Project decision intelligent evaluation method and system based on rule fusion

Abstract

The invention relates to a project decision intelligent evaluation method and system based on rule fusion, and belongs to the technical field of data analysis. The method comprises the steps that research and development premise attributes and result attributes are acquired, research and development premise attribute reference values and result attribute evaluation grades are obtained through Gaussian membership function calculation, and a judgment rule base is constructed; a project name is obtained, a database is retrieved according to the project name to obtain a project related file, a fact object set is extracted according to the project related file, a cleaning object set is obtained through calculation of a cleaning model according to the fact object set, and a research and development score is obtained through calculation of an inference engine according to the fact object set and a judgment rule base; a research and development attribute threshold value is preset, judgment is carried out according to the research and development attribute threshold value and the research and development score to obtain a judgment result, the intelligent terminal carries out research and development project statistics according to the judgment result to obtain a statistical result, and the statistical result is transmitted to the human-computer interaction interface. The research and development project can be intelligently found from historical projects.

Description

Technical Field

[0001] The present invention belongs to the technical field of data analysis, and particularly relates to an intelligent evaluation method and system for project decision-making based on rule fusion. Background Art

[0002] The intensity of R & D funding investment is an important indicator reflecting a company's innovation ability and has been incorporated into the performance evaluation index system of central enterprises by the State-owned Assets Supervision and Administration Commission of the State Council. The analysis of R & D funding investment is a core management task for evaluating the capital allocation and benefits in the company's technology R & D field. With the wide application of smart grid and new energy technologies and the strong support of policies for scientific and technological innovation, power grid enterprises are facing the pressure and challenges of increasing R & D investment, optimizing resource allocation, and enhancing innovation ability.

[0003] The intensity of R & D funding investment is adjusted according to the evaluation results of project decision-making. The benign adjustment of the R & D investment intensity can rationally allocate R & D resources, improve the R & D efficiency and the achievement conversion rate, and can also discover and solve problems such as resource waste and duplicate investment, thereby improving the quality and benefits of the company's R & D.

[0004] In the current project management field, with the continuous expansion of enterprise business and the increasing complexity of projects, how to efficiently and accurately identify and evaluate R & D projects has become a major challenge for enterprises. Most traditional project decision-making evaluation methods rely on manual review and experience judgment. This method not only takes time and effort but is also easily affected by personal subjective factors, making it difficult to ensure the accuracy and consistency of evaluation results.

[0005] Especially in large enterprises and scientific research institutions, there are numerous projects involving a wide range of fields, and each project has its unique background, objectives, and constraints. Traditional evaluation methods often have difficulty in comprehensively and deeply analyzing project information, thus being unable to accurately identify projects with R & D potential. In addition, with the continuous development of big data and artificial intelligence technologies, how to extract valuable information from massive data to further support project decision-making has also become an urgent problem to be solved. Summary of the Invention

[0006] To solve the above problems existing in the prior art, the present invention provides an intelligent evaluation method and system for project decision-making based on rule fusion.

[0007] The objective of the present invention can be achieved through the following technical solutions:

[0008] An intelligent evaluation method for project decision-making based on rule fusion includes:

[0009] Obtain the R & D prerequisite attributes and result attributes, calculate the R & D prerequisite attribute reference value and the result attribute evaluation level through the Gaussian membership function according to the R & D prerequisite attributes and the result attributes, and construct a judgment rule base according to the R & D prerequisite attribute reference value, the result attribute evaluation level, the R & D prerequisite attributes, and the result attributes;

[0010] Obtain the project name, retrieve the project-related files from the database according to the project name, calculate the set of fact objects through the extraction model according to the project-related files, calculate the set of cleaned objects through the cleaning model according to the set of fact objects, and calculate the R & D score through the inference engine according to the set of fact objects and the judgment rule base;

[0011] Preset the R & D attribute threshold, make a judgment according to the R & D attribute threshold and the R & D score to obtain a judgment result, the intelligent terminal performs R & D project statistics according to the judgment result to obtain a statistical result, and transmits the statistical result to the human-computer interaction interface.

[0012] As a preferred technical solution of the present invention, the project-related files include but are not limited to project profiles, task books, and research reports.

