Intelligent investment subject identification method and system based on map visualization

Abstract

The invention discloses an intelligent identification method and system for an investment subject based on map visualization, and the method comprises the following steps: intelligently judging the type label of the investment subject based on the research and judgment data of the investment subject; generating a weak early warning prompt in response to the fact that the type tag of the investment subject is inconsistent with the type tag selected by the user; positioning the investment subject on a map, and displaying a type label of the investment subject; and providing a map visualization service to allow a user to view the type labels of the investment subjects in the different investment areas of the map through a webpage or a mobile terminal. Therefore, big data analysis and artificial intelligence technologies can be utilized, and the advantages of map visualization are combined, so that the information of the investment subjects can be displayed more intuitively and comprehensively, and a user can intuitively know the distribution conditions of the investment subjects in different investment areas.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of investment subject intelligent identification, and more particularly, to an investment subject intelligent identification method and system based on map visualization. BACKGROUND

[0002] With the rapid development of the global economy, investment activities are becoming more frequent in various fields. In the electricity handling process, correctly identifying the investment subject can help determine the ownership of the business. Different investment subjects may involve different power demand, electricity nature, electricity price policy, etc., so the investment subject research and judgment can ensure that the business is correctly classified and processed. In addition, the investment subject research and judgment can help conduct compliance audits. According to the attributes and types of the investment subject, it can be determined whether it meets the requirements of relevant regulations, policies and standards, so as to ensure the compliance of the electricity handling process.

[0003] The existing investment subject identification method is usually based on a series of artificial rules, such as park boundary division, etc. In this process, manual intervention and confirmation are required, which increases the operation complexity and time cost. In addition, fixed artificial rules may not cover all cases, and the identification of investment subjects in specific cases may need to be adjusted flexibly. Therefore, an optimized investment subject intelligent identification method and system are expected. SUMMARY

[0004] To solve the above technical problems, the present application is proposed. The embodiments of the present application provide an investment subject intelligent identification method and system based on map visualization, which intelligently judges the type label of the investment subject based on the research and judgment data of the investment subject; generates a weak warning prompt in response to the type label of the investment subject being inconsistent with the type label selected by the user; locates the investment subject on the map and displays the type label of the investment subject; provides a map visualization service to allow the user to view the type label of the investment subject in different investment areas of the map through a webpage or a mobile terminal. In this way, big data analysis and artificial intelligence technology can be used to combine the advantages of map visualization to more intuitively and comprehensively display the information of the investment subject, so that the user can intuitively understand the distribution of the investment subject in different investment areas.

[0005] In a first aspect, an investment subject intelligent identification method based on map visualization is provided, which includes:

[0006] Intelligently judging the type label of the investment subject based on the research and judgment data of the investment subject;

[0007] Generating a weak warning prompt in response to the type label of the investment subject being inconsistent with the type label selected by the user;

[0008] positioning the investment subject on the map and displaying the type label of the investment subject;

[0009] providing a map visualization service to allow users to view the type labels of investment subjects in different investment areas of the map through a webpage or a mobile terminal.

[0010] In a second aspect, an investment subject intelligent identification system based on map visualization is provided, which comprises:

[0011] a type label determination module configured to intelligently determine the type label of the investment subject based on the research and judgment data of the investment subject;

[0012] a weak warning prompt generation module configured to generate a weak warning prompt in response to the type label of the investment subject being inconsistent with the type label selected by the user;

[0013] a type label display module configured to position the investment subject on the map and display the type label of the investment subject;

[0014] a type label viewing module configured to provide a map visualization service to allow users to view the type labels of investment subjects in different investment areas of the map through a webpage or a mobile terminal.

[0015] Compared with the prior art, the investment subject intelligent identification method and system based on map visualization of the present application intelligently determines the type label of the investment subject based on the research and judgment data of the investment subject, generates a weak warning prompt in response to the type label of the investment subject being inconsistent with the type label selected by the user, positions the investment subject on the map and displays the type label of the investment subject, and provides a map visualization service to allow users to view the type labels of investment subjects in different investment areas of the map through a webpage or a mobile terminal. In this way, big data analysis and artificial intelligence technology can be utilized to combine the advantages of map visualization to more intuitively and comprehensively display the information of investment subjects, so that users can intuitively understand the distribution of investment subjects in different investment areas. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0017] Figure 1 The flowchart of the investment subject intelligent identification method based on map visualization according to the embodiments of the present application.

[0018] Figure 2Flowchart of the sub-step of step 110 of the investment subject intelligent identification method based on map visualization according to the embodiment of the present application.

[0019] Figure 3 Block diagram of the investment subject intelligent identification system based on map visualization according to the embodiment of the present application.

