Intelligent management application system for power system

Abstract

The invention provides an intelligent management application system for a power system, which comprises an information identification module, a grade division module, an information evaluation module and an information terminal display module, and is characterized in that intellectual property information is divided into a key text and a conventional text, and based on the key text, a first key vocabulary is extracted; the method comprises the following steps: extracting a first keyword of intellectual property information, taking the first keyword as a keyword of intellectual property achievement retrieval to obtain a similar conversion achievement, calculating a predicted achievement coefficient of the intellectual property information based on the similar conversion achievement, extracting a second keyword based on a conventional text, performing retrieval by using the second keyword, performing similar value processing to obtain secondary information, and performing retrieval by using the second keyword based on the secondary information. And then combining the similar conversion result with the comprehensive similar value to determine a comprehensive risk value of the intellectual property information, thereby enhancing risk identification of enterprise intellectual property result assets, and effectively reducing the risk of an enterprise.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to an intelligent management application system for power systems. Background Art

[0002] With the emphasis on intellectual property rights, the proportion of intellectual property products in enterprises is increasing. At present, the power system is growing stronger in scientific and technological innovation and market competition, and intellectual property rights and achievements such as papers and patents are playing an increasingly important role in promoting the development of enterprises.

[0003] With the rapid accumulation of intellectual property and achievement information within power supply companies, asset evaluation and management of intellectual property achievement transformation is becoming increasingly important. If intellectual property achievements are blindly transformed through manual management, there will be low efficiency and potential risks and economic losses. Summary of the invention

[0004] The purpose of the present invention is to solve the problems in the background technology and to propose an intelligent management application system for power systems.

[0005] The present invention adopts the following technical solution: an intelligent management application system for a power system, comprising an information identification module, a level classification module, an information evaluation module and a terminal management module;

[0006] An information identification module is used to identify the intellectual property information in the enterprise information and transmit it to the classification module;

[0007] A level classification module is used to analyze the intellectual property information, determine the pre-emphasis combination by combining the font formats in the intellectual property information, and then determine the characteristic title according to the text information corresponding to the pre-emphasis combination and the position of the text information, and then identify the characteristic title to determine the characteristic protection area of ​​the intellectual property information, and divide the intellectual property information into a first level and a second level based on the characteristic protection area, and then the level classification module and the information evaluation module are connected in a one-way communication manner;

[0008] The information evaluation module specifically divides the intellectual property into key text and regular text based on the first level and the second level, and extracts the information keywords of the key text and the regular text respectively, and marks them as the first key word and the second key word. The first key word is firstly taken as the keyword for the search of intellectual property results to obtain similar transformation results. Based on the similar transformation results, the expected results coefficient is calculated. Then, the second key word is taken and used as the keyword for search to obtain matching information, and the information similarity value of the matching information is calculated, and the secondary information is determined. Then, according to the characteristic titles in the regular text, the regular information is divided into multiple different information paragraphs. At the same time, based on the information paragraphs and the secondary information, the comprehensive similarity value is determined. Then, the similar transformation results are combined with the comprehensive similarity value to determine the comprehensive risk value of the intellectual property information. Then, the information evaluation module and the terminal management module are connected in a one-way communication manner.

[0009] The terminal management module is used to calculate the comprehensive risk value of the intellectual property information in the enterprise information and display the comprehensive risk value of the intellectual property information on the terminal device.

[0010] As a further solution of the present invention, a method for determining a characteristic title includes:

[0011] Marking the identified intellectual property information as target information, and identifying the font format in the target information, wherein the font format combination includes font, font size, bold, and font italic;

[0012] The combination with the most font format combinations in the target information is taken as the basic combination, and the text corresponding to the basic combination is marked as the main text content, and the remaining font format combinations are marked as pre-emphasis combinations;

[0013] Get the text information corresponding to the pre-emphasis combination, identify the position of the text information in the target information, and then identify the content before and after the text information. If the text information is preceded and followed by characteristic symbols, mark the text information as a characteristic title. Otherwise, if there is a place before and after the information text that is not a characteristic symbol, mark the text information as the main content, where characteristic symbols include periods, colons, and spaces.

