A system and method for object definition

Abstract

The present invention discloses a system and method for object definition, belonging to the technical field of object definition. The system of the present invention includes: a model management module that nests and combines a derivative industry prediction model and a derivative industry identification model to generate a combined model; an object definition module, and the object definition module includes: an object acquisition unit that acquires statistical data information through a data interface; an object definition unit that sets a stripping coefficient for the statistical data information, accesses the combined model generated by the model management module with the statistical data information with the stripping coefficient set, performs operations on the combined model, acquires predicted or identified data information, and determines the derivative industry to which the data information belongs, thereby completing object definition. The present invention is more accurate in statistical summarization and more targeted for statistical survey work.

Description

Technical Field

[0001] The present invention relates to the technical field of object definition, and more particularly, to a system and method for object definition. Background Art

[0002] Object definition refers to selecting data sets that meet feature requirements from all data based on the data information obtained, setting stripping coefficients (weights) for specified data attributes, forming an object library, and conducting subsequent statistical work. Object definition is generally performed for cross-industry production industries.

[0003] At present, the definition of objects in the market is generally based only on data attribute information, and the definition method is mainly to screen data attributes. With the further development of national statistical work, object definition needs to rely on machine learning related algorithms to make the definition more targeted and flexible.

[0004] The existing technology relies solely on data attribute information to define objects, which is very limited and has low accuracy and work efficiency. Summary of the invention

[0005] In view of the above problems, the present invention proposes a system for object definition, comprising:

[0006] A model management module, wherein the model management module performs nested combination of the derived industry prediction model and the derived industry identification model to generate a combined model;

[0007] An object definition module, the object definition module comprising:

[0008] An object acquisition unit, wherein the object acquisition unit acquires statistical data information through a data interface;

[0009] The object definition unit sets a stripping coefficient for the statistical data information, connects the statistical data information with the set stripping coefficient to the combination model generated by the model management module, operates the combination model, obtains predicted or identified data information, and determines the derived industry to which the data information belongs, thereby completing the object definition.

[0010] Optionally, a derived industry prediction model is constructed based on a logistic regression classification algorithm and trained with derived industry data;

[0011] The model is as follows:

[0012]

[0013] Among them, k is the number of categories, x is the feature parameter, θ is the regression coefficient, and w is the weight;

[0014] The indicators of the derived industry data include: name, address, industry, scope, zoning, and main business activities; weights are added to the indicators.

[0015] Optionally, the derived industry identification model is identified through a clustering algorithm, and the clustering center is set to the number of derived industries. An arbitrary one is selected as the centroid point in the derived industry, and the Euclidean distance is used as the distance measure.

[0016] Optionally, the data interface includes: database interface, file interface, and XML interface.

[0017] Optionally, the object definition module further includes: an object display unit, which displays the object definition result in the form of a list or a graph, and saves the list and the graph.

[0018] The present invention also proposes a method for object definition, including:

[0019] Nest and combine the derived industry prediction model and the derived industry identification model to generate a combined model;

[0020] Obtain statistical data information through the data interface;

[0021] Set a stripping coefficient for the statistical data information, connect the statistical data information with the set stripping coefficient to the combined model generated by the model management module, perform operations on the combined model, obtain the predicted or identified data information, and determine the derived industry to which the data information belongs, thereby completing the object definition.

[0022] Optionally, the derived industry prediction model is constructed according to the logistic regression classification algorithm and trained with the derived industry data;

[0023] The model is as follows:

[0024]

[0025] Where k is the number of classifications, x is the feature parameter, θ is the regression coefficient, and w is the weight;

[0026] The indicators of the derived industry data include: name, address, industry, scope, zoning, and main business activities; weights are added to the indicators.

[0027] Optionally, the derived industry identification model is identified through a clustering algorithm, and the clustering center is set to the number of derived industries. An arbitrary one is selected as the centroid point in the derived industry, and the Euclidean distance is used as the distance measure.

[0028] Optionally, the data interface includes: database interface, file interface, and XML interface.

[0029] Optionally, the method further includes: presenting the object definition result in the form of a list or a graph, and saving the list and the graph.

[0030] Based on the definition using data attribute information, the present invention develops a model algorithm for object definition, creatively realizes the combined nesting of the model algorithms, and sets the stripping coefficient (weight) on this basis, making the object definition more reasonable, the statistical summary more accurate, and the statistical survey work more targeted. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a structural diagram of a system for object definition according to the present invention;

[0032] Figure 2 It is a flowchart of a method for object definition according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] Now, exemplary embodiments of the present invention will be described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely, and to fully convey the scope of the present invention to those skilled in the art. The terms in the exemplary embodiments shown in the drawings are not intended to limit the present invention. In the drawings, the same unit / element is denoted by the same reference numeral.

[0034] Unless otherwise specified, the terms (including scientific and technical terms) used herein have the ordinary meaning understood by those skilled in the art. In addition, it can be understood that the terms defined in the commonly used dictionary should be understood as having a meaning consistent with the context of their related fields, and should not be understood as idealized or overly formal meanings.

