Developing intelligent chatbots powered by large language models
In the rapidly evolving landscape of e-commerce, customer service stands as a pivotal component of consumer satisfaction and business success. This project centers on developing a system to assess the effectiveness of Large Language Models (LLMs) in augmenting the intelligence of chatbots for e-comm...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project (FYP) |
| Language: | English |
| Published: |
Nanyang Technological University
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/176199 |
| _version_ | 1811697681166237696 |
|---|---|
| author | Ng, Kai Teck |
| author2 | Lihui Chen |
| author_facet | Lihui Chen Ng, Kai Teck |
| author_sort | Ng, Kai Teck |
| collection | NTU |
| description | In the rapidly evolving landscape of e-commerce, customer service stands as a pivotal component of consumer satisfaction and business success. This project centers on developing a system to assess the effectiveness of Large Language Models (LLMs) in augmenting the intelligence of chatbots for e-commerce sellers. With the digital marketplace in Singapore experiencing significant growth, alongside an increase in consumer complaints regarding online transactions, there is an urgent need for more responsive and sophisticated customer service solutions. By tweaking and integrating LLMs, particularly OpenAI's GPT-3.5 Turbo, with a chat web application, this research aims to enhance chatbot effectiveness and user experience. The study employs a methodological approach involving the development of a basic web app interface using bubble.io, integration with multiple AI models via OpenRouter.ai, and advanced training with Pinecone.io to facilitate vector embeddings for improved chatbot responses. Results indicate that LLM-enhanced chatbots demonstrate great performance in problem-solving, understanding context and nuance, and multilingual capabilities, thereby offering significant potential to improve e-commerce chat experience by providing more accurate, contextually relevant, and personalized or customized user interactions. This project not only fills a gap in understanding the impact of pre-trained language models on chatbot performance but also proposes practical solution for e-commerce owners aiming to improve their customer interaction and satisfaction levels, ensuring Singapore remains at the forefront of e-commerce innovation and adaptability. |
| first_indexed | 2024-10-01T07:59:07Z |
| format | Final Year Project (FYP) |
| id | ntu-10356/176199 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:59:07Z |
| publishDate | 2024 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1761992024-05-17T15:44:08Z Developing intelligent chatbots powered by large language models Ng, Kai Teck Lihui Chen School of Electrical and Electronic Engineering ELHCHEN@ntu.edu.sg Computer and Information Science In the rapidly evolving landscape of e-commerce, customer service stands as a pivotal component of consumer satisfaction and business success. This project centers on developing a system to assess the effectiveness of Large Language Models (LLMs) in augmenting the intelligence of chatbots for e-commerce sellers. With the digital marketplace in Singapore experiencing significant growth, alongside an increase in consumer complaints regarding online transactions, there is an urgent need for more responsive and sophisticated customer service solutions. By tweaking and integrating LLMs, particularly OpenAI's GPT-3.5 Turbo, with a chat web application, this research aims to enhance chatbot effectiveness and user experience. The study employs a methodological approach involving the development of a basic web app interface using bubble.io, integration with multiple AI models via OpenRouter.ai, and advanced training with Pinecone.io to facilitate vector embeddings for improved chatbot responses. Results indicate that LLM-enhanced chatbots demonstrate great performance in problem-solving, understanding context and nuance, and multilingual capabilities, thereby offering significant potential to improve e-commerce chat experience by providing more accurate, contextually relevant, and personalized or customized user interactions. This project not only fills a gap in understanding the impact of pre-trained language models on chatbot performance but also proposes practical solution for e-commerce owners aiming to improve their customer interaction and satisfaction levels, ensuring Singapore remains at the forefront of e-commerce innovation and adaptability. Bachelor's degree 2024-05-15T00:35:25Z 2024-05-15T00:35:25Z 2024 Final Year Project (FYP) Ng, K. T. (2024). Developing intelligent chatbots powered by large language models. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/176199 https://hdl.handle.net/10356/176199 en A3031-231 application/pdf Nanyang Technological University |
| spellingShingle | Computer and Information Science Ng, Kai Teck Developing intelligent chatbots powered by large language models |
| title | Developing intelligent chatbots powered by large language models |
| title_full | Developing intelligent chatbots powered by large language models |
| title_fullStr | Developing intelligent chatbots powered by large language models |
| title_full_unstemmed | Developing intelligent chatbots powered by large language models |
| title_short | Developing intelligent chatbots powered by large language models |
| title_sort | developing intelligent chatbots powered by large language models |
| topic | Computer and Information Science |
| url | https://hdl.handle.net/10356/176199 |
| work_keys_str_mv | AT ngkaiteck developingintelligentchatbotspoweredbylargelanguagemodels |