[0014] As a further solution of the present invention, a method for determining the first level and the second level includes:

[0015] Identify the characteristic title and determine the content title and paragraph title, wherein the content title refers to the title that specifically summarizes the overall content of the article in the target information, and the paragraph title refers to the title of the content to be expressed in a paragraph;

[0016] According to the paragraph title, determine the characteristic protection area of ​​the target information, which refers to the unique protection area in different intellectual property rights;

[0017] The feature protection area is then marked as the first level, and the remaining body content is marked as the second level.

[0018] As a further solution of the present invention, the calculation method of the expected achievement coefficient includes:

[0019] SS1: Extract the first-level information content and the second-level information content in the target information respectively, and mark them as key text and regular text;

[0020] SS2: Semantically split the information in the key text and regular text using natural language processing technology, and extract information keywords in the key text and regular text using the TF-IDF algorithm. Information keywords are phrases with TF-IDF values ​​greater than or equal to the threshold X1.

[0021] At the same time, the information keywords extracted from the key text and the regular text are marked as the first key words and the second key words respectively;

[0022] SS3: Take the first key word as the keyword for searching intellectual property achievements, search in the big data network, and mark the search results as similar transformation achievements;

[0023] Using the formula Calculate the expected achievement coefficient CG of the target information;

[0024] Among them, Ys is the actual income of similar transformation results within one cycle, Br is the input cost of similar transformation results, Tz is the investment time of similar transformation results, α1 and α2 are both proportional coefficients, and C is a fixed constant.

[0025] As a further solution of the present invention, if there are multiple similar transformation results in the detection results, the expected result coefficient of each similar transformation result is calculated respectively, and the average of all the expected result coefficients is taken as the expected result coefficient of the final target information. The lower the expected result coefficient, the lower the result transformation value of the corresponding intellectual property information. Conversely, the higher the expected result coefficient, the higher the result transformation value of the corresponding intellectual property information.

[0026] As a further solution of the present invention, a method for calculating the comprehensive similarity value includes:

[0027] Take the second key word, search the second key word in the big data, mark the search result as matching information, use the pre-trained Word2Vec algorithm to convert the matching information and the regular text into vectors, and use the cosine similarity algorithm to calculate the information similarity value between each matching information and the regular text;

[0028] Take the matching information whose information similarity value is greater than or equal to the similarity threshold X2 and mark it as secondary information;

[0029] Acquire characteristic titles in regular text, and divide the regular text into multiple information paragraphs based on the characteristic titles as dividing points;

[0030] Then, taking the information paragraph in the regular text as the unit, the similarity value between each information paragraph and the secondary information is calculated and marked as the paragraph similarity value, where i represents different information paragraphs, and i∈[1, I], indicating that I and information paragraphs coexist in the regular text;

[0031] Take the maximum value of the paragraph similarity value of each information paragraph and mark it as the similarity extreme value Zi;

[0032] Using the formula Calculate the comprehensive similarity value CL of the secondary information, where βi represents the weight coefficient of the feature title corresponding to information paragraph i.

[0033] As a further solution of the present invention, the calculation method of the comprehensive risk value is:

[0034] Combine the expected achievement coefficient of the target information with the comprehensive similarity value, and use the formula Fx = CL × θ1 + CG -1 ×θ2 to obtain the comprehensive risk value Fx of the target information, where θ1 and θ2 are both proportional coefficients.

[0035] In operation, the present invention analyzes the intellectual property information of an enterprise, determines the characteristic protection area of ​​the intellectual property information according to the information title in the intellectual property information, and divides the intellectual property information into key text and regular text according to the characteristic protection area, extracts the first key word based on the key text, and uses it as a keyword for searching the intellectual property results to obtain similar transformation results, and then calculates the expected result coefficient of the intellectual property information based on the similar transformation results, and at the same time, extracts the second key word based on the regular text, uses the second key word to search, and processes the retrieved information with similarity values ​​to obtain secondary information, and then determines the comprehensive similarity value based on the secondary information, and then combines the similar transformation results with the comprehensive similarity value to determine the comprehensive risk value of the intellectual property information, thereby enhancing the risk identification of the intellectual property results assets of the enterprise, effectively reducing the economic risk of the enterprise, and at the same time improving the efficiency of automatic risk processing and management of intellectual property information. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0038] Reference Figure 1 , an intelligent management application system for a power system, comprising an information identification module, a classification module, an information evaluation module and a terminal management module;