[0035] The present invention provides a system for object definition, as Figure 1 shown, including:

[0036] A model management module, which nests and combines the derivative industry prediction model and the derivative industry identification model to generate a combined model;

[0037] An object definition module, which includes:

[0038] An object acquisition unit, which acquires statistical data information through a data interface;

[0039] The object definition unit sets a stripping coefficient for the statistical data information, accesses the combined model generated by the model management module with the statistical data information with the stripping coefficient set, performs operations on the combined model, obtains the predicted or identified data information, and determines the derivative industry to which the data information belongs, thus completing the object definition.

[0040] Among them, the object definition module further includes: an object display unit, which displays the object definition result in the form of a list or a graph, and saves the list and the graph.

[0041] Among them, the derivative industry prediction model is constructed according to the logistic regression classification algorithm and trained with derivative industry data;

[0042] The model is as follows:

[0043]

[0044] Among them, k is the number of classifications, x is the feature parameter, θ is the regression coefficient, and w is the weight;

[0045] The indicators of the derivative industry data include: name, address, industry, scope, zoning, and main business activities; weights are added to the indicators.

[0046] Among them, the derivative industry identification model is identified through a clustering algorithm, and the clustering center is set as the number of derivative industries. Any one is selected as the centroid point in the derivative industry, and the Euclidean distance is used as the distance measure.

[0047] Among them, the data interface includes: a database interface, a file interface, and an XML interface.

[0048] The present invention will be further described below in conjunction with embodiments:

[0049] The system of the present invention includes: a model management module and an object definition module.

[0050] The model management module constructs relevant statistical models based on big data algorithms. Most standard big data algorithms in the market are built into the module, and stable statistical models are generated through training data and verification data for use in the object definition process.

[0051] Currently, the system includes a derivative industry prediction model and a derivative industry identification model, and supports the nested combination of models. More models can be independently trained and generated based on big data algorithms.

[0052] (1) The derivative industry prediction model.

[0053] This model is constructed based on the Logistic Regression classification algorithm and trained according to the currently known derivative industry data. The main characteristic parameters include indicators such as name, address, industry, scope, zoning, and main business activities. Weights can be set for each indicator. Since a piece of data information may belong to multiple derivative industry classifications, a multi-classification Logistic Regression model is adopted. The specific formula is:

[0054]

[0055] k represents the number of classifications, x represents the characteristic parameters, θ represents the regression coefficient, w represents the characteristic weights, and the calculation result of the model is the probability that a piece of data information belongs to a certain class.

[0056] If K is set to 3, which are the New Four Industries, Blue Economy, and New Culture Industry respectively. The specific characteristic parameters are as follows: for the name, containing mobile Internet, cloud computing, big data, Internet of Things, and mobile payment, with a weight of 0.8 and 0.2 for others; for the address, being the high-tech development zone, with a weight of 0.6 and 0.4 for others; for the zoning code, being the development zone, with a weight of 0.7 and 0.3 for others; for the main business being software development, manufacturing, and e-commerce, with a weight of 0.75 and 0.25 for others. For data with these characteristics, the model calculates that the probability of belonging to the New Four Industries is the highest, reaching 70%, the probability of belonging to the Blue Economy is 23%, and the probability of belonging to the New Culture Industry is 7%, which basically conforms to the actual situation of the data. For dynamically added data information (verification data), it can be directly input into the model to predict the derivative industry to which it belongs according to the data characteristics and dynamically adjust the current object definition result, greatly improving the efficiency of the statistical work in defining derivative industries.

[0057] (2) Derivative Industry Identification Model.

[0058] This model is identified through the clustering algorithm (K-MEANS). Set the clustering center K as the number of known derivative industries, and select one centroid point in each derivative industry. The distance metric uses the Euclidean distance, that is

[0059] “l

[0060] ;lkjhgfdx

[0061] Zxc vdan is the number of sample characteristics, and x, y are sample points.

[0062] The derivative industry identification model is to minimize the distance between each sample data X and the centroid point K, and finally converge to the centroid point K. Through training with training data, the model has relatively good results in clustering identification, and basically each category is a known derivative industry. For the identification of new derivative industries, set the clustering center as K+1. Select one centroid point from each derivative industry, and add a data without a derivative industry label as the centroid point. Input the data into the model to identify the new derivative industries. Manually judge the identification results, and by adjusting the parameters, more satisfactory results can be obtained, and the identification of new derivative industries can be achieved statistically.

[0063] (3) Model nested combination.

[0064] The so-called model nested combination means the combined application of multiple models in a certain order. For example, the derivative industry identification model and the derivative industry prediction model can be combined. Use the clustering result as the input of classification, reduce the amount of data input into the classification model, and make the prediction more efficient and accurate. Or the classification result can be used as the input of clustering to achieve more refined knowledge discovery on the basis of classification, such as which data in the four new economies are new technologies, which are new industries, which are new business forms, and which are new models.

[0065] The object definition module is based on the algorithm model to define the derivative industry to which the data information belongs, including an object acquisition unit, an object definition unit, and an object display unit.

[0066] (1) Object acquisition unit. Automatically acquire the data information provided by the statistical bureau through various methods such as database interfaces, file interfaces, and XML interfaces.

