A digital project application intelligent review system and method

Abstract

The present invention discloses a digital project application intelligent review system and method, which belongs to the field of project application intelligent review technology. It includes: taking the project with current content changes as the target project, obtaining the changed content of the target project, and extracting keywords in the changed content; collecting the approval records corresponding to each approval link, and obtaining the correlation coefficient corresponding to the changed content and each approval link; obtaining and analyzing historical projects, and obtaining the associated approval links corresponding to each approval link; and determining the re-approval process of the target project according to the correlation coefficient. The present invention analyzes the keywords in the changed content and the approval records of the target project, and determines the re-approval process of the target project according to the historical projects; reasonably sets the re-approval process of the project after the content changes, and does not re-examine the approval links that do not involve the changed content, which has the advantages of improving the efficiency and quality of the review.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent review of project applications, and in particular to a digital intelligent review system and method for project applications. Background Art

[0002] Project approval plays an important role in ensuring the smooth implementation of projects, promoting economic development and protecting public interests. During the project approval process, various contents of the project will be reviewed to ensure that the project complies with the requirements of relevant laws, regulations, policies and standards. As the scale and complexity of the project increase, the difficulty and workload of approval will also increase accordingly. During the project review, it is necessary to conduct a comprehensive assessment and review of all aspects of the project. If the project approval process is regarded as an approval process, the approval process includes several approval links, and each approval link will be recorded after approval. If the content needs to be changed in the project after the project approval process is completed, the staff usually needs to re-examine all approval links, which not only wastes time and delays the progress of the project, but also consumes a lot of human resources. Therefore, it is necessary to conduct intelligent review of project applications. According to the content of the project changes, a reasonable re-approval process for the changed project should be set up, and the approval links that do not involve the changed content should not be re-examined, which will help improve the efficiency and quality of the review. Summary of the invention

[0003] The purpose of the present invention is to provide a digital project application intelligent review system and method to solve the problems raised in the above background technology.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0005] A digital project application intelligent review method includes the following steps:

[0006] Step S100: taking the project with the current content change as the target project, obtaining the change content of the target project, analyzing the change content, and extracting keywords from the change content;

[0007] Step S200: The approval process of the target project includes several approval links, each of which has a corresponding approval record. The approval record corresponding to each approval link is collected; based on the approval record and keywords, the correlation coefficient between the change content and each approval link is obtained;

[0008] The general project approval process includes multiple review links, among which the review links mainly include: material review, feasibility assessment, risk assessment, budget review, compliance review, etc. Each review link has its specific purpose, and after the review, the relevant reviewer will output the corresponding approval record based on the review results. The approval record is a record of the results of this approval, including the approval time, approval content, etc.;

[0009] Step S300: obtaining a number of historical items, and obtaining the associated approval links corresponding to each approval link according to the number of approval links corresponding to each historical item and the approval records corresponding to each approval link;

[0010] Step S400: Determine the re-approval process of the target project according to the correlation coefficient and the associated approval links corresponding to each approval link.

[0011] Furthermore, step S100 includes:

[0012] Step S110: obtaining the changed content CH, and performing preprocessing to extract the text changed content and the digital changed content in the changed content CH;

[0013] Step S120: taking the content between two adjacent punctuation marks in the text as a sentence fragment; if the digitally changed content is not empty, obtaining the sentence fragment of the digitally changed content in the text corresponding to the target item, and extracting characteristic keywords in the sentence fragment;

[0014] The textual change content in the change content is to record the textual information of the change content; the digital change content is usually a description of the products and equipment involved in the number. For example, if a unit wants to request funding to purchase equipment, digital information will appear. Through the digital change content, the corresponding textual information, that is, the characteristic keywords in the sentence fragment, can be obtained; and according to the textual change content, pre-processing is performed to obtain the keywords in the change content CH;

[0015] The product name extracted from the sentence fragment in step S120 and the text keywords extracted from the text change content in step S130 can be obtained by a text summarization algorithm or a text classification algorithm in the prior art;

[0016] Step S130: If the text change content is not empty, extract the text keywords in the text change content, and combine them with the characteristic keywords, extract all the keywords in the change content CH through word frequency statistics, and then obtain a keyword set, which includes each keyword and the number corresponding to each keyword.