[0039] The information identification module is used to identify the intellectual property information in the enterprise information, wherein the intellectual property information includes papers, patents, various standards and group innovations in the enterprise, and in this embodiment, the enterprise refers to the power supply company, and then the information identification module transmits the identified intellectual property information to the level classification module;

[0040] It should be further explained that since the power supply company is divided into multiple departments and regions, such as county regional companies and city regional companies, the information identification module can identify the intellectual property information in all departments and regions through the company's internal system when identifying intellectual property information, thereby breaking down the information barriers between regions and departments and realizing cross-departmental and cross-level intellectual property information sharing within the company;

[0041] The level classification module is used to analyze the intellectual property rights and classify the information in the intellectual property rights into different information levels, wherein the information level includes a first level and a second level, and further, in the information level, the first level>the second level;

[0042] Among them, the methods for classifying information levels of intellectual property rights include:

[0043] S1: Mark the identified intellectual property information as target information, and first identify the characteristic titles in the target information;

[0044] The specific identification method of the characteristic title is: identifying the font format in the target information, wherein the font format combination includes font, font size, bold, font italic and other formats;

[0045] The combination of font format combinations with the most existence in the target information is selected as the basic combination, and the text corresponding to the basic combination is marked as the main text content, and the remaining font format combinations are marked as pre-emphasis combinations;

[0046] Obtain the text information corresponding to the pre-emphasis combination, and identify the position of the text information in the target information, and then identify the content before and after the text information. If the text information before and after are both feature symbols, mark this text information as a feature title. On the contrary, if there is a place before and after the information text that is not a feature symbol, mark this text information as the main text content;

[0047] Among them, characteristic symbols include period, colon, and space, etc.;

[0048] S2: Taking feature titles, and identifying the feature titles, determining content titles and paragraph titles, wherein the content title refers to the title that specifically summarizes the overall content of the article in the target information, and the paragraph title refers to the title of the content to be expressed in a paragraph;

[0049] Then, the characteristic protection area of ​​the target information is determined according to the paragraph title, where the characteristic protection area refers to the unique protection area in different intellectual property rights. For example, in a patent, the paragraph title is divided into claims, instructions, etc. At this time, the characteristic protection area of ​​the patent is the corresponding text content in the claims;

[0050] Then the feature protection area is marked as the first level, and the remaining text content is marked as the second level, and a one-way communication connection is established between the level classification module and the information evaluation module;

[0051] The information evaluation module is used to evaluate the target information and determine the comprehensive risk value of the target information. The specific method for determining the comprehensive risk value includes:

[0052] SS1: extract the first-level information content and the second-level information content in the target information respectively, and mark them as key text and regular text. Further, the key text is the first-level information content, and the regular text is the second-level information content;

[0053] SS2: Semantically split the information in the key text and the regular text using natural language processing technology, and extract the information keywords in the key text and the regular text using the TF-IDF algorithm. At the same time, the extracted information keywords in the key text and the regular text are marked as the first keyword and the second keyword, respectively. Furthermore, the first keyword is the information keyword in the key text, and the second keyword is the information keyword in the regular text.

[0054] Among them, natural language processing technology and TF-IDF algorithm are both existing technologies, which will not be described in detail here, and the information keyword is determined by the TF-IDF value, that is, the information keyword is a phrase whose TF-IDF value is greater than or equal to the threshold X1, and the specific value of the threshold X1 is obtained by those skilled in the art after big data calculation;

[0055] SS3: Take the first key word as the keyword for searching intellectual property achievements, search in the big data network, and mark the search results as similar transformation achievements;

[0056] Using the formula Calculate the expected achievement coefficient CG of the target information;

[0057] Among them, Ys is the actual income of similar transformation results within a cycle time, Br is the investment cost of similar transformation results, Tz is the investment time of similar transformation results, α1 and α2 are both proportional coefficients, C is a fixed constant, and the specific values ​​of α1, α2 and C are obtained by technicians in this field after big data calculation;

[0058] It should be further explained that the lower the expected achievement coefficient, the lower the achievement conversion value of the corresponding intellectual property information. On the contrary, the higher the expected achievement coefficient, the higher the achievement conversion value of the corresponding intellectual property information.

