基于LSTM的食品安全自动问答系统方法研究.PDFVIP

20 19 年7 月 农 业 机 械 学 报 第50 卷 增刊 doi:10.6041/ j.issn.1000鄄1298.2019.S0.058 基于LSTM 的食品安全自动问答系统方法研究 陈摇 瑛摇 陈昂轩摇 董玉博摇 赵筱钰摇 侯文俊 (中国农业大学信息与电气工程学院,北京 100083) 摘要:为高效、准确、全面获取食品安全相关信息,以食品安全文本为研究对象,采用Lucene全文检索架构和长短 期记忆神经网络(Longshort鄄termmemory,LSTM)构建了食品安全自动问答系统。 依托于从互联网爬取的文本作为 非结构化数据集,利用检索架构扩充人工标注的问题答案对规模,并以此训练了可以判断问题和答案候选句匹配 程度的LSTM模型。 基于Lucene检索机制进行答案候选集提取和基于LSTM模型进行答案提取,构建了一个可根 据食品安全相关问题给出答案所在句子的问答系统,并对比了基于Lucene直接检索的答案抽取和基于LSTM 的答 案抽取这两种方法。 结果表明,当候选文档数量增加时,基于 LSTM模型的问题答案匹配方法,其平均准确度始终 高于基于Lucene检索方法的平均准确度;而候选句子数量较小时,基于 LSTM 模型的问题答案匹配方法的平均准 确度也高于基于Lucene检索方法的平均准确度。 关键词:食品安全;问答系统;答案抽取模型;LSTM技术;深度学习 中图分类号:TP182;TS201郾6 文献标识码:A 文章编号:1000鄄1298(2019)S0鄄0380鄄05 Methods of Food Safety Question Answering System Based on LSTM CHEN Ying摇 CHEN Angxuan摇 DONG Yubo摇 ZHAO Xiaoyu摇 HOU Wenjun (College of Information and Electrical Engineering,ChinaAgricultural University,Beijing 100083,China) Abstract:Nowadays, food safety issues have been concerned by both governments and consumers. However,the increasing number of food safety related articles makes it difficult to retrieve useful information from the articles in a short time with high accuracy. In order to improve the efficiency and accuracy of accessing food safety information, a question answering system was proposed,which was based onlong short鄄termmemory (LSTM)andinformationretrievaltechniques. The systemreliedonthe food safety unstructured texts obtained by Web crawler technologies, and question answer pairs were selected by using Lucene,and LSTMwasusedtopredict answersaccordingto matching degreesbetween question and candidate sentences. Based on Lucene爷s retrieval mechanism and the LSTM model, a question answering system was built which can select sentences that were mo