Enterprise information retrieval method and device based on knowledge engine and electronic equipment

Abstract

The invention provides an enterprise information retrieval method and device based on a knowledge engine and electronic equipment, and relates to the field of information retrieval. In the method, to-be-retrieved requirements input by a user are obtained, and the to-be-retrieved requirements comprise a system retrieval requirement and a policy retrieval requirement; performing information crawling on a preset large model knowledge engine according to the system retrieval demand and the policy retrieval demand to obtain initial system information and initial policy information; the initial system information and the initial policy information are decomposed to obtain a plurality of sub-retrieval items, and the information is conveniently managed through information decomposition, so that subsequent accurate recommendation is facilitated; correlation calculation is performed on each sub-retrieval item to obtain correlation analysis, and correlation systems and policies can be recommended through correlation analysis, so that convenience is provided for user retrieval, and relatively easy information acquisition is realized; and performing information retrieval recommendation according to the correlation analysis to obtain a retrieval result, so that information meeting the input requirements of the user can be recommended.

Description

Technical Field

[0001] The present application relates to the field of information retrieval, and in particular to a knowledge engine-based enterprise information retrieval method, device and electronic device. Background Art

[0002] As the scale of enterprises expands, a sound internal management system can optimize internal processes, clarify the division of responsibilities, and improve collaboration efficiency. At the same time, timely understanding and adapting to external policy changes can ensure that enterprises operate efficiently under the premise of compliance. Therefore, the acquisition and update of internal management systems and external policy information of enterprises have become increasingly important. At present, most enterprises still use traditional document management methods to store and retrieve systems and policies, which cannot achieve intelligent retrieval, making it difficult for users to obtain the required systems or policies. Summary of the invention

[0003] The present application provides a knowledge engine-based enterprise information retrieval method, device and electronic device, which can perform intelligent retrieval and easily obtain corresponding systems and policies.

[0004] The technical solution of the embodiment of the present application is as follows: In a first aspect, an embodiment of the present application provides an enterprise information retrieval method based on a knowledge engine, the method comprising: Acquire the to-be-searched requirements input by the user, wherein the to-be-searched requirements include system search requirements and policy search requirements, wherein the to-be-searched requirements are voice search; According to the system retrieval requirements and policy retrieval requirements, information is crawled on a preset large model knowledge engine to obtain initial system information and initial policy information; Decomposing the initial system information and the initial policy information respectively to obtain a plurality of sub-search items; Calculate the correlation of each of the sub-search items to obtain a correlation analysis; An information retrieval recommendation is performed based on the association analysis to obtain a retrieval result, wherein the retrieval result is a voice answer corresponding to the voice retrieval.

[0005] In the above technical scheme, the requirements to be searched input by the user are obtained, and the requirements to be searched include system search requirements and policy search requirements, wherein the requirements to be searched are voice search, and by obtaining the corresponding requirements to be searched, it is convenient to automatically recommend the corresponding system or policy in the subsequent process; information crawling is performed on the preset large model knowledge engine according to the system search requirements and policy search requirements to obtain initial system information and initial policy information, and by automatically crawling information through the large model knowledge engine, not only the corresponding information can be obtained in time, but also the newer and more comprehensive system and policy information can be obtained; the initial system information and the initial policy information are decomposed respectively to obtain multiple sub-search items, and by performing information decomposition, it is convenient to manage the information for subsequent accurate recommendation; the correlation calculation is performed on each sub-search item to obtain the association analysis, and by performing the association analysis, the related systems and policies can be recommended, which provides convenience for user search and realizes easier information acquisition; information search recommendation is performed according to the association analysis to obtain the search result, wherein the search result is the voice answer corresponding to the voice search, which can recommend the information that meets the user input requirements, facilitates search and search, and the use of voice mode can also enable users to use the search function easily and quickly.

[0006] In some embodiments of the present application, the sub-search items include a plurality of system sub-items and a plurality of policy sub-items; The performing correlation calculation on each of the sub-search items to obtain the association analysis includes: Performing feature graph processing on each of the system sub-items to obtain a system feature sub-graph; Performing feature graph processing on each of the policy sub-items to obtain a policy feature sub-graph; The institutional characteristic subgraph and the policy characteristic subgraph are associated to obtain the association analysis.

[0007] In the above technical scheme, the feature graphs of the system sub-items and the policy sub-items are first processed respectively to obtain the system feature sub-graph and the policy feature sub-graph, and then the system feature sub-graph and the policy feature sub-graph are associated to obtain the association analysis. By calculating the feature sub-graph, the association relationship between the features can be calculated more clearly for subsequent retrieval and recommendation.

[0008] In some embodiments of the present application, associating the institutional feature subgraph with the policy feature subgraph to obtain the association analysis includes: Calculate the similarity between each institutional feature node in the institutional feature subgraph and the policy feature node in the policy feature subgraph to obtain a similarity; Calculating the correlation between the system feature subgraph and the policy feature subgraph using a preset association rule to obtain the correlation, wherein the association rule is obtained by extracting the relationship between the historical system and the historical policy; The similarity and the correlation are weightedly calculated to obtain the correlation analysis.

[0009] In the above technical solution, by calculating the similarity between the system characteristic subgraph and the policy characteristic subgraph, and calculating using association rules, and then performing weighted calculation, the related systems and policies can be retrieved, simplifying the complexity of retrieval and screening.

[0010] In some embodiments of the present application, the initial system information and the initial policy information are decomposed respectively to obtain multiple sub-search items, including: Classify the initial system information according to various departments in the enterprise to obtain department search sub-items; Divide the department search sub-items according to functions and role permissions to obtain multiple system sub-items; Dividing the initial policy information according to the business scope of the enterprise to obtain multiple policy sub-items; The various system sub-items and the various policy sub-items are combined into a plurality of the sub-search items.