[0017] Further, step S200 includes:

[0018] Step S210: The approval record includes the approval time and approval content; a certain approval record is set as AR, and the approval content corresponding to the approval record AR is AC. The approval content AC is preprocessed and relations are extracted to obtain all entities and definition relations in the approval content AC, and the knowledge graph corresponding to the approval content AC is obtained based on the entities and definition relations;

[0019] The knowledge graph is a semantic network that uses a graph structure to represent the relationships and attributes between entities. Entities refer to specific things or concepts, and definition relationships represent the way these entities are connected, describing their interactions or attributes. Through entities and definition relationships, we can better understand the associations and attributes between entities. In this embodiment, through the knowledge graph, we can obtain the associations between the entities in the approval content AC. Through this association information, we can obtain the degree of relationship between the keywords in the change content CH and the approval content AC.

[0020] Step S220: According to the keyword set and the knowledge graph, the semantic similarity between the kth keyword in the keyword set and the mth entity or nth definition relationship in the knowledge graph is obtained, and the result of the semantic similarity is normalized. If the semantic similarity is greater than the similarity threshold, the mth entity or nth definition relationship is marked at the corresponding position in the knowledge graph, and the weight of the marked position is the number corresponding to the kth keyword; then all the marked positions of the knowledge graph regarding the changed content CH and the weight of each marked position are obtained;

[0021] Step S230: Set the approval link corresponding to the approval record AR to AP, and the initial correlation coefficient X between the change content CH and the approval link AP = 0. If the mth entity and the nth definition relationship in the knowledge graph are related to each other and both are marked, obtain the weight W corresponding to the mth entity. m The weight W corresponding to the nth defined relationship n , and increase the value of the correlation coefficient X by W m +W0*W n , W0 is the relationship weight coefficient; and then the final correlation coefficient X between the change content CH and the approval link AP is obtained.

[0022] Furthermore, step S300 includes:

[0023] Step S310: Taking the application time of the target project as the starting time, obtain the H historical projects that have completed approval starting from the starting time; obtain the pth and qth approval links corresponding to the hth historical project, as well as the corresponding approval records and Approval Record and The corresponding knowledge graph is and The approval time is taken as and the approval time is the time when the corresponding approval link starts;

[0024] Step S320: According to all the entities and defined relationships corresponding in the knowledge graph and the combinations of entities and defined relationships that are mutually related are taken as relationship combinations; thus, an entity set and a defined relationship set and and a relationship combination set and The number of elements that are the same in the two entity sets is denoted as The number of elements that are the same in the two defined relationship sets is The number of elements that are the same in the two relationship combination sets is Thus, the approval record and of the h-th historical project and the corresponding data correlation coefficient are obtained;

[0025] Among them, in the knowledge graph, the combination of entities and defined relationships that are mutually related is the combination of the connections established between different entities in the knowledge graph through specific defined relationships; in this solution, the value of the data correlation coefficient is: Among them, 1>c>b>a>0, a + b + c = 1;

[0026] Step S330: Obtain the data correlation coefficients corresponding to the 1st, 2nd,..., H-th historical projects. If there are H0 data correlation coefficients greater than the correlation coefficient threshold and H0 < H, it is determined that there is a correlation between the p-th and q-th approval links; then, it is judged whether there is a correlation between any two approval links;

[0027] Step S340: According to the p-th and q-th approval links with correlations, obtain the p-th and q-th approval times corresponding to the 1st, 2nd,..., H-th historical projects. Denote the number of times when the p-th approval time is earlier than the q-th approval time as N1, and the number of times when the p-th approval time is not earlier than the q-th approval time as N2; if N1 > N2, then the q-th approval link is the associated approval link of the p-th approval link, and if N1 ≤ N2, then the p-th approval link is the associated approval link of the q-th approval link; thus, the associated approval link corresponding to each approval link is obtained.

[0028] If the p-th approval link is usually earlier than the q-th approval link, it can be said that the q-th approval link is an associated approval link of the p-th approval link, because the p-th approval link is carried out first, and its result may affect the approval of the q-th approval link. The approval of the q-th approval link depends to a certain extent on the result or condition of the p-th approval link.