[0013] Specifically, the construction of the judgment rule base according to the R & D prerequisite attribute reference value, the result attribute evaluation level, the R & D prerequisite attributes, and the result attributes includes:

[0014] Calculate the component confidence of each component of the R & D prerequisite attributes and the result confidence of the result attributes;

[0015] Concatenate the R & D prerequisite attributes, the component confidence, the result attributes, and the result confidence to obtain a rule;

[0016] Collect each of the rules to obtain the judgment rule base.

[0017] Specifically, the specific calculation steps of the extraction model are:

[0018] S201: Obtain the word segmentation data through jieba word segmentation according to the project-related files;

[0019] S202: Calculate the left word segmentation vector and the right word segmentation vector through the embedding layer according to the word segmentation data, and concatenate the left word segmentation vector and the right word segmentation vector to obtain an input feature vector;

[0020] S203: Calculate the semantic features through the Bi-LSTM layer according to the input feature vector;

[0021] S204: Calculate entities and entity categories through the entity recognition layer according to the semantic features, classify the entities according to the entity categories to obtain character entities and numerical entities, quantify the character entities to obtain character entity quantification values, and calculate numerical entity label vectors according to the numerical entities and the entity categories through the label embedding layer;

[0022] S205: Calculate entity relationships through the relationship extraction layer according to the entity label vectors and the parameters of the Bi-LSTM layer;

[0023] S206: Construct numerical entity relationship binary tuples according to the entity relationships and the numerical entities, construct character entity relationship binary tuples according to the character entity quantification values and the character entities, preset a storage template, and store the numerical entity relationship binary tuples and the character entity relationship binary tuples according to the storage template to obtain a set of fact objects.

[0024] Specifically, the set of fact objects is stored in the form of a two-dimensional array, the array includes character-type data and numerical-type data, and the calculation steps of the cleaning model are as follows:

[0025] S301: Preset a first pointer and a second pointer, point the first pointer to the first row of the array, and point the second pointer to the second row of the array;

[0026] S302: Preset a similarity threshold, calculate the similarity of the character-type data of the first pointer and the second pointer, and judge whether the similarity is greater than the similarity threshold. If so, execute step S305; otherwise, move the second pointer one position backward and execute step S303;

[0027] S303: Judge whether the second pointer is empty. If so, move the first pointer one position backward and execute step S304; otherwise, execute step S302;

[0028] S304: Judge whether the first pointer is empty. If so, output the two-dimensional array to obtain the set of cleaning objects; otherwise, move the second pointer to the position one after the first pointer and execute step S302;

[0029] S305: Judge whether the numerical-type data of the first pointer and the second pointer are equal. If so, delete the element pointed to by the second pointer; otherwise, do nothing, move the second pointer one position backward, and execute step S302.

[0030] Specifically, the calculation process of the inference engine is as follows:

[0031] Preset a rule activation weight threshold, calculate the rule activation weight between the cleaning object set and each rule in the rule base, determine a calculation rule according to the rule activation weight and the rule activation weight threshold, perform rule synthesis through the ER algorithm according to the calculation rule to obtain the confidence level of the result attribute evaluation level, and calculate the R & D score according to the confidence level of the result attribute evaluation level.

[0032] Specifically, the judgment result includes R & D projects and non-R & D projects. The judgment result obtained by judging according to the R & D attribute threshold and the R & D score includes:

[0033] Judge whether the R & D score is greater than the R & D attribute threshold. If so, the judgment result is an R & D project; if not, the judgment result is a non-R & D project.

[0034] An intelligent project decision-making evaluation system based on rule fusion includes: a rule base construction module, a data processing module, and an inference module;

[0035] The rule base construction module is used to obtain R & D premise attributes and result attributes, calculate the R & D premise attribute reference values and result attribute evaluation levels through the Gaussian membership function according to the R & D premise attributes and result attributes, and construct a judgment rule base according to the R & D premise attribute reference values, the result attribute evaluation levels, the R & D premise attributes, and the result attributes;

[0036] The data processing module is used to obtain the project name, retrieve project-related files from the database according to the project name, calculate the fact object set through an extraction model according to the project-related files, calculate the cleaning object set through a cleaning model according to the fact object set, and calculate the R & D score through an inference engine according to the fact object set and the judgment rule base;

[0037] The inference module is used to preset an R & D attribute threshold, make a judgment according to the R & D attribute threshold and the R & D score to obtain a judgment result. The intelligent terminal performs R & D project statistics according to the judgment result to obtain a statistical result, and transmits the statistical result to the human-computer interaction interface.