[0059] In another embodiment of the present invention, if there are multiple similar transformation results in the detection results, the expected result coefficient of each similar transformation result is calculated respectively, and the average of all the expected result coefficients is taken as the expected result coefficient of the final target information;

[0060] SS4: Take the second key word again, search for the second key word in the big data, mark the search results as matching information, use the pre-trained Word2Vec algorithm to convert the matching information and regular text into vectors respectively, and use the cosine similarity algorithm to calculate the information similarity value between each matching information and the regular text;

[0061] Among them, the Word2Vec algorithm and the cosine similarity algorithm are both existing technologies and will not be described in detail here;

[0062] SS5: Take the matching information whose information similarity value is greater than or equal to the similarity threshold X2 and mark it as secondary information;

[0063] Acquire characteristic titles in regular text, and divide the regular text into multiple information paragraphs based on the characteristic titles as dividing points;

[0064] Then, taking the information paragraph in the regular text as the unit, the similarity value between each information paragraph and the secondary information is calculated and marked as the paragraph similarity value, where i represents different information paragraphs, and i∈[1, I], indicating that I and information paragraphs coexist in the regular text;

[0065] Take the maximum value of the paragraph similarity value of each information paragraph and mark it as the similarity extreme value Zi;

[0066] Using the formula Calculate the comprehensive similarity value CL of the secondary information, βi represents the weight coefficient of the feature title corresponding to the information paragraph i, further, the specific value of βi is obtained by the technicians in this field after big data calculation, and before calculating the similarity value between each information paragraph and the secondary information, it is necessary to convert the information paragraph and the secondary information into vectors using the Word2Vec algorithm, and then calculate the similarity value using the cosine similarity algorithm;

[0067] SS6: Then combine the expected achievement coefficient of the target information with the comprehensive similarity value, using the formula Fx = CL × θ1 + CG -1 ×θ2 to obtain the comprehensive risk value Fx of the target information, θ1 and θ2 are both proportional coefficients, and the specific values ​​of θ1 and θ2 are obtained by technicians in this field after big data calculation;

[0068] There is a one-way communication connection between the information evaluation module and the terminal management module;

[0069] The terminal management module is used to calculate the comprehensive risk value Fx of the intellectual property information in the enterprise information, and display the comprehensive risk value of the intellectual property information on the terminal device for relevant staff to view.

[0070] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An intelligent management application system for power system, characterized in that: It includes an information identification module, a level classification module, an information evaluation module and a terminal management module which are connected in sequence; Information identification module, used to identify intellectual property information in corporate information; A level classification module is used to analyze the intellectual property information, determine the pre-emphasis combination by combining the font formats in the intellectual property information, and then determine the characteristic title according to the text information corresponding to the pre-emphasis combination and the position of the text information, and then identify the characteristic title to determine the characteristic protection area of ​​the intellectual property information, and classify the intellectual property information into a first level and a second level based on the characteristic protection area; The information evaluation module specifically divides the intellectual property rights into key texts and regular texts based on the first level and the second level, and extracts information keywords from the key texts and regular texts respectively, and marks them as the first key words and the second key words. The first key words are first used as keywords for searching the intellectual property results to obtain similar transformation results. Based on the similar transformation results, the expected results coefficient is calculated. Then, the second key words are taken and used as keywords for searching to obtain matching information. The information similarity value of the matching information is calculated, and the secondary information is determined. Then, according to the characteristic titles in the regular text, the regular information is divided into multiple different information paragraphs. At the same time, based on the information paragraphs and the secondary information, the comprehensive similarity value is determined. Then, the similar transformation results are combined with the comprehensive similarity value to determine the comprehensive risk value of the intellectual property information. The terminal management module is used to calculate the comprehensive risk value of the intellectual property information in the enterprise information and display the comprehensive risk value of the intellectual property information on the terminal device.