[0029] Furthermore, step S400 includes:

[0030] Step S410: Determine the approval weight V corresponding to the approval link AP according to the correlation coefficient between the change content CH and the approval link AP. AP ;

[0031] The correlation coefficient indicates that there is a correlation between the change content and the approval link. Since there is also a correlation between the approval links, this solution uses the approval weight to determine whether each approval link should be re-approved based on the relationship between the approval link and its corresponding associated approval link;

[0032] Step S420: According to the approval weight V AP , and the associated approval link corresponding to the approval link AP, the approval weight corresponding to the associated approval link of the approval link AP is V RAP =V AP / B, B is the correlation weight coefficient; obtain the approval weight V obtained by the approval link AP according to the correlation coefficient AP , and the approval weight when it is an associated approval link, and add them together to get the final approval weight of the approval link AP; then get the final approval weight corresponding to each review link, and get the re-approval process of the target project, as well as the importance of each approval link in the re-approval process.

[0033] The higher the final approval weight corresponding to each review link, the more important it is; and based on the final approval weight, the review links are divided into no review, simple review, general review and important review; ultimately, the approval links without review are removed, and the resulting review and approval process only has the approval links of simple review, general review and important review.

[0034] A digital project application intelligent review system, including a keyword module, a correlation coefficient module, a correlation approval link module and a re-approval process module;

[0035] Keyword module: used to take the current project with changed content as the target project, obtain the changed content of the target project, analyze the changed content, and extract keywords from the changed content;

[0036] Correlation coefficient module: It is used to collect the approval records corresponding to each approval link according to the approval process of the target project, which includes several approval links. According to the approval records and keywords, the correlation coefficient corresponding to the change content and each approval link is obtained;

[0037] Related approval link module: used to obtain several historical projects, and obtain the related approval links corresponding to each approval link according to the several approval links corresponding to each historical project and the approval records corresponding to each approval link;

[0038] Re-approval process module: used to determine the re-approval process of the target project based on the correlation coefficient and the associated approval links corresponding to each approval link.

[0039] Furthermore, the association coefficient module includes a knowledge graph unit, a tag weight unit and an association coefficient unit;

[0040] Knowledge graph unit: used to obtain all entities and definition relationships in the approval content according to the approval content in the approval record, and obtain the knowledge graph corresponding to the approval content according to the entities and definition relationships;

[0041] Marking weight unit: used to obtain all markings in the knowledge graph and the weight of each marking;

[0042] Correlation coefficient unit: used to obtain the correlation coefficient between the change content and each approval link based on the approval records and keywords.

[0043] Further, the associated approval link module includes a historical project unit, a data association coefficient unit, a correlation unit, and an associated approval link unit;

[0044] Historical project unit: used to obtain any two approval links and corresponding approval records corresponding to historical projects, and obtain the knowledge graph and approval time corresponding to any two approval records;

[0045] Data association coefficient unit: used to obtain the data association coefficient corresponding to any two approval records according to the entities, defined relationships and relationship combinations corresponding to the knowledge graphs corresponding to any two approval records;

[0046] Relevance unit: used to obtain the data correlation coefficients of any two approval records corresponding to each historical item, and to determine whether there is a correlation between any two approval links;

[0047] Related approval link unit: used to obtain the related approval link corresponding to each approval link according to the correlation between any two approval links.

[0048] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: the present invention provides a digital project application intelligent review system and method, including: taking the project with the current content change as the target project, obtaining the change content of the target project, and extracting the keywords in the change content; collecting the approval records corresponding to each approval link, and obtaining the correlation coefficient corresponding to the change content and each approval link; obtaining and analyzing the historical projects, and obtaining the associated approval links corresponding to each approval link; and determining the re-approval process of the target project according to the correlation coefficient. The present invention analyzes the keywords in the change content and the approval records of the target project, and determines the re-approval process of the target project according to the historical projects; reasonably sets the re-approval process of the project after the content change, and does not re-examine the approval links that do not involve the change content, which has the advantages of improving the efficiency and quality of the review. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0050] Figure 1 It is a flowchart of a digital project application intelligent review method of the present invention;

[0051] Figure 2 It is a structural diagram of a digital project application intelligent review system of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] See also Figure 1-Figure 2 The present invention provides a digital project application intelligent review system and method, and a corresponding flowchart of a digital project application intelligent review method is as follows: Figure 1 As shown. The following steps are included:

[0054] Step S100: taking the project with the current content change as the target project, obtaining the change content of the target project, analyzing the change content, and extracting keywords from the change content;

[0055] Step S110: Acquire the changed content CH, perform preprocessing, and extract the text changed content and the digital changed content in the changed content CH.