[0020] Figure 4 Scenario schematic diagram of the investment subject intelligent identification method based on map visualization according to the embodiment of the present application. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0022] Unless otherwise defined, all technical and scientific terms used in the embodiments of the present application have the same meanings as those commonly understood by one of ordinary skill in the art of the present application. The terms used in the present application are only for the purpose of describing the specific embodiments of the present application, and are not intended to limit the scope of the present application.

[0023] In the description of the embodiments of the present application, it should be noted that unless otherwise specified and limited, the term "connection" should be understood broadly, for example, it can be an electrical connection, or a connection between two elements, it can be directly connected, or indirectly connected through an intermediate medium. For those of ordinary skill in the art, the specific meaning of the above-mentioned term can be understood according to the specific circumstances.

[0024] It should be noted that the terms "first", "second", and "third" involved in the embodiments of the present application are only to distinguish similar objects, and do not represent a specific order of the objects. Understandably, "first", "second", and "third" can be interchanged in a specific order or sequence as allowed. It should be understood that the objects distinguished by "first", "second", and "third" can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0025] Correctly identifying the investment subject helps determine the ownership of the business. Different investment subjects may involve different electricity demand, electricity nature, and electricity price policy. Through the research and judgment of the investment subject, the business can be correctly classified and processed, avoiding confusion and problems caused by unclear business ownership. Research and judgment of the investment subject can help compliance audit. According to the attributes and types of the investment subject, it can be determined whether it meets the requirements of relevant regulations, policies and standards, which helps to ensure that the power process is carried out within the legal framework and avoid risks and penalties caused by illegal behavior. Through the research and judgment of the investment subject, potential risk factors can be identified. Different investment subjects may have different risks such as credit risk and market risk. Research and judgment of the investment subject can help assess and manage these risks and take appropriate measures to reduce the likelihood of risk. Understanding the attributes and types of the investment subject helps optimize resource allocation. Different investment subjects may have different electricity demand and behavior characteristics. Through research and judgment, power supply planning can be better adjusted, power distribution can be optimized, resource utilization efficiency can be improved, and costs can be reduced.

[0026] Research and judgment of the investment subject is very necessary in the power process, which helps determine business ownership, conduct compliance audit, manage risks, and optimize resource allocation, so as to ensure the smooth progress and compliance of the power process.

[0027] However, the existing investment subject identification method is usually based on a series of artificial rules, such as park boundary division. These rules may be static and fixed, and cannot flexibly respond to complex and changing actual situations. The identification of investment subjects in specific situations may need to be adjusted flexibly, while artificial rules often cannot cover all situations. The existing method requires manual intervention and confirmation, increasing the complexity and time cost of operation. Manual operation is prone to misjudgment, omission and other problems, affecting the accuracy and consistency of the identification results. Sometimes the existing investment subject identification method may be limited by the completeness of information. If the relevant data is incomplete or inaccurate, it may lead to deviation of the identification results, affecting the accuracy of subsequent business processing. With the change of business environment, the existing investment subject identification method may be difficult to adjust and adapt to new situations, which may lead to inconsistencies between the identification results and the actual situation, affecting the subsequent business. The existing method lacks intelligence and automation, and cannot fully utilize advanced technologies to improve identification efficiency and accuracy. Intelligent technologies such as machine learning and artificial intelligence can help the identification model better adapt to changing situations.

[0028] The existing investment subject identification method has some obvious defects, including artificial rule limitation, manual intervention and confirmation, information incompleteness, difficulty in adapting to changes, and lack of intelligence and automation. Solving these defects requires the use of advanced technical means and methods to improve the accuracy, efficiency and adaptability of the identification method.

[0029] Figure 1 A flowchart of the method for intelligent identification of investment subjects based on map visualization according to the embodiment of the present application is shown in FIG. 1. As shown in FIG. 1, the method for intelligent identification of investment subjects based on map visualization comprises the following steps. Figure 1 110, intelligently judging the type label of the investment subject based on the research and judgment data of the investment subject; 120, generating a weak warning prompt in response to the type label of the investment subject being inconsistent with the type label selected by the user; 130, positioning the investment subject on the map and displaying the type label of the investment subject; 140, providing a map visualization service to allow the user to view the type label of the investment subject in different investment areas of the map through a webpage or a mobile terminal.

[0030] That is, by using big data analysis and artificial intelligence technology and combining the advantages of map visualization, the information of the investment subject can be more intuitively and comprehensively displayed, so that the user can intuitively understand the distribution of the investment subject in different investment areas.

[0031] In step 110, it is expected to use an intelligent natural language processing method to comprehensively utilize the text semantic information of the work order business type, the urban planning range, the provincial development zone range, the time information of the land use right taking, the electricity price information and the user extended attribute information, to mine the type information of the investment subject from them, and to intelligently judge the category of the investment subject.