[0038] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the intelligent project decision-making evaluation method based on rule fusion as described above.

[0039] A storage medium containing computer-executable instructions, where the computer-executable instructions are used to execute the intelligent project decision-making evaluation method based on rule fusion as described above when executed by a computer processor.

[0040] The beneficial effects of the present invention are:

[0041] (1) By constructing a judgment rule base, establishing an extraction model, a cleaning model, and an inference engine, an automated processing flow is realized, significantly reducing the workload of manual review, shortening the project evaluation cycle, improving the evaluation efficiency, and quickly finding innovative projects from a large number of historical projects.

[0042] (2) The inference mechanism based on the rule base can comprehensively and deeply analyze project information, avoiding the influence of personal subjective factors and improving the accuracy and consistency of the evaluation results.

[0043] (3) By accurately identifying R & D projects, enterprises can allocate resources more reasonably, ensure that R & D projects receive sufficient support and attention, and thus improve the R & D success rate.

[0044] (4) The application of this method and system promotes the development of project management towards intelligence, providing more scientific and efficient decision-making support for enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings.

[0046] Figure 1 It is a schematic flowchart of an intelligent evaluation method for project decision-making based on rule fusion of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following will, with reference to the accompanying drawings and preferred embodiments, describe in detail the specific embodiments, structures, features, and their effects according to the present invention.

[0048] Please refer to Figure 1 , an intelligent evaluation method for project decision-making based on rule fusion,

[0049] Obtain the R & D premise attributes and result attributes, calculate the R & D premise attribute reference values and result attribute evaluation grades through a Gaussian membership function according to the R & D premise attributes and result attributes, and construct a judgment rule base according to the R & D premise attribute reference values, the result attribute evaluation grades, the R & D premise attributes, and the result attributes;

[0050] Obtain the project name, retrieve project-related files from the database according to the project name, calculate the fact object set through the extraction model according to the project-related files, calculate the cleaned object set through the cleaning model according to the fact object set, and calculate the R & D score through the inference engine according to the fact object set and the judgment rule base;

[0051] Set a threshold for R & D attributes, make a judgment based on the R & D attribute threshold and the R & D score to obtain a judgment result, and the intelligent terminal performs R & D project statistics based on the judgment result to obtain a statistical result, and transmits the statistical result to the human-computer interaction interface.

[0052] Specifically, the project-related documents include, but are not limited to, project profiles, task books, and research reports.

[0053] Specifically, constructing the judgment rule base according to the R & D premise attribute reference value, the result attribute evaluation level, the R & D premise attribute, and the result attribute includes:

[0054] Calculate the component confidence of each component of the R & D premise attribute and the result confidence of the result attribute;

[0055] The formula for calculating the component confidence is:

[0056]

[0057] Among them, Y ij k Is the confidence of the jth component of the ith premise attribute of the Kth data, A ij k Is the reference value of the jth component of the ith premise attribute of the Kth data, A i(j+1) k Is the reference value of the j + 1th component of the ith premise attribute of the Kth data, X i k Is the ith premise attribute of the kth data;

[0058] It should be noted that the calculation process of the result confidence is the same as that of the component confidence calculation process;

[0059] Concatenate the R & D premise attribute, the component confidence, the result attribute, and the result confidence to obtain a rule;

[0060] The expression of the rule is:

[0061]

[0062] Among them, x Tk Is the Tkth R & D premise attribute, R k Is the kth rule, Ai j Is the jth reference value of the ith R & D premise attribute, a ij k Is the confidence of the jth reference value of the ith R & D premise attribute of the kth rule, T k Is the total number of R & D premise attributes of the kth rule, J iis the number of reference values of the i-th R & D prerequisite attribute, θ k The relative weight of the k-th rule, represents the relative attribute weight of the i-th prerequisite attribute in the k-th rule; D j represents the j-th result attribute evaluation level, β j k is D j is the confidence level of D; N represents the total number of result evaluation levels;

[0063] The judgment rule base is obtained by aggregating each of the said rules.