[0056] Step S120: the content between two adjacent punctuation marks in the text is regarded as a sentence fragment; if the digitally changed content is not empty, the sentence fragment of the digitally changed content in the text corresponding to the target item is obtained, and characteristic keywords in the sentence fragment are extracted.

[0057] The textual changes in the change content record the textual information of the change content; the digital changes are usually descriptions of the products and equipment involved in the number. For example, if a unit requests funding to purchase equipment, digital information will appear. Through the digital changes, the corresponding textual information can be obtained, that is, the characteristic keywords in the sentence fragment; and pre-processing is performed based on the textual changes to obtain the keywords in the change content CH.

[0058] The product name extracted from the sentence fragment in step S120 and the text keywords extracted from the text change content in step S130 can be obtained by using a text summarization algorithm or a text classification algorithm in the prior art, and the specific process will not be repeated here.

[0059] Step S130: If the text change content is not empty, extract the text keywords in the text change content, and combine them with the characteristic keywords, extract all the keywords in the change content CH through word frequency statistics, and then obtain a keyword set, which includes each keyword and the number corresponding to each keyword.

[0060] Through word frequency statistics, if the number of keywords in the text keywords or key product names is greater than 1 or 2 (determined according to actual conditions), they can be used as keywords for the changed content CH.

[0061] Step S200: The approval process of the target project includes several approval links, each of which has a corresponding approval record. The approval record corresponding to each approval link is collected; based on the approval record and keywords, the correlation coefficient between the change content and each approval link is obtained;

[0062] The general project approval process includes multiple review links, among which the review links mainly include: material review, feasibility assessment, risk assessment, budget review, compliance review, etc. Each review link has its specific purpose, and after the review, the relevant reviewers will output the corresponding approval records based on the review results. The approval record is a record of the results of this approval, including the approval time, approval content, etc.

[0063] Step S210: The approval record includes the approval time and approval content; set a certain approval record as AR, and the approval content corresponding to the approval record AR is AC, pre-process the approval content AC, and perform relationship extraction to obtain all entities and definition relationships in the approval content AC, and obtain the knowledge graph corresponding to the approval content AC based on the entities and definition relationships.

[0064] The knowledge graph is a semantic network that uses a graph structure to represent the relationships and attributes between entities. Entities refer to specific things or concepts, and definition relationships represent the way these entities are connected, describing their interactions or attributes. Through entities and definition relationships, we can better understand the associations and attributes between entities. In this embodiment, through the knowledge graph, we can obtain the associations between the entities in the approval content AC. Through this association information, we can obtain the degree of relationship between the keywords in the change content CH and the approval content AC.

[0065] Step S220: Based on the keyword set and the knowledge graph, calculate the semantic similarity between the kth keyword in the keyword set and the mth entity or nth definition relationship in the knowledge graph, and normalize the result of the semantic similarity. If the semantic similarity is greater than the similarity threshold, the mth entity or nth definition relationship is marked at the corresponding position in the knowledge graph, and the weight of the marked position is the number corresponding to the kth keyword; then all the marked positions of the knowledge graph about the changed content CH and the weight of each marked position are obtained.

[0066] Step S230: Set the approval link corresponding to the approval record AR to AP, and the initial correlation coefficient X between the change content CH and the approval link AP = 0. If the mth entity and the nth definition relationship in the knowledge graph are related to each other and both are marked, obtain the weight W corresponding to the mth entity. m The weight W corresponding to the nth defined relationship n , and increase the value of the correlation coefficient X by W m +W0*W n , W0 is the relationship weight coefficient; and then the final correlation coefficient X between the change content CH and the approval link AP is obtained.