[0032] Figure 2 A flowchart of the sub-step of step 110 in the method for intelligent identification of investment subjects based on map visualization according to the embodiment of the present application is shown in FIG. 1. As shown in FIG. 1, 110, intelligently judging the type label of the investment subject based on the research and judgment data of the investment subject comprises the following steps. Figure 2 111, obtaining the text description of the research and judgment data, wherein the text description of the research and judgment data comprises the work order business type, the urban planning range, the provincial development zone range, the time information of the land use right taking, the electricity price information and the user extended attribute information; 112, performing text preprocessing and text semantic understanding analysis on the text description of the research and judgment data to obtain the work order business text semantic feature vector, the urban planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector; 113, performing global correlation semantic interaction on the work order business text semantic feature vector, the urban planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector to obtain a semantic global correlation investment subject comprehensive representation feature vector; 114, determining the research and judgment result based on the semantic global correlation investment subject comprehensive representation feature vector.

[0033] Based on this, in the technical solution of the present application, based on the investment subject's research and judgment materials, the specific processing process of the type label of the investment subject is intelligently judged, including: first, obtaining the text description of the research and judgment materials, wherein the text description of the research and judgment materials includes work order business type, urban planning range, provincial development zone range, land use right acquisition time information, electricity price information and user extension attribute information. Here, the work order business type can reflect the specific business nature and demand of the investment subject, and different types of business may need different power resource support, so as to help determine the power demand and power consumption characteristics of the investment subject. The urban planning range can reflect the geographical position and planning range of the investment subject, and the planning policy and power supply situation of different areas may be different, which affects the power consumption mode and electricity price policy of the investment subject. The provincial development zone range can indicate the development situation and policy support degree of the area where the investment subject is located, which has a certain influence on the power consumption planning and policy application of the investment subject. The land use right acquisition time information can reflect the land use situation and development history of the investment subject, which affects the power demand and power consumption attribute of the investment subject. The electricity price information can directly relate to the power consumption cost of the investment subject, and different electricity price policies have important influence on the economic benefit and operation cost of the investment subject. And the user extension attribute information can include the scale, industry attribute, development stage and other information of the investment subject, which has important reference significance for determining the type and characteristics of the investment subject. By analyzing the text description of the research and judgment materials, the business characteristics, geographical position, power demand, development stage and other key information of the investment subject can be comprehensively understood, so as to help accurately research and judge the type of the investment subject, and provide important basis for the classification, processing and management of the investment subject.

[0034] In one specific embodiment of the present application, the text description of the research material is subjected to text preprocessing and text semantic understanding analysis to obtain a work order business text semantic feature vector, a town planning text semantic feature vector, a provincial development text semantic feature vector, a land use right taking text semantic feature vector, an electricity price semantic feature vector, and a user extension attribute text semantic feature vector, including: dividing the text description of the research material according to the text field dimension to obtain a work order business type text description, a town planning range text description, a provincial development zone range text description, a land use right taking time information text description, an electricity price information text description, and a user extension attribute information text description; and performing semantic coding on the work order business type text description, the town planning range text description, the provincial development zone range text description, the land use right taking time information text description, the electricity price information text description, and the user extension attribute information text description to obtain the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector, and the user extension attribute text semantic feature vector.

[0035] Next, the text description of the research material is divided according to the text field dimension to obtain a work order business type text description, a town planning range text description, a provincial development zone range text description, a land use right taking time information text description, an electricity price information text description, and a user extension attribute information text description; and the work order business type text description, the town planning range text description, the provincial development zone range text description, the land use right taking time information text description, the electricity price information text description, and the user extension attribute information text description are subjected to semantic coding to obtain a work order business text semantic feature vector, a town planning text semantic feature vector, a provincial development text semantic feature vector, a land use right taking text semantic feature vector, an electricity price semantic feature vector, and a user extension attribute text semantic feature vector. Here, dividing the text description of the research material according to different text field dimensions helps to finely process complex text information, extract key content under each field, and make data easier to analyze. Then, through semantic coding, the semantic information in the text description of each different text field dimension can be represented by a vector, and important semantic content in the text description can be captured.

[0036] During the above processing, the text description of the work order business type, the text description of the town planning range, the text description of the provincial development zone range, the text description of the time information of the land use right taking, the text description of the electricity price information, and the text description of the user extended attribute information are subjected to semantic coding, and the text semantic feature information of the text description of the work order business type, the text description of the town planning range, the text description of the provincial development zone range, the text description of the time information of the land use right taking, the text description of the electricity price information, and the text description of the user extended attribute information are extracted respectively, but lack of feature information exchange and interaction between each other. It can be understood that, in the actual application scenario of the present application, for the investment subject, the text semantic feature information of the text description of the work order business type, the text description of the town planning range, the text description of the provincial development zone range, the text description of the time information of the land use right taking, the text description of the electricity price information, and the text description of the user extended attribute information should not exist independently, and the correlation between the text descriptions under each text field is very important for accurately judging the investment subject. That is, by analyzing the semantic correlation between the text descriptions of different text fields, the internal relationship and influence between each other can be recognized. For example, the work order business type can be related to the town planning range. Generally speaking, a specific business type needs to be invested in a specific planning range to respond to the actual policy or planning purpose.