[0067] (2) Object definition unit. Based on the data information acquired by the object acquisition unit, by setting the stripping coefficient (weight) of the data, connect the data to the model, and automatically predict or identify the derivative industry to which the data information belongs to achieve object definition.

[0068] (3) Object display unit. Display the results of object definition in various ways such as rendering into lists and graphs, and save the results for statistical investigation use.

[0069] The present invention also proposes a method for object definition, as Figure 2 shown, including:

[0070] Perform nested combination on the derivative industry prediction model and the derivative industry identification model to generate a combined model;

[0071] Obtain statistical data information through the data interface;

[0072] For statistical data information, a stripping coefficient is set. The statistical data information with the stripping coefficient set is connected to the combined model generated by the model management module, and the combined model is operated to obtain the predicted or identified data information, and the derivative industry to which the data information belongs is determined, that is, the object definition is completed.

[0073] The object definition result is displayed in the form of a list or a graph, and the list and the graph are saved.

[0074] Among them, the derivative industry prediction model is constructed according to the logistic regression classification algorithm and trained with derivative industry data.

[0075] The model is as follows:

[0076]

[0077] Among them, k is the number of classifications, x is the feature parameter, θ is the regression coefficient, and w is the weight.

[0078] The indicators of the derivative industry data include: name, address, industry, scope, zoning, and main business activities; weights are added to the indicators.

[0079] Among them, the derivative industry identification model is identified through a clustering algorithm, and the clustering center is set as the number of derivative industries, and an arbitrary one is selected from the derivative industries as the centroid point, and the Euclidean distance is used as the distance measure.

[0080] Among them, the data interface includes: database interface, file interface, and XML interface.

[0081] Based on the definition based on data attribute information, the present invention develops a model algorithm for object definition, creatively realizes the combined nesting of the model algorithm, and sets the stripping coefficient (weight) on this basis, making the object definition more reasonable, the statistical summary more accurate, and the statistical survey work more targeted.

[0082] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript.

[0083] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce a means for implementing the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or one or more of the blocks.

[0084] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction means that implements the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or one or more of the blocks.

[0085] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or one or more of the blocks.

[0086] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present application.

[0087] Obviously, those skilled in the art can make various changes and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these changes and variations.

Claims

1. A system for object definition, the system comprising: A model management module, which nests and combines a derivative industry prediction model and a derivative industry identification model to generate a combined model; An object definition module, the object definition module comprising: An object acquisition unit, which acquires statistical data information through a data interface; An object definition unit, which sets a stripping coefficient for the statistical data information, accesses the combined model generated by the model management module with the statistical data information with the stripping coefficient set, performs operations on the combined model, acquires prediction or identification data information, and determines the derivative industry to which the data information belongs, thereby completing object definition; The derivative industry prediction model is constructed according to a logistic regression classification algorithm and trained with derivative industry data; The model is as follows: Where k is the number of classifications, x is the feature parameter, θ is the regression coefficient, and w is the weight; The indicators of the derivative industry data include: name, address, industry, scope, zoning, and main business activities; weights are added to the indicators; The derivative industry identification model is identified through a clustering algorithm, and the clustering center is set to the number of derivative industries, and any one is selected as the centroid point in the derivative industry, and the Euclidean distance is used as the distance measure; The data interface includes: a database interface, a file interface, and an XML interface; The nested combination of the derivative industry prediction model and the derivative industry identification model means that multiple models in the derivative industry prediction model and the derivative industry identification model are combined and applied in a certain order. During the combined application process, the clustering result is used as the input of classification or the classification result is used as the input of clustering.

2. The system according to claim 1, wherein the object definition module further comprises: An object display unit, which displays the object definition result in the form of a list or a graph and saves the list and the graph.

3. A method for object definition, the method comprising: Nest and combine a derivative industry prediction model and a derivative industry identification model to generate a combined model; Acquire statistical data information through a data interface; Set a stripping coefficient for the statistical data information, access the combined model generated by the model management module with the statistical data information with the stripping coefficient set, perform operations on the combined model, acquire prediction or identification data information, and determine the derivative industry to which the data information belongs, thereby completing object definition; The derivative industry prediction model is constructed according to a logistic regression classification algorithm and trained with derivative industry data; The model is as follows: Where k is the number of classifications, x is the feature parameter, θ is the regression coefficient, and w is the weight; The indicators of the derivative industry data include: name, address, industry, scope, zoning, and main business activities; weights are added to the indicators; The derivative industry identification model is identified through a clustering algorithm, and the clustering center is set to the number of derivative industries, and any one is selected as the centroid point in the derivative industry, and the Euclidean distance is used as the distance measure; The data interface includes: a database interface, a file interface, and an XML interface; The nested combination of the derivative industry prediction model and the derivative industry identification model is to combine and apply multiple models in the derivative industry prediction model and the derivative industry identification model in a certain order. During the combined application process, the result of clustering is used as the input of classification or the result of classification is used as the input of clustering.

4. The method according to claim 3, wherein the method further comprises: The object definition result is displayed in the form of a list or a graph, and the list and the graph are saved.

Images