Project information
- Title: LLM-Enhanced Cross-Platform Web Search Application Combining AI and Traditional Search Techniques
- Supervisor: Professor Heming Cui
- Student: Sze Shing Fung (3035930060)
Project description
Recent advancements in Large Language Models (LLMs) have led to the development of AI-powered chatbots such as ChatGPT and Claude, which are now sometimes used for information retrieval in place of traditional search engines like Google. This has led to products such as Perplexity, which uses LLMs to specifically target the problem of web searching and information retrieval. However, chatbots face the problem commonly referred to as hallucinations, where they can potentially provide inaccurate information to unknowing users. Traditional search engines, on the other hand, present results as a list of website links. They are therefore constantly challenged by search engine optimization techniques and the proliferation of AI-generated content, which lead to degradations in the quality of search results.
To tackle these problems, I propose a cross-platform application that combines the strengths of LLMs and search engines. The proposed system will process multiple batches of search results, which will then be summarized and presented to the user in a traditional search engine format. This approach will help mitigate the hallucination problem by providing various perspectives and sources. The search results will be further enhanced using other technologies, including AI image generation, to provide a more interactive experience for users. The deliverables shall include a web application, a mobile application, and a supporting backend system.