[0037] In one specific embodiment of the present application, the text description of the work order business type, the text description of the town planning range, the text description of the provincial development zone range, the text description of the time information of the land use right taking, the text description of the electricity price information, and the text description of the user extended attribute information are subjected to semantic coding to obtain the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector, and the user extended attribute text semantic feature vector, including: performing word segmentation processing on the text description of the work order business type to convert the text description of the work order business type into a word sequence composed of multiple words; using the embedding layer of the context encoder containing the embedding layer to map each word in the word sequence to a word vector to obtain a sequence of word vectors; and using the context semantic coding based on the global of the context encoder containing the embedding layer to obtain the work order business text semantic feature vector.

[0038] performing word segmentation on the text description of the urban planning range to convert the text description of the urban planning range into a first word sequence composed of a plurality of first words; mapping each first word in the first word sequence to a word vector using the embedding layer of the context encoder comprising the embedding layer to obtain a sequence of first word vectors; and performing global context semantic coding on the sequence of first word vectors using the context encoder comprising the embedding layer to obtain the urban planning text semantic feature vector.

[0039] performing word segmentation on the text description of the provincial development zone range to convert the text description of the provincial development zone range into a second word sequence composed of a plurality of second words; mapping each second word in the second word sequence to a word vector using the embedding layer of the context encoder comprising the embedding layer to obtain a sequence of second word vectors; and performing global context semantic coding on the sequence of second word vectors using the context encoder comprising the embedding layer to obtain the provincial development text semantic feature vector.

[0040] performing word segmentation on the text description of the land use right taking time information to convert the text description of the land use right taking time information into a third word sequence composed of a plurality of third words; mapping each third word in the third word sequence to a word vector using the embedding layer of the context encoder comprising the embedding layer to obtain a sequence of third word vectors; and performing global context semantic coding on the sequence of third word vectors using the context encoder comprising the embedding layer to obtain the land use right taking text semantic feature vector.

[0041] performing word segmentation on the text description of the electricity price information to convert the text description of the electricity price information into a fourth word sequence composed of a plurality of fourth words; mapping each fourth word in the fourth word sequence to a word vector using the embedding layer of the context encoder comprising the embedding layer to obtain a sequence of fourth word vectors; and performing global context semantic coding on the sequence of fourth word vectors using the context encoder comprising the embedding layer to obtain the electricity price semantic feature vector.

[0042] performing word segmentation on the text description of the user extension attribute information to convert the text description of the user extension attribute information into a fifth word sequence composed of a plurality of fifth words; mapping each fifth word in the fifth word sequence to a word vector using the embedding layer of the context encoder comprising the embedding layer to obtain a sequence of fifth word vectors; and performing global context semantic coding on the sequence of fifth word vectors using the context encoder comprising the embedding layer to obtain the user extension attribute text semantic feature vector.

[0043] In the technical solutions of the present application, in order to obtain more comprehensive semantic feature representation, the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector are processed by a text internal correlation global interaction module to extract and integrate the semantic correlation of each text description, so as to obtain a semantic global correlation investment subject comprehensive representation feature vector. The text internal correlation global interaction module introduces an internal interaction mechanism, so that the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector can interact based on global constraint guidance, so as to capture the semantic correlation and dependency relationship between each other.

[0044] In one specific embodiment of the present application, the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector are globally correlated and interacted to obtain a semantic global correlation investment subject comprehensive representation feature vector, including: the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector are processed by a text internal correlation global interaction module to obtain the semantic global correlation investment subject comprehensive representation feature vector.

[0045] Further, the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector are processed by a text internal correlation global interaction module to obtain the semantic global correlation investment subject comprehensive representation feature vector, including: the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right taking text semantic feature vector, the electricity price semantic feature vector and the user extended attribute text semantic feature vector are processed by a text internal correlation global interaction formula to obtain the semantic global correlation investment subject comprehensive representation feature vector; wherein the text internal correlation global interaction formula is:

[0046]

[0047] S = [s1; s2; s3; s4; s5; s6]

[0048] Wherein, E1 to E6 are the work order business text semantic feature vector, the town planning text semantic feature vector, the provincial development text semantic feature vector, the land use right text semantic feature vector, the electricity price semantic feature vector and the user extension attribute text semantic feature vector respectively, a ij And a kj Is the association matrix between the feature vector E i And the feature vector E j , the feature vector E i And the feature vector E k , the value of n is 5, s1 to s6 are the internal association work order business text semantic feature vector, the internal association town planning text semantic feature vector, the internal association provincial development text semantic feature vector, the internal association land use right text semantic feature vector, the internal association electricity price semantic feature vector and the internal association user extension attribute text semantic feature vector respectively, S is the semantic global association investment subject comprehensive representation feature vector, exp(·) represents the exponential function operation, [·;·] represents the cascade processing.

