XINJIE ZHANG 家具智能导购 ChatbotFurniture Shopping Assistant Chatbot
把家庭家具选购中的真实信息整理成一个可以访问和测试的智能导购原型。 A testable shopping-assistant prototype built from real product information in a family furniture retail context.
为什么做Why I made it
在北京大学"大模型：从基础到实战"暑期课程中，我希望让 Agent 概念落到一个熟悉的真实场景，因此选择家庭家具选购作为实践方向。 During Peking University’s summer course “Large Language Models: From Foundations to Practice,” I wanted to apply the idea of an AI agent to a familiar, concrete setting. I chose furniture shopping as the use case.
我做了什么What I did
我提出使用场景、整理需求，录入约 8 款真实产品信息，学习并参与部分 Python 实现，定义页面内容，测试回答准确性和幻觉问题，并整合课程报告。 I proposed the scenario, structured the requirements, entered information for around eight real products, learned and contributed to parts of the Python implementation, defined the page content, tested response accuracy and hallucinations, and integrated the course report.
AI 帮了什么How AI helped
AI 辅助编程、信息整理和页面实现。 AI supported programming, information organisation, and page implementation.
我负责的判断What I decided
我判断产品信息是否正确、回答是否可信、页面是否符合购物场景，并决定哪些结果可以保留。 I checked whether product information was accurate, whether responses were trustworthy, and whether the page suited a real shopping scenario. I also decided which outputs were usable.