2. The intelligent management application system for power system according to claim 1, characterized in that: Methods for determining characteristic titles include: Marking the identified intellectual property information as target information, and identifying the font format in the target information, wherein the font format combination includes font, font size, bold, and font italic; The combination with the most font format combinations in the target information is taken as the basic combination, and the text corresponding to the basic combination is marked as the main text content, and the remaining font format combinations are marked as pre-emphasis combinations; Get the text information corresponding to the pre-emphasis combination, identify the position of the text information in the target information, and then identify the content before and after the text information. If the text information is preceded and followed by characteristic symbols, mark the text information as a characteristic title. Otherwise, if there is a place before and after the information text that is not a characteristic symbol, mark the text information as the main content, where characteristic symbols include periods, colons, and spaces.

3. The intelligent management application system for power system according to claim 2, characterized in that: The method for determining the first and second levels includes: Identify the characteristic title and determine the content title and paragraph title, wherein the content title is a title that specifically summarizes the overall content of the article in the target information, and the paragraph title is a title for the content to be expressed in a paragraph; According to the paragraph title, determine the characteristic protection area of ​​the target information, which is a unique protection area in different intellectual property rights; The feature protection area is then marked as the first level, and the remaining body content is marked as the second level.

4. The intelligent management application system for power system according to claim 1, characterized in that: The calculation method of the expected achievement coefficient includes: SS1: Extract the first-level information content and the second-level information content in the target information respectively, and mark them as key text and regular text; SS2: Semantically split the information in the key text and regular text using natural language processing technology, and extract information keywords in the key text and regular text using the TF-IDF algorithm. Information keywords are phrases with TF-IDF values ​​greater than or equal to the threshold X1. At the same time, the information keywords extracted from the key text and the regular text are marked as the first key words and the second key words respectively; SS3: Take the first key word as the keyword for searching intellectual property achievements, search in the big data network, and mark the search results as similar transformation achievements; Using the formula Calculate the expected achievement coefficient CG of the target information; Among them, Ys is the actual income of similar transformation results within one cycle, Br is the input cost of similar transformation results, Tz is the investment time of similar transformation results, α1 and α2 are both proportional coefficients, and C is a fixed constant.

5. The intelligent management application system for power system according to claim 4, characterized in that: If there are multiple similar transformation results in the detection results, the expected result coefficient of each similar transformation result is calculated respectively, and the average of all expected result coefficients is taken as the expected result coefficient of the final target information. The lower the expected result coefficient, the lower the result transformation value of the corresponding intellectual property information. Conversely, the higher the expected result coefficient, the higher the result transformation value of the corresponding intellectual property information.

6. The intelligent management application system for power system according to claim 4, characterized in that: The calculation method of the comprehensive similarity value includes: Take the second key word, search the second key word in the big data, mark the search result as matching information, use the pre-trained Word2Vec algorithm to convert the matching information and the regular text into vectors, and use the cosine similarity algorithm to calculate the information similarity value between each matching information and the regular text; Take the matching information whose information similarity value is greater than or equal to the similarity threshold X2 and mark it as secondary information; Acquire characteristic titles in regular text, and divide the regular text into multiple information paragraphs based on the characteristic titles as dividing points; Then, taking the information paragraph in the regular text as the unit, the similarity value between each information paragraph and the secondary information is calculated and marked as the paragraph similarity value, where i represents different information paragraphs, and i∈[1, I], indicating that I and information paragraphs coexist in the regular text; Take the maximum value of the paragraph similarity value of each information paragraph and mark it as the similarity extreme value Zi; Using the formula Calculate the comprehensive similarity value CL of the secondary information, where βi represents the weight coefficient of the feature title corresponding to information paragraph i.

7. The intelligent management application system for power system according to claim 6, characterized in that: The calculation method of comprehensive risk value is: Combine the expected achievement coefficient of the target information with the comprehensive similarity value, and use the formula Fx = CL × θ1 + CG -1 ×θ2 to obtain the comprehensive risk value Fx of the target information, where θ1 and θ2 are both proportional coefficients.

Images