[0049] In one specific embodiment of the present application, based on the semantic global association investment subject comprehensive representation feature vector, the research and judgment result is determined, including: performing feature distribution clustering optimization on the semantic global association investment subject comprehensive representation feature vector to obtain an optimized semantic global association investment subject comprehensive representation feature vector; passing the optimized semantic global association investment subject comprehensive representation feature vector through a classifier-based investment subject research and judgment module to obtain the research and judgment result, which is used to represent the type label of the investment subject.

[0050] Then, the optimized semantic global association investment subject comprehensive representation feature vector is passed through a classifier-based investment subject research and judgment module to obtain a research and judgment result, which is used to represent the type label of the investment subject.

[0051] In an embodiment of the present application, the investment subject intelligent identification method based on map visualization further comprises a training step of training the text internal correlation global interaction module and the investment subject research and judgment module based on the classifier. The training step comprises: obtaining training data, wherein the training data comprises text descriptions of training research and judgment materials and real values of type labels of investment subjects; dividing the text descriptions of the training research and judgment materials according to the text field dimension to obtain text descriptions of training work order business types, text descriptions of training town planning ranges, text descriptions of training provincial development zone ranges, text descriptions of training time information of land use rights, text descriptions of training electricity price information and text descriptions of training user extension attribute information; performing semantic coding on the text descriptions of the training work order business types, the text descriptions of the training town planning ranges, the text descriptions of the training provincial development zone ranges, the text descriptions of the training time information of land use rights, the text descriptions of the training electricity price information and the text descriptions of the training user extension attribute information to obtain training work order business text semantic feature vectors, training town planning text semantic feature vectors, training provincial development text semantic feature vectors, training land use right taking text semantic feature vectors, training electricity price semantic feature vectors and training user extension attribute text semantic feature vectors; passing the training work order business text semantic feature vectors, the training town planning text semantic feature vectors, the training provincial development text semantic feature vectors, the training land use right taking text semantic feature vectors, the training electricity price semantic feature vectors and the training user extension attribute text semantic feature vectors through the text internal correlation global interaction module to obtain training semantic global correlation investment subject comprehensive representation feature vectors; performing feature distribution clustering optimization on the training semantic global correlation investment subject comprehensive representation feature vectors to obtain optimized training semantic global correlation investment subject comprehensive representation feature vectors; passing the optimized training semantic global correlation investment subject comprehensive representation feature vectors through the investment subject research and judgment module based on the classifier to obtain a classification loss function value; and training the text internal correlation global interaction module and the investment subject research and judgment module based on the classifier according to the classification loss function value.

[0052] In the technical solution, the training work order business text semantic feature vector, the training town planning text semantic feature vector, the training provincial development text semantic feature vector, the training land use right taking text semantic feature vector, the training electricity price semantic feature vector and the training user extension attribute text semantic feature vector respectively express the text semantic coding features of the text description of the training work order business type, the text description of the training town planning range, the text description of the training provincial development zone range, the text description of the training land use right taking time information, the text description of the training electricity price information and the text description of the training user extension attribute information. Thus, considering the inconsistent text semantic feature distribution caused by the inconsistent source text semantics of each text paragraph, although the long-distance context association representation is performed through the text internal correlation global interaction module, the training semantic global association investment subject comprehensive representation feature vector still has local feature distribution dispersion corresponding to each text semantic distribution.

[0053] Thus, when the training semantic global association investment subject comprehensive representation feature vector as a whole is classified through the classifier, the convergence difficulty of the predetermined class probability through the classifier for class regression is caused by the local feature distribution dispersion of the training semantic global association investment subject comprehensive representation feature vector, thereby affecting the training speed of the classifier and the accuracy of the final classification result.

