Development and Application of an Automatic Question Answering System for Silkworm Disease Knowledge

Silkworm farmers need to quickly understand the relevant knowledge of silkworm disease, help the healthy growth of silkworms and increase the yield of silkworms. In order to assist silkworm farmers in their breeding work, this paper proposes and builds an automatic question and answer system for silkworm disease knowledge based on the Android mobile terminal. The system fills the blank of the automatic question answering system in the field of silkworm disease. The system adopts C/S mode architecture and develops the server based on Django framework. The server first realizes named entity recognition of silkworm disease names and symptom names based on the Bi-LSTM-CRF network, then implements the classification of silkworm disease questions based on the Multinomial Naive Bayes algorithm, and finally retrieves and generates answers in the graph database. The system test results show that the correct rate of questions and answers is 85.5%, and the average response time is within 1s. The high accuracy and low response time prove that the system has good application value.