[0064] Specifically, the specific calculation steps of the extraction model are as follows:

[0065] S201: Obtain tokenized data by Jieba tokenization according to the project-related documents;

[0066] S202: Calculate the left token vector and the right token vector through the embedding layer according to the tokenized data, and splice the left token vector and the right token vector to obtain an input feature vector;

[0067] S203: Calculate semantic features through the Bi-LSTM layer according to the input feature vector;

[0068] S204: Calculate entities and entity categories through the entity recognition layer according to the semantic features, classify the entities according to the entity categories to obtain character entities and numerical entities, quantify the character entities to obtain character entity quantization values, and calculate numerical entity label vectors through the label embedding layer according to the numerical entities and the entity categories;

[0069] S205: Calculate entity relationships through the relationship extraction layer according to the entity label vectors and the parameters of the Bi-LSTM layer;

[0070] S206: Construct a numerical entity relationship binary tuple according to the entity relationships and the numerical entities, construct a character entity relationship binary tuple according to the character entity quantization values and the character entities, preset a storage template, and store the numerical entity relationship binary tuple and the character entity relationship binary tuple according to the storage template to obtain a set of fact objects.

[0071] Specifically, the set of fact objects is stored in the form of a two-dimensional array, the array includes character data and numerical data, and the calculation steps of the cleaning model are as follows:

[0072] S301: Preset a first pointer and a second pointer, point the first pointer to the first row of the array, and point the second pointer to the second row of the array;

[0073] S302: Preset a similarity threshold, calculate the similarity of the character-type data of the first pointer and the second pointer, and determine whether the similarity is greater than the similarity threshold. If yes, execute step S305; if no, move the second pointer one position backward and execute step S303;

[0074] The similarity calculation formula is:

[0075]

[0076] where C is the similarity, T A is the character-type data of the first pointer, T B is the character-type data of the second pointer, T A ·T B represents dot product calculation, ||T A ||||T B || represents norm calculation;

[0077] S303: Determine whether the second pointer is empty. If yes, move the first pointer one position backward and execute step S304; if no, execute step S302;

[0078] S304: Determine whether the first pointer is empty. If yes, output the two-dimensional array to obtain the cleaned object set; if no, move the second pointer to the position one after the first pointer and execute step S302;

[0079] S305: Determine whether the numerical data of the first pointer and the second pointer are equal. If yes, delete the element pointed to by the second pointer; if no, do nothing, move the second pointer one position backward and execute step S302.

[0080] Specifically, the calculation process of the inference engine is as follows:

[0081] Preset a rule activation weight threshold, calculate the rule activation weight between the cleaned object set and each rule in the rule library, determine the calculation rule according to the rule activation weight and the rule activation weight threshold, perform rule synthesis through the ER algorithm according to the calculation rule to obtain the confidence level of the result attribute evaluation level, and calculate the R & D score according to the confidence level of the result attribute evaluation level,

[0082] The activation weight calculation formula is

[0083]

[0084] where W k is the activation weight, b kis the relative rule weight of the k-th calculation rule, i is a variable parameter, k is the total number of calculation rules, a i k is the matching degree between the cleaning object set and the i-th R & D prerequisite attribute in the k-th calculation rule, l is a variable parameter, L is the total number of rules, b l is the relative rule weight of the l-th rule, i is a variable parameter, a i l is the matching degree between the cleaning object set and the i-th R & D prerequisite attribute in the l-th rule, c i is the average weight of R & D prerequisite attributes;

[0085] The R & D score calculation formula is as follows:

[0086]

[0087] where f is the R & D score, j is a variable parameter, N is the total number of calculation rules, μ(D j ) is the utility value of the j-th result attribute evaluation level, β j is the confidence level of the j-th result attribute evaluation level, D j is the j-th result attribute evaluation level.

[0088] Specifically, the judgment result includes R & D projects and non-R & D projects. The judgment result obtained by judging according to the R & D attribute threshold and the R & D score includes:

[0089] Judge whether the R & D score is greater than the R & D attribute threshold. If so, the judgment result is an R & D project; if not, the judgment result is a non-R & D project.