[0054] Based on this, the applicant of the present application performs clustering optimization on the training semantic global association investment subject comprehensive representation feature vector, that is, firstly, the feature values of the training semantic global association investment subject comprehensive representation feature vector are clustered, for example, clustering based on the distance between feature values, and then the feature intra-class and inter-class representation after clustering is optimized, expressed as: the training semantic global association investment subject comprehensive representation feature vector is subjected to feature distribution clustering optimization to obtain an optimized training semantic global association investment subject comprehensive representation feature vector through the following optimization formula; wherein the optimization formula is:

[0055]

[0056] Wherein, f is each feature value of the training semantic global association investment subject comprehensive representation feature vector, n is the number of feature sets corresponding to the training semantic global association investment subject comprehensive representation feature vector, and k is the number of clustering features. Indicates a clustering feature set, f ′ is each feature value of the optimized training semantic global association investment subject comprehensive representation feature vector.

[0057] Specifically, by taking the intra-class features and extra-class features of the training semantic global correlation investment subject comprehensive representation feature vector as different instance roles to perform class instance description based on clustering proportion distribution, and introducing a clustering response history based on intra-class and extra-class dynamic context, a global perspective of keeping the intra-class distribution and extra-class distribution of the overall features of the training semantic global correlation investment subject comprehensive representation feature vector in coordination is maintained, so that the optimized feature clustering operation of the training semantic global correlation investment subject comprehensive representation feature vector can maintain the coherent and consistent response of intra-class and extra-class features, so that the regression convergence path based on feature clustering in the class regression process remains coherent and consistent, improves the convergence effect of the training semantic global correlation investment subject comprehensive representation feature vector facing the predetermined class probability, and improves the training speed of the classifier and the accuracy of the classification result.

[0058] In summary, the investment subject intelligent identification method based on map visualization according to the embodiments of the present application is illustrated, and the intelligent natural language processing method is used to comprehensively utilize the text semantic information of the work order business type, the urban planning range, the provincial development zone range, the land use right taking time information, the electricity price information and the user extension attribute information, to mine the type information about the investment subject, and to realize intelligent research and judgment of the category of the investment subject.

[0059] In an embodiment of the present application, Figure 3 The block diagram of the investment subject intelligent identification system based on map visualization according to the embodiments of the present application is shown. As Figure 3 shown, the investment subject intelligent identification system based on map visualization according to the embodiments of the present application 200 includes: a type label judgment module 210, configured to intelligently judge the type label of the investment subject based on the research and judgment data of the investment subject; a weak early warning prompt generation module 220, configured to generate a weak early warning prompt in response to the type label of the investment subject being inconsistent with the type label selected by the user; a type label display module 230, configured to position the investment subject on the map and display the type label of the investment subject; and a type label viewing module 240, configured to provide a map visualization service to allow the user to view the type label of the investment subject in different investment areas of the map through a webpage or a mobile terminal.

[0060] It can be understood that the investment interface automatic identification application construction realizes the classification of investment areas by developing the investment interface automatic identification function, developing the boundary drawing map visualization service, establishing an intelligent judgment model through planning park boundary identification, electricity price, user classification, user extension attribute and other parameters, to realize the purpose of automatic judgment of investment subjects, optimize government-enterprise cooperation, standardize investment interface, and strengthen business management and control, to promote the electricity application process to be more optimal, more efficient and lower in cost, and effectively improve the customer's sense of gain and satisfaction in applying for electricity.

[0061] Investment subject research and judgment of applicable business types include: low-voltage resident new installation, low-voltage resident capacity expansion, low-voltage non-resident new installation, low-voltage non-resident capacity expansion, high-voltage new installation, high-voltage capacity expansion, and temporary electricity installation.

[0062] In the field investigation link, after the selection of the electricity price and the saving, the system will automatically research and judge the investment subject according to the work order business type, whether it belongs to the urban planning range, whether it belongs to the provincial development zone, the time of taking the land use right, the electricity price, and the user expansion attribute. If the investment subject field is empty, the system research result will be written into the investment subject. If the investment subject has been selected before the electricity price is saved, the investment subject selected manually will be used as the criterion, and the investment subject can also be manually modified. If the automatically researched investment subject is inconsistent with the finally selected investment subject, the work order will give a weak prompt when it is submitted in the field investigation.

[0063] Here, those skilled in the art can understand that the specific functions and operations of each unit and module in the above-mentioned investment subject intelligent identification system based on map visualization have been described in detail above with reference to the description of the investment subject intelligent identification method based on map visualization Figures 1 to 2 , and therefore, the repeated description will be omitted.

[0064] As described above, the investment subject intelligent identification system based on map visualization 200 according to the embodiments of the present application can be implemented in various terminal devices, such as a server for investment subject intelligent identification based on map visualization, etc. In one example, the investment subject intelligent identification system based on map visualization 200 according to the embodiments of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the investment subject intelligent identification system based on map visualization 200 can be a software module in the operating system of the terminal device, or can be an application program developed for the terminal device; of course, the investment subject intelligent identification system based on map visualization 200 can also be one of the many hardware modules of the terminal device.