[0067] In this embodiment, since the weight W m and W n It is related to the number corresponding to the kth keyword. Since the definition relationship can better explain the relationship between the change content CH and the approval link AP, the weight of the definition relationship in the correlation coefficient X is higher than that of the entity. In this embodiment, W0=2.

[0068] Step S300: obtaining a number of historical items, and obtaining the associated approval links corresponding to each approval link according to the number of approval links corresponding to each historical item and the approval records corresponding to each approval link;

[0069] Approval links are usually related. For example, in the various approval links of a software project, there may be an impact between material review and feasibility assessment: if the project application materials are found to be incomplete or inaccurate during the material review, it may lead to a lack of necessary information in the feasibility assessment, thereby affecting the accuracy and comprehensiveness of the feasibility assessment; there may be an impact between feasibility assessment and risk assessment: the analysis results of the project technology, economy, market, etc. in the feasibility assessment will provide a basis for risk assessment. This solution analyzes each approval link to obtain the associated approval links corresponding to each approval link, among which a certain approval link may correspond to multiple associated approval links.

[0070] Step S310: Taking the application time of the target project as the starting time, obtain the H historical projects that have completed approval starting from the starting time; obtain the pth and qth approval links corresponding to the hth historical project, as well as the corresponding approval records and Approval Record and The corresponding knowledge graph is and Approval time as and The approval time is the time when the corresponding approval link starts.

[0071] Step S320: Based on the knowledge graph and All entities and definition relationships corresponding to the above are combined into a set of related entities and definition relationships as a relationship combination; then the entity set is obtained. and Defining a Relationship Set and And the relationship combination set and The number of elements that are the same in the two entity sets is recorded as The number of elements in common in the two defined relation sets is The number of elements in the combination sets of two relations that are the same is Then get the approval record of the hth historical project and The corresponding data correlation coefficient.

[0072] Among them, in the knowledge graph, there is a combination of interrelated entities and definition relationships, which is a combination of connections established between different entities in the knowledge graph through specific definition relationships.

[0073] In this embodiment, the value of the data correlation coefficient is as follows: where 1 > c > b > a > 0 and a + b + c = 1. The correlation between two approval processes is obtained through the data correlation coefficient. In this embodiment, a = 0.2, b = 0.3, and c = 0.5.

[0074] Step S330: Obtain the data correlation coefficients corresponding to the 1st, 2nd, …, Hth historical projects. If there are H0 data correlation coefficients greater than the correlation coefficient threshold and H0 < H, it is determined that there is a correlation between the pth and qth approval processes; then, it is determined whether there is a correlation between any two approval processes. In this embodiment, the number of H0 is H / 2.

[0075] Step S340: According to the pth and qth approval processes with a correlation, obtain the pth and qth approval times corresponding to the 1st, 2nd, …, Hth historical projects. Denote the number of times when the pth approval time is earlier than the qth approval time as N1, and the number of times when the pth approval time is not earlier than the qth approval time as N2; if N1 > N2, the qth approval process is the associated approval process of the pth approval process, and if N1 ≤ N2, the pth approval process is the associated approval process of the qth approval process; then, the associated approval process corresponding to each approval process is obtained.

[0076] If the pth approval process is usually earlier than the qth approval process, it can be said that the qth approval process is the associated approval process of the pth approval process because the pth approval process is carried out first, and its result may affect the approval of the qth approval process. The approval of the qth approval process depends to a certain extent on the result or conditions of the pth approval process.

[0077] Step S400: Determine the review process of the target project according to the correlation coefficient and the associated approval process corresponding to each approval process.

[0078] Step S410: Determine the approval weight V corresponding to the approval process AP according to the correlation coefficient between the change content CH and the approval process AP AP .

[0079] The correlation coefficient indicates that there is a correlation between the change content and the approval process. Since there is also a correlation between each approval process, in this solution, according to the relationship between the approval process and its associated approval process, the approval weight is used to determine whether each approval process needs to be re-approved. In this embodiment, according to the correlation coefficient X, if 10 ≤ X < 20, the approval weight V AP = 1; if 20 ≤ X < 30, the approval weight V AP = 2; if 30 ≤ X, the approval weight V AP = 3.

