Design And Implementation Of The Smart Question Answering System For Postgraduate Enrollment Consultation

Postgraduate enrollment consultation is a crucial part of the enrollment process of major colleges and universities.Traditional postgraduate enrollment consultations usually use manual methods to deal with counseling issues.As the number of applicants increases annually,the pressure on counseling staff is gradually increasing.The major colleges and universities also set up an enrollment information Q&A community on the Internet,and organize the staff to answer the questions of students in a unified way.However,it is easy to cause problems accumulation since most of the consultation questions are similar and repeated.In view of the various problems in the process of enrollment consultation,this paper designs and implements a smart question answering system for postgraduate enrollment consultation.The specific work content is divided into the following points:1.Collect and analysis the data of relevant field,and divide the question and answer system structure into multiple layers,including interaction layer,platform layer and data layer according to the data characteristics of the postgraduate enrollment consultation question and answer and the functions required by the system.Design and implement an appropriate and efficient Q&A process to meet requirements of the smart Q&A function of graduate admissions counseling.2.Propose the W-Sen2vec vector representation algorithm based on the CBOW model of Word2vec,and adds text word order information in the process of training word vector to obtain richer sentence semantic information.The experimental results show that the vector generated by W-Sen2vec vector representation algorithm can be used in the topic filtering model,question clustering model and question classification model of the question and answer system to effectively improve the model effect and the accuracy of question and answer system.3.An automated model update process is proposed based on the architecture of the intelligent question answering system.The theme filtering model is used to filter and collect the new questions and answers generated by the user to ensure the purity of the corpus data.Automatic updating of models to ensure the effectiveness of the system model and further improve the accuracy of the question and answer system.At the same time,through the human-computer interaction interface,the combination of automatic question and answer and manual processing is used to solve the negative feedback of users and the lack of corpus of system answers.The method can also reduce labor costs and improve user satisfaction.This paper designs and implements the Postgraduate Enrollment Consultation Intelligent Question Answering System(PECIQA),which uses W-Sen2vec vector representation algorithm and model automatic update function to ensure the availability of the question and answer system.Combined the intelligent question-answering process with manual processing,the system can automatically return the answer of the natural language question proposed by user.The result shows that PECIQA can filter out irrelevant questions and answer the questions raised by the user correctly and reasonably.