[0065] Alternatively, in another example, the investment subject intelligent identification system based on map visualization 200 and the terminal device can also be separate devices, and the investment subject intelligent identification system based on map visualization 200 can be connected to the terminal device through a wired and / or wireless network, and transmit interactive information in a conventional data format.

[0066] Figure 4 A scene diagram of the investment subject intelligent identification method based on map visualization according to the embodiments of the present application. As shown in Figure 4 , in this application scenario, first, the text description of the research and judgment data is obtained, wherein the text description of the research and judgment data includes the work order business type (for example, as shown inFigure 4 C1), a city planning range (e.g., as shown in FIG. 2A Figure 4 C2), a provincial development zone range (e.g., as shown in FIG. 2B Figure 4 C3), time information of land use right acquisition (e.g., as shown in FIG. 2C Figure 4 C4), electricity price information (e.g., as shown in FIG. 2D Figure 4 C5), and user extended attribute information (e.g., as shown in FIG. 2E Figure 4 C6); then, the acquired text description is input into a server (e.g., as shown in FIG. 3) in which a map-visualized investment subject intelligent recognition algorithm is deployed, where the server is capable of processing the text description based on the map-visualized investment subject intelligent recognition algorithm to determine a research and judgment result. Figure 4

[0067] It should also be noted that in the apparatus, device and method of the present application, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions of the present application.

[0068] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0069] Finally, it should be noted that the terms such as first and second, etc., are merely used to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between such entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or terminal device. Without more limitations, the element defined by the statement "including a" does not exclude the presence of additional identical elements in the process, method, article or terminal device including the element.

[0070] The above description is given for illustrative and descriptive purposes. In addition, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain modifications, alterations, changes, additions and sub-combinations thereof.​

Claims

1. A method for intelligent identification of investment entities based on map visualization, characterized in that, include: Based on the analysis data of the investment entities, the system intelligently determines the type label of the investment entities; A weak warning is generated in response to the inconsistency between the type label of the investment entity and the type label selected by the user. Locate the investment entity on the map and display its type label; Provide map visualization services to allow users to view the type labels of investment entities in different investment areas of the map via web pages or mobile devices.

2. The intelligent identification method for investment entities based on map visualization according to claim 1, characterized in that, Based on the analysis data of the investment entity, the system intelligently determines the type label of the investment entity, including: The text description of the assessment data includes the work order business type, urban planning scope, provincial development zone scope, land use right acquisition time information, electricity price information, and user extended attribute information. Text preprocessing and semantic understanding analysis are performed on the text descriptions of the assessment data to obtain semantic feature vectors of work order business texts, urban planning texts, provincial development texts, land use right texts, electricity price texts, and user extended attribute texts. Global semantic interaction is performed on the semantic feature vectors of the work order business text, the urban planning text, the provincial development text, the land use right text, the electricity price text, and the user extended attribute text to obtain a semantically globally associated comprehensive representation feature vector of the investment entity. The judgment result is determined based on the comprehensive characteristic vector of the semantically globally related investment entities.

3. The intelligent identification method for investment entities based on map visualization according to claim 2, characterized in that, The text descriptions of the aforementioned assessment data undergo text preprocessing and semantic understanding analysis to obtain semantic feature vectors for work order business texts, urban planning texts, provincial development texts, land use rights texts, electricity prices, and user extended attribute texts, including: The text descriptions of the analysis data are divided according to the text field dimension to obtain text descriptions of work order business types, urban planning scope, provincial development zone scope, land use right acquisition time information, electricity price information, and user extended attribute information. Semantic encoding is performed on the text descriptions of the work order business type, the urban planning scope, the provincial development zone scope, the land use right acquisition time information, the electricity price information, and the user extended attribute information to obtain the work order business text semantic feature vector, the urban planning text semantic feature vector, the provincial development text semantic feature vector, the land use right acquisition text semantic feature vector, the electricity price semantic feature vector, and the user extended attribute text semantic feature vector.

4. The intelligent identification method for investment entities based on map visualization according to claim 3, characterized in that, Semantic encoding is performed on the text descriptions of the work order business type, the urban planning scope, the provincial development zone scope, the land use right acquisition time information, the electricity price information, and the user extended attribute information to obtain the work order business text semantic feature vector, the urban planning text semantic feature vector, the provincial development zone text semantic feature vector, the land use right acquisition text semantic feature vector, the electricity price semantic feature vector, and the user extended attribute text semantic feature vector, including: The text description of the work order business type is segmented into words to transform it into a word sequence composed of multiple words. The embedding layer of the context encoder, which includes an embedding layer, maps each word in the word sequence to a word vector to obtain a sequence of word vectors; and The sequence of word vectors is subjected to global contextual semantic encoding using the context encoder containing the embedding layer to obtain the semantic feature vector of the work order business text.

