A topic-enhanced python question-answering model

With the development of large language model technology, the application of retrieval enhancement in the field of education has become one of the hot research directions, with the aim of alleviating the hallucination problem of large language models and improving the accuracy of large language models in answering educational questions. Questions in the field of education are usually more complex and highly personalized. When traditional retrieval methods are applied to educational questions and answers, they often have problems such as inaccurate semantic matching, insufficient context understanding, and difficulty in data processing, resulting in poor answer quality. To address the above challenges, this paper proposes a retrieval enhancement technology based on a neural topic model, which can effectively improve the accuracy of large language models in answering Python programming education questions. This technology reorders the retrieved external knowledge so that information that is more relevant to the question in the educational scenario is used to prompt the large language model to answer the question. Experimental results show that the Python question-answering model built based on the proposed topic enhancement technology generates higher-quality answers than the comparison models.