Extraction And Stratification Of Public Value(s) Preference

Public value(s)has become an important proposition in the post-new public management era.The proposal of public value provides the possibility to solve the lack of publicity in the public sector caused by the introduction of private sector management methods in the new public management and excessive pursuit of "performance".The academic community's research on public value is mainly focused on the research of public value connotation,public value generation,and public value conflict under the boundaries of result-oriented public value and consensus-oriented public values.Most of the existing public value researches start from western backgrounds or local cases,and the value items often have deficiencies in translation,context,and extrapolation.In the context of China,research on the overall public value of the whole society has become a significant issue.This study starts with sorting out the definition of public value(s),and defines the need for value with the social organization of people as the yardstick for judging the value and the degree of satisfaction of that need for the value subject.It is considered that public value(s)as a value is actually a construction process.Public value(s)preference is the basis for constructing public value.Identifying the common and public value preferences of different subjects in our society is helpful for further research on public value(s).In the era of big data,the involvement of computer technology makes large sample text analysis possible.With the help of natural language processing methods in computer science,the distributed representation of text is used to represent text features as word embedding vectors through word2 vec,and clustering is performed based on the obtained text feature vectors to obtain deep similarity relationships between texts.To achieve value extraction and preliminary clustering.Based on the classification of existing studies,this study avoids the loss of value preferences as much as possible during the classification stage,providing space for further stratification.Under the theory of social stratification,the value preferences of different subjects are further identified to understand the differences in value preferences among different subjects.At the same time,lexical attention analysis is performed inside each value,analyzing the value preference levels within each subject,trying to distinguish the internal levels for each subject's value preferences,and analyzing several enlightenments brought by the layering of public value(s).The innovation of this research lies in the use of large sample data in the Chinese context to identify public value(s)that are in line with national conditions,and the first discussion of the internal levels of public value(s)preferences.