively.
Tags from this library: No tags from this library for this title. Log in to add tags.

Undergraduate Thesis (BSCS) - Computer Science Camarines Sur Polytechnic Colleges, 2024

Includes bibliographic references.

This study explored the development of a conversational agent for the Citizen's Charter of Camarines Sur Polytechnic Colleges (CSPC), utilizing a pre-trained Transformer model. The agent aimed to simplify access to the charter's information, improving transparency and enhancing user interaction. The research focused on leveraging a pre-trained LLaMA-3 model to provide accurate, contextually relevant responses to queries related to CSPC's institutional processes, such as admissions, enrollment, and student services. The method-ology involved processing a dataset derived from the CSPC Citizen's Charter Handbook and official website. The evaluation of the agent's performance was carried out in three CSPC offices, with 33 respondents providing feedback on its usability. Results showed that the agent performed well in structured queries, with pass rates of 93.6% and 98.5% in the Institutional Planning and Development Unit and the Student Registration and Records office, respectively. However, the Office of Student Affairs and Services showed a lower pass rate of 86.9%, highlighting challenges with complex queries. The study concluded that the conversational agent provided a valuable tool for enhancing access to CSPC's in-stitutional information, with ongoing development needed to handle a wider range of user queries more effectively.

There are no comments on this title.

to post a comment.