5. The intelligent identification method for investment entities based on map visualization according to claim 4, characterized in that, Globally correlated semantic feature vectors are performed on the semantic feature vectors of the work order business text, the urban planning text, the provincial development text, the land use right text, the electricity price text, and the user extended attribute text to obtain a semantically globally correlated comprehensive representation feature vector of the investment entity, including: The semantic feature vectors of the work order business text, the urban planning text, the provincial development text, the land use right text, the electricity price text, and the user extended attribute text are used together through the text internal association global interaction module to obtain the semantically globally associated investment entity comprehensive representation feature vector.

6. The intelligent identification method for investment entities based on map visualization according to claim 5, characterized in that, The semantic feature vectors of the work order business text, the urban planning text, the provincial development text, the land use right text, the electricity price text, and the user extended attribute text are used through the text-internal association global interaction module to obtain the semantically globally associated investment entity comprehensive representation feature vector, including: The semantic feature vectors of the work order business text, the urban planning text, the provincial development text, the land use right text, the electricity price text, and the user extended attribute text are processed using the following text-internal association global interaction formula to obtain the semantically globally associated comprehensive representation feature vector of the investment entity; wherein, the text-internal association global interaction formula is: S=[s1;s2;s3;s4;s5;s6] Wherein, E1 to E6 are the semantic feature vectors of the work order business text, the urban planning text, the provincial development text, the land use right text, the electricity price text, and the user extended attribute text, respectively. ij and a kj E represents the eigenvectors. i and eigenvector E j eigenvector E i and eigenvector E k The correlation matrix between them has a value of 5 for n. S1 to S6 are the semantic feature vectors of the internal correlation work order business text, the internal correlation urban planning text, the internal correlation provincial development text, the internal correlation land use right text, the internal correlation electricity price text, and the internal correlation user extended attribute text. S is the semantic global correlation investment entity comprehensive representation feature vector. exp(·) represents the exponential function operation, and [·; ·] represents cascading processing.

7. The intelligent identification method for investment entities based on map visualization according to claim 6, characterized in that, Based on the semantically globally related investment entity comprehensive characterization feature vector, the judgment result is determined, including: The feature distribution clustering optimization is performed on the semantically globally related investment entity comprehensive representation feature vector to obtain the optimized semantically globally related investment entity comprehensive representation feature vector; The optimized semantically globally associated comprehensive feature vector of investment entities is passed through a classifier-based investment entity judgment module to obtain the judgment result, which is used to represent the type label of the investment entity.

8. The intelligent identification method for investment entities based on map visualization according to claim 7, characterized in that, It also includes a training step: training the text internal association global interaction module and the classifier-based investment entity judgment module.

9. The intelligent identification method for investment entities based on map visualization according to claim 8, characterized in that, The training steps include: Acquire training data, which includes textual descriptions of training and analysis materials, and the true values ​​of investment entity type labels; The text descriptions of the training and analysis data are divided according to the text field dimension to obtain text descriptions of training work order business types, training urban planning scope, training provincial development zone scope, training land use right acquisition time information, training electricity price information, and training user extended attribute information. Semantic encoding is performed on the text descriptions of the training work order business type, the training town planning scope, the training provincial development zone scope, the training land use right acquisition time information, the training electricity price information, and the training user extended attribute information to obtain the training work order business text semantic feature vector, the training town planning text semantic feature vector, the training provincial development text semantic feature vector, the training land use right acquisition text semantic feature vector, the training electricity price semantic feature vector, and the training user extended attribute text semantic feature vector. The training work order business text semantic feature vector, the training town planning text semantic feature vector, the training provincial development text semantic feature vector, the training land use right text semantic feature vector, the training electricity price semantic feature vector, and the training user extended attribute text semantic feature vector are used through the text internal association global interaction module to obtain the training semantic global association investment entity comprehensive representation feature vector. The feature distribution clustering optimization is performed on the trained semantic global association investment entity comprehensive representation feature vector to obtain the optimized trained semantic global association investment entity comprehensive representation feature vector; The optimized trained semantic global association investment entity comprehensive representation feature vector is passed through the classifier-based investment entity judgment module to obtain the classification loss function value; The classification loss function value is used to train the text internal association global interaction module and the classifier-based investment subject judgment module.

10. An intelligent identification system for investment entities based on map visualization, characterized in that, include: The type label determination module is used to intelligently determine the type label of the investment entity based on the analysis data of the investment entity; The weak warning prompt generation module is used to generate a weak warning prompt in response to the inconsistency between the type label of the investment entity and the type label selected by the user. A type label display module is used to locate the investment entity on the map and display the type label of the investment entity; The type label viewing module provides map visualization services, allowing users to view the type labels of investment entities in different investment areas of the map via web pages or mobile devices.

Images