[0080] Step S420: According to the approval weight V AP , and the associated approval link corresponding to the approval link AP, the approval weight corresponding to the associated approval link of the approval link AP is V RAP =V AP / B, B is the correlation weight coefficient; obtain the approval weight V obtained by the approval link AP according to the correlation coefficient AP , and the approval weight when it is an associated approval link, and add them together to get the final approval weight of the approval link AP; then get the final approval weight corresponding to each review link, and get the re-approval process of the target project, as well as the importance of each approval link in the re-approval process.

[0081] In this embodiment, the associated weight coefficient B=3, that is, if the approval weight V corresponding to the approval link AP AP =3, then the approval weight corresponding to the associated approval link of the approval link AP is 1.

[0082] According to the final approval weight corresponding to each review link, the higher the final approval weight, the higher the importance of the corresponding review link; in this embodiment, if the final approval weight is between 0-1, the approval link is not reviewed; if the final approval weight is between 1-3, the approval link is a simple review; if the final approval weight is between 3-5, the approval link is a general review; if the final approval weight is greater than 5, the approval link is an important review. Finally, the approval links that are not reviewed are removed, and the review process obtained only has the approval links of simple review, general review and important review.

[0083] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0084] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A digital project application intelligent review method, characterized in that: The following steps are involved: Step S100: taking the project with the current content change as the target project, obtaining the change content of the target project, analyzing the change content, and extracting keywords from the change content; Step S200: The approval process of the target project includes several approval links, each of which has a corresponding approval record, and the approval record corresponding to each approval link is collected; based on the approval record and keywords, the correlation coefficient between the change content and each approval link is obtained; Step S300: obtaining a number of historical items, and obtaining the associated approval links corresponding to each approval link according to the number of approval links corresponding to each historical item and the approval records corresponding to each approval link; Step S400: Determine the re-approval process of the target project according to the correlation coefficient and the associated approval links corresponding to each approval link; Step S110: obtaining the changed content CH, and performing preprocessing to extract the text changed content and the digital changed content in the changed content CH; Step S120: taking the content between two adjacent punctuation marks in the text as a sentence fragment; If the digital change content is not empty, obtain the sentence fragment of the digital change content in the text corresponding to the target project, and extract the characteristic keywords in the sentence fragment; Step S130: if the text change content is not empty, extract the text keywords in the text change content, and combine with the characteristic keywords, extract all the keywords in the change content CH through word frequency statistics, and then obtain a keyword set, which includes each keyword and the number of each keyword; Step S210: The approval record includes the approval time and approval content; a certain approval record is set as AR, and the approval content corresponding to the approval record AR is AC. The approval content AC is preprocessed and relations are extracted to obtain all entities and definition relations in the approval content AC, and the knowledge graph corresponding to the approval content AC is obtained based on the entities and definition relations; Step S220: According to the keyword set and the knowledge graph, the semantic similarity between the kth keyword in the keyword set and the mth entity or nth definition relationship in the knowledge graph is obtained, and the result of the semantic similarity is normalized. If the semantic similarity is greater than the similarity threshold, the mth entity or nth definition relationship is marked at the corresponding position in the knowledge graph, and the weight of the marked position is the number corresponding to the kth keyword; and then all the marked positions of the knowledge graph about the changed content CH and the weight of each marked position are obtained; Step S230: Set the approval link corresponding to the approval record AR to AP, and the initial correlation coefficient X=0 between the change content CH and the approval link AP. If the mth entity and the nth definition relationship in the knowledge graph are related to each other and both are marked, obtain the weight W corresponding to the mth entity. m The weight W corresponding to the nth defined relationship n , and increase the value of the correlation coefficient X by W m +W0*W n , W0 is the relationship weight coefficient; and then the final correlation coefficient X between the change content CH and the approval link AP is obtained.

