The Mining Of People's Concern Based On Public Feedback And Suggestions From Government Platforms

Ensuring people's livelihood is the foundation of the country,and the data of people's livelihood can reflect people's concerns and the current situation of social public services.It is of great significance to accurately tap people's livelihood concerns and improve the precision of social governance ability for improving the efficiency of government governance and system intelligence.The balance between people's concerns and the supply of social function services is one of the difficult problems in current social governance.Paying attention to and obtaining people's concerns in time can promote the rationality of relevant decisions and make social resources equally and efficiently distributed.People's livelihood concerns are related to the people's reflection of clothing,food,housing and transportation,medical care,education,environmental protection,employment and other aspects,reflecting the current basic social problems.The traditional livelihood projects decided by experts and government departments in the prescribed procedures are reasonable in some ways,but they cannot be implemented on the basis of the concerns of the masses.In the current situation of information expression increasing and diversifying,the 12345 hotlines is not only the way for people to make suggestion.Official account of Wechat,Weibo,the websites of municipal government and municipal Party committee are also issued by relevant departments for people to make suggestions.It is very significant to extract people's livelihood hot spots from multi-source people's livelihood data and policy texts,and support local governments to invest resources in people's livelihood.This paper intelligently collects the data from public feedback and suggestions on the government platform,comprehensively uses the algorithms of big data,neural network and keyword extraction to collect and process the relevant data,and constructs the mining model to serves the resource allocation and economic construction of the national and local governments.For the purpose of this study,firstly,tens of thousands of multi-source livelihood data are collected from multiple platforms for preprocessing.And then,based on the LDA theme model,this study forecasts the field words and classifies the people's livelihood data.On the basis of LDA topic classification,the classification model is trained again by LSTM(Long Short-Term Memory)neural network to improve the classification accuracy and performance.After that,the data of three topic categories were selected and used by TF-IDF algorithm and the algorithm based on Word2 vec and Text Rank to extract the key words,and the feature indexes of keywords were extracted by constructing the coword network and comparing the two algorithms.At last,according to the content of key words and keyword groups,this paper takes education as an example to analyze and mine the focus of people's livelihood,and expounds the extracted focus of people's livelihood from practical problems.