Top-k Query For Relational Databases With Conditional Preferences

In daily life,user preference determines the choice,and the difference of user preference leads to the difference of user selection results,so the research on preference query in relational database has become a hot research topic.In specific applications,preferences can be divided into quantitative preferences and qualitative preferences.Quantitative preference is to express user preference in the form of specific numbers,that is to say,the specific choice of user is transformed into a numerical value.Although this method is simple and intuitive,not all problems can be expressed in the form of this function,especially in the face of automatic decision-making problems,so qualitative preference can be applied more widely than quantitative preference.CP-net is a simple and intuitive graphical preference representation tool,especially suitable for multiple attribute qualitative preference decision with dependent relationship between attributes in the case of incomplete information.Now,the processing of preference query is mainly to extend the Pareto compound mechanism to CP-net.Based on the "Lattice" framework theory,the Top-k query of conditional preference relational database is realized,and the query satisfying user preference is completed.In this paper,conducted the following research:(1)Pareto preference compound: for people undefineds multi-attribute preference,preference combination is the core of personalized success.The traditional Pareto composite only considers the preference relationship of preference,while CP-net has more complex characteristics,so it is necessary to expand Pareto composite to represent the preference relationship of preference.(2)Top-k query technology: the purpose of Top-k query is to retrieve the related keywords given by the user,and to extract the best k objects distributed in the specified area.In the field of database,the application of preference integration to database is an important research direction and an important preference query.Among them,Top-kquery represents the sort relationship of data,so it is also called order sensitive query.Thus,the search space of the database is effectively reduced and the query efficiency is improved.(3)Top-k query based on PCA dimensionality reduction technology: In real life,people often need to describe a large number of complex things and phenomena with multivariate.These variables are abstracted into high-dimensional data.High-dimensional data provides extremely rich and detailed information about objective phenomena,but on the other hand,the dramatic increase in data dimensions has brought unprecedented difficulties to subsequent data processing.In this paper,the PCA dimension reduction technology is combined with Top-k query technology to reduce the dimensionality of high-dimensional data before the Top-k query,thereby reducing the data dimension and improving the query rate.