[0011] In the above technical solution, the initial system information is first classified according to each department, and then divided according to function and role to obtain multiple system sub-items, which can manage the system and facilitate retrieval, making it easier to obtain retrieval results. The initial policy information is divided according to the business scope of the enterprise to obtain multiple policy sub-items, which realizes the management of the policy, which is conducive to the subsequent easier retrieval of policies related to the enterprise.

[0012] In some embodiments of the present application, performing information retrieval recommendation based on the association analysis to obtain retrieval results includes: According to the association analysis, the corresponding institutional feature nodes whose association analysis values ​​are greater than a preset threshold are selected as institutional association nodes, and the corresponding policy feature nodes whose association analysis values ​​are greater than a preset threshold are selected as policy association nodes; The system corresponding to the system association node and the policy corresponding to the policy association node are connected, and the connected results are recommended to obtain the search results.

[0013] In the above technical solution, the screening is performed according to the association analysis, and then the screened systems and policies are connected and recommended to obtain the search results. The related systems and policies can be obtained and the search needs of users can be met.

[0014] In some embodiments of the present application, performing information retrieval recommendation based on the association analysis to obtain retrieval results includes: In response to the search and screening operation of the user, screening the association analysis to obtain a screened association analysis; The filtered association analysis is matched with the user's operation records within the enterprise to generate the search result.

[0015] In the above technical solution, through user-defined screening operations, user-personalized retrieval can be achieved, and corresponding retrieval results can be obtained to meet the user's retrieval needs.

[0016] In some embodiments of the present application, after obtaining the to-be-searched requirement input by the user, the method further includes: Verify the role of the user, and if the role of the user is a project team member, verify the authority of the user to obtain access rights; In a case where the access right has multiple access rights, determining that the to-be-searched requirement input by the user is a parallel search requirement; In a case where the access permission is non-parallel access, it is determined that the to-be-retrieved requirement input by the user is a single retrieval requirement.

[0017] In the above technical solution, by verifying the role and confirming that the user is a member of the project team, access rights verification is performed to ensure the security of the search. In addition, when the user has access rights, parallel search requirements and single search requirements are determined to achieve parallel search and meet diverse search requirements.

[0018] In a second aspect, an embodiment of the present application provides an enterprise information retrieval device based on a knowledge engine, the device comprising: A data acquisition module is used to acquire the to-be-searched requirements input by the user, wherein the to-be-searched requirements include system search requirements and policy search requirements, wherein the to-be-searched requirements are voice searches; A data search module is used to crawl information on a preset large model knowledge engine according to the system retrieval requirements and policy retrieval requirements to obtain initial system information and initial policy information; A data decomposition module, used for decomposing the initial system information and the initial policy information respectively to obtain a plurality of sub-search items; A correlation calculation module, used to calculate the correlation of each of the sub-search items to obtain a correlation analysis; The retrieval recommendation module is used to perform information retrieval recommendation based on the association analysis to obtain retrieval results, wherein the retrieval results are voice answers corresponding to the voice retrieval.

[0019] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor, a memory, a user interface, a communication bus and a network interface, wherein the processor, the memory, the user interface and the network interface are respectively connected to the communication bus, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes any one of the methods provided in the first aspect.

[0020] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed, any one of the methods provided in the first aspect is executed.

[0021] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Since the user inputs the to-be-searched requirements, the to-be-searched requirements include the system search requirements and the policy search requirements, wherein the to-be-searched requirements are voice searches. By obtaining the corresponding to-be-searched requirements, it is convenient to automatically recommend the corresponding system or policy in the future; information crawling is performed on the preset large model knowledge engine according to the system search requirements and the policy search requirements to obtain the initial system information and the initial policy information. By automatically crawling the information through the large model knowledge engine, not only the corresponding information can be obtained in time, but also the newer and more comprehensive system and policy information can be obtained; the initial system information and the initial policy information are decomposed respectively to obtain multiple sub-search items. By performing information decomposition, it is convenient to manage the information for subsequent accurate recommendation; the relevance of each sub-search item is calculated to obtain the association analysis. By performing the association analysis, the related systems and policies can be recommended, which provides convenience for user search and realizes easier information acquisition; information search recommendation is performed according to the association analysis to obtain the search result, wherein the search result is the voice answer corresponding to the voice search, which can recommend the information that meets the user input requirements, facilitates search and search, and the use of voice mode can also enable users to use the search function easily and quickly. Therefore, the problem that relevant technologies cannot realize intelligent retrieval, making it difficult for users to obtain the required systems or policies is effectively solved.

[0022] 2. By associating and matching the association analysis with the user's historical operation records, the recommended search results are more in line with the user's search needs.

[0023] 3. By judging the user's role and permissions, corresponding systems and policies can be recommended based on the user's search permissions to achieve security management and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1This is one of the flow diagrams of a knowledge engine-based enterprise information retrieval method provided by an embodiment of the present application; Figure 2 yes Figure 1 A schematic flow chart of a sub-step of step S300; Figure 3 yes Figure 1 A schematic flow chart of a sub-step of step S400; Figure 4 yes Figure 3 A schematic flow chart of a sub-step of step S430; Figure 5 yes Figure 1 A schematic flow chart of a sub-step of step S500; Figure 6 It is a schematic diagram of the structure of a knowledge engine-based enterprise information retrieval device provided by an embodiment of the present application; Figure 7 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0025] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0026] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0027] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0028] The embodiment of the present application provides a