Research And Application Of Intelligent Question Answering System On Social Security

"Made in China 2025" states that it is necessary to rapidly develop information technology,and intelligent customer service robots,as an important part of this field,have received great attention from all walks of life.With the popularization of computers and the rapid development of the Internet,the use of online consulting issues has also increased.The requirements of customers for products are not only the quality of products,but more customers have also put forward higher requirements for service quality.In recent years,the cost of artificial customer service has become higher and higher,the turnover rate has also increased year by year,and there have been problems such as long training periods.The intelligent customer service robot as a form of information technology,it can serve customers more efficient and efficient.This article uses web crawler technology to obtain relevant data in areas such as social security,and based on the Spark big data platform,uses WeChat public number as an interactive platform to construct an intelligent question answering system in the field of social security so that users can accurately and quickly obtain social security-related replies.,The main work is as follows:1.Build a Spark big data platform,write reptiles to obtain social security and other required corpus,and store the acquired data in big data platforms to accumulate material for the next study.2.Process the acquired data,first remove the dirty data and adjust the data format,and then use techniques such as word segmentation and word vector to build a corpus.3.Use the sentence classification,sentence matching technology to build the model,using a variety of algorithms for comparative analysis and select the most suitable model as the final model,the comparison of the model includes:in recent years is very popular convolutional neural network:CNN classification;Natural language processing is very suitable for the recurrent neural network:QA-Lstm question and answer model;traditional machine learning algorithm:TF-IDF+SVM;and the newer algorithm proposed in 2014:Doc2vec.4.We designed and implemented a WeChat public number,which provides an intelligent question answering system based on social security and evaluates the system from various aspects.The intelligent answering system for social security professional issues was established through artificial intelligence,Spark data platform and other technologies,which verified the feasibility and effectiveness of the proposed method.