[0090] An intelligent evaluation system for project decision-making based on rule fusion includes: a rule base construction module, a data processing module, and an inference module;

[0091] The rule base construction module is used to obtain R & D prerequisite attributes and result attributes, calculate R & D prerequisite attribute reference values and result attribute evaluation levels through a Gaussian membership function according to the R & D prerequisite attributes and result attributes, and construct a judgment rule base according to the R & D prerequisite attribute reference values, the result attribute evaluation levels, the R & D prerequisite attributes, and the result attributes;

[0092] The data processing module is used to obtain the project name, retrieve project-related files from the database according to the project name, calculate a fact object set through an extraction model according to the project-related files, calculate a cleaning object set through a cleaning model according to the fact object set, and calculate an R & D score through an inference engine according to the fact object set and the judgment rule base;

[0093] The inference module is used to preset a R & D attribute threshold, make a judgment based on the R & D attribute threshold and the R & D score to obtain a judgment result. The intelligent terminal performs R & D project statistics according to the judgment result to obtain a statistical result, and transmits the statistical result to the human-computer interaction interface.

[0094] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the intelligent evaluation method for project decision-making based on rule fusion as described above.

[0095] A storage medium containing computer-executable instructions, where the computer-executable instructions are used to execute the intelligent evaluation method for project decision-making based on rule fusion as described above when executed by a computer processor.

[0096] The computer storage medium of the embodiments of the present invention can be any combination of one or more computer-readable media. The computer-readable media can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer-readable storage medium can be any tangible medium that contains or stores a program, which can be used by or in combination with an instruction execution system, apparatus, or device.

[0097] The computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries the computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, which can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0098] The program code contained on a computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber cable, RF, etc., or any suitable combination of the foregoing. The computer program code for performing the operations of the present invention can be written in one or more programming languages or combinations thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and also including conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or, alternatively, can be connected to an external computer (e.g., using an Internet service provider to connect via the Internet).

[0099] As described above, the above are only the preferred embodiments of the present invention and do not impose any formal restrictions on the present invention. Although the present invention has been disclosed as above in preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art, without departing from the scope of the technical solution of the present invention, can make some changes or modifications to the above-disclosed technical content to obtain equivalent embodiments with equivalent changes, but as long as the content of the technical solution of the present invention is not departed from, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A project decision intelligent evaluation method based on rule fusion, characterized in that: include: Acquire research and development premise attributes and result attributes, calculate the research and development premise attribute reference value and the result attribute evaluation grade through Gaussian membership function according to the research and development premise attributes and the result attributes, and build a judgment rule library according to the research and development premise attribute reference value, the result attribute evaluation grade, the research and development premise attributes, and the result attributes; Obtain a project name, search a database according to the project name to obtain project-related documents, calculate a fact object set according to the project-related documents through an extraction model, calculate a cleansing object set according to the fact object set through a cleaning model, and calculate a research and development score according to the fact object set and the judgment rule base through an inference engine; A research and development attribute threshold is preset, and a judgment result is obtained based on the research and development attribute threshold and the research and development score. The intelligent terminal performs research and development project statistics based on the judgment result to obtain statistical results, and transmits the statistical results to the human-computer interaction interface.

2. The project decision intelligent evaluation method based on rule fusion according to claim 1 is characterized in that: The project-related documents include but are not limited to project brief, task book, and research report.

3. The project decision intelligent evaluation method based on rule fusion according to claim 1 is characterized in that: The constructing of a judgment rule base according to the R&D premise attribute reference value, the result attribute evaluation level, the R&D premise attribute, and the result attribute comprises: Calculating the component confidence of each component of the R&D premise attribute and the result confidence of the result attribute; The research and development premise attribute, the component confidence, the result attribute, and the result confidence are spliced ​​to obtain a rule; Each of the rules is aggregated to obtain the judgment rule library.

