基于强关联规则与关键词的AI问数机器人训练算法  

AI Chatbot training algorithm based on strong association rules and keywords

作　　者：陈晓云 窦波 王鹤森 刘静 Chen Xiaoyun;Dou Bo;Wang Hesen;Liu Jing(Marketing Service Center,State Grid Xinjiang Power Supply Company,Xinjiang Urumqi,830063,China;State Grid Xinjiang Power Co.,Ltd.,Xinjiang Urumqi,830002,China)

机构地区：[1]国网新疆电力有限公司营销服务中心,新疆乌鲁木齐830063 [2]国网新疆电力有限公司,新疆乌鲁木齐830002

出　　处：《机械设计与制造工程》2025年第4期99-103,共5页Machine Design and Manufacturing Engineering

摘　　要：为提升AI问数机器人回答问题的准确性,采用Apriori算法挖掘AI问数机器人的问题关键词与数据库答案之间的强关联规则,结合IOGA算法形成增量化关联规则挖掘算法,通过结合部分历史数据和新增数据挖掘新增数据的关联规则,增强AI问数机器人回答问题的准确性,并根据标签推荐出最符合关键词的问题答案。实验结果表明:Apriori算法能够挖掘出更多的高支持度、高置信度的关联规则,并且能够更快挖掘出新增数据的关联规则,令AI问数机器人能够及时更新回答问题的准确性,体现了算法的有效性。To improve the accuracy of AI Chatbot answering questions,the Apriori algorithm is used to establish strong association rules between the question keywords of AI Chatbot and the database answers.Combined with the IOGA algorithm,an incremental association rule mining algorithm is formed.By combining some historical data and newly added data to mine the association rules of the newly added data,the accuracy of AI Chatbot answering questions is enhanced,and the most suitable question answers for the keywords are recommended based on the tags.The experimental results show that the Apriori algorithm can mine more high support and high confidence association rules,and can quickly mine association rules for newly added data,which enables the AI Chatbot to update the accuracy of answering questions in a timely manner,demonstrates the effectiveness of the method.

关 键 词：关联规则 AI问数机器人 增量化 APRIORI算法 IOGA 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]