2. According to claim 1, a digital project application intelligent review method is characterized in that: Step S300 includes: Step S310: Taking the application time of the target project as the starting time, obtain the H historical projects that have completed approval starting from the starting time; obtain the pth and qth approval links corresponding to the hth historical project, as well as the corresponding approval records and , the approval record and The corresponding knowledge graph is and , approval time as and , the approval time is the time when the corresponding approval link starts; Step S320: Based on the knowledge graph and All entities and definition relationships corresponding to the above are combined into a set of related entities and definition relationships as a relationship combination; then the entity set is obtained. and , define a set of relationships and And the relationship combination set and ; The number of elements in the two entity sets that are the same is recorded as , the number of elements in the two defined relation sets that are the same is , the number of elements in the two relational combination sets that are the same is , and then get the approval record of the hth historical project and The corresponding data correlation coefficient; Step S330: Obtain the data correlation coefficients corresponding to the first, second, ... up to the Hth historical project. If there are H0 data correlation coefficients greater than the correlation coefficient threshold, where H0 < H, it is determined that there is a correlation between the pth and qth approval links. Furthermore, determine whether there is a correlation between any two approval links; Step S340: According to the associated pth and qth approval links, obtain the 1st, 2nd, For the p-th and q-th approval times corresponding to H historical projects, the number of times the p-th approval time is earlier than the q-th approval time is recorded as N1, and the number of times the p-th approval time is not earlier than the q-th approval time is recorded as N2; if N1>N2, then the q-th approval link is the associated approval link of the p-th approval link, if N1≤N2, then the p-th approval link is the associated approval link of the q-th approval link; and then the associated approval links corresponding to each approval link are obtained.

3. A digital project application intelligent review method according to claim 2, characterized in that: Step S400 includes: Step S410: Determine the approval weight V corresponding to the approval link AP according to the correlation coefficient between the change content CH and the approval link AP. AP ; Step S420: According to the approval weight V AP , and the associated approval link corresponding to the approval link AP, the approval weight corresponding to the associated approval link of the approval link AP is V RAP =V AP / B, B is the correlation weight coefficient; obtain the approval weight V obtained by the approval link AP according to the correlation coefficient AP , and the approval weight when it is an associated approval link, and add them together to get the final approval weight of the approval link AP; then get the final approval weight corresponding to each review link, and get the re-approval process of the target project, as well as the importance of each approval link in the re-approval process.

4. A project application intelligent review system, used to implement a digital project application intelligent review method as described in any one of claims 1 to 3, characterized in that: The system includes a keyword module, a correlation coefficient module, a correlation approval link module and a re-approval process module; Keyword module: used to take the current project with changed content as the target project, obtain the changed content of the target project, analyze the changed content, and extract keywords from the changed content; Correlation coefficient module: used to collect the approval records corresponding to each approval link according to the approval process of the target project, which includes several approval links, each of which has a corresponding approval record; and obtain the correlation coefficient corresponding to the change content and each approval link according to the approval record and keywords; Related approval link module: used to obtain several historical projects, and obtain the related approval links corresponding to each approval link according to the several approval links corresponding to each historical project and the approval records corresponding to each approval link; Re-approval process module: used to determine the re-approval process of the target project according to the correlation coefficient and the associated approval links corresponding to each approval link.

5. The project application intelligent review system according to claim 4 is characterized in that: The association coefficient module includes a knowledge graph unit, a mark weight unit and an association coefficient unit; Knowledge graph unit: used to obtain all entities and definition relationships in the approval content according to the approval content in the approval record, and obtain the knowledge graph corresponding to the approval content according to the entities and definition relationships; Marking weight unit: used to obtain all markings in the knowledge graph and the weight of each marking; Correlation coefficient unit: used to obtain the correlation coefficient corresponding to the change content and each approval link according to the approval record and keywords.

6. The project application intelligent review system according to claim 5 is characterized in that: The associated approval link module includes a historical project unit, a data association coefficient unit, a correlation unit and an associated approval link unit; Historical project unit: used to obtain any two approval links and corresponding approval records corresponding to historical projects, and obtain the knowledge graph and approval time corresponding to any two approval records; Data association coefficient unit: used to obtain the data association coefficient corresponding to any two approval records according to the entities, definition relationships and relationship combinations corresponding to the knowledge graphs corresponding to any two approval records; Relevance unit: used to obtain the data correlation coefficient corresponding to any two approval records corresponding to each historical item, and determine whether there is a correlation between any two approval links; The associated approval link unit is used to obtain the associated approval link corresponding to each approval link according to the association between any two approval links.

Images