4. The project decision intelligent evaluation method based on rule fusion according to claim 1 is characterized in that: The specific calculation steps of the extraction model are: S201: Obtaining word segmentation data through Jieba word segmentation according to the project-related files; S202: Calculate a left word segmentation vector and a right word segmentation vector through an embedding layer according to the word segmentation data, and concatenate the left word segmentation vector and the right word segmentation vector to obtain an input feature vector; S203: Obtaining semantic features through Bi-LSTM layer calculation according to the input feature vector; S204: Calculate entities and entity categories through an entity recognition layer according to the semantic features, classify the entities according to the entity categories to obtain character entities and numerical entities, quantize the character entities to obtain character entity quantization values, and calculate numerical entity label vectors through a label embedding layer according to the numerical entities and the entity categories; S205: Obtaining entity relations through calculation by a relation extraction layer according to the entity label vector and the parameters of the Bi-LSTM layer; S206: construct a numerical entity relationship bigram according to the entity relationship and the numerical entity, construct a character entity relationship bigram according to the character entity quantization value and the character entity, preset a storage template, and store the numerical entity relationship bigram and the character entity relationship bigram according to the storage template to obtain a fact object set.

5. The project decision intelligent evaluation method based on rule fusion according to claim 1 is characterized in that: The fact object set is stored in the form of a two-dimensional array, and the array includes character data and numerical data. The cleaning model calculation steps are: S301: Preset a first pointer and a second pointer, set the first pointer to point to the first row of the array, and set the second pointer to point to the second row of the array; S302: Preset a similarity threshold, calculate the similarity of the character data of the first pointer and the second pointer, and determine whether the similarity is greater than the similarity threshold. If yes, execute step S305; if no, move the second pointer backward by one position and execute step S303; S303: Determine whether the second pointer is empty, if yes, move the first pointer backward by one position, and execute step S304; If no, then execute step S302; S304: Determine whether the first pointer is empty. If so, output the two-dimensional array to obtain the cleaning object set. If not, move the second pointer to the position after the first pointer and execute step S302. S305: Determine whether the numerical data of the first pointer and the second pointer are equal, if yes, delete the element pointed to by the second pointer; If not, no processing is performed, and the second pointer is moved backward by one position, and step S302 is executed.

6. The project decision intelligent evaluation method based on rule fusion according to claim 1 is characterized in that: The calculation process of the inference engine is: A rule activation weight threshold is preset, and the rule activation weight between the cleaning object set and each rule in the rule base is calculated. The calculation rule is determined according to the rule activation weight and the rule activation weight threshold. According to the calculation rule, rule synthesis is performed through the ER algorithm to obtain the result attribute evaluation level confidence, and the R&D score is calculated according to the result attribute evaluation level confidence.

7. The project decision intelligent evaluation method based on rule fusion according to claim 1 is characterized in that: The judgment result includes R&D projects and non-R&D projects, and the judgment result obtained by judging according to the R&D attribute threshold and the R&D score includes: It is determined whether the R&D score is greater than the R&D attribute threshold. If yes, the determination result is an R&D project; if not, the determination result is not a R&D project.

8. A project decision intelligent evaluation system based on rule fusion, characterized in that: include: Rule base construction module, data processing module, and reasoning module; The rule base construction module is used to obtain the R&D premise attribute and the result attribute, calculate the R&D premise attribute reference value and the result attribute evaluation grade through Gaussian membership function according to the R&D premise attribute and the result attribute, and construct a judgment rule base according to the R&D premise attribute reference value, the result attribute evaluation grade, the R&D premise attribute, and the result attribute; The data processing module is used to obtain a project name, retrieve a database according to the project name to obtain project-related documents, calculate a fact object set according to the project-related documents through an extraction model, calculate a cleansing object set according to the fact object set through a cleaning model, and calculate an R&D score according to the fact object set and the judgment rule base through an inference engine; The reasoning module is used to preset a research and development attribute threshold, and to obtain a judgment result based on the research and development attribute threshold and the research and development score. The intelligent terminal performs research and development project statistics based on the judgment result to obtain a statistical result, and transmits the statistical result to the human-computer interaction interface.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the project decision intelligent evaluation method based on rule fusion as described in any one of claims 1-7 is implemented.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the project decision intelligent evaluation method based on rule fusion as described in any one of claims 1-7 when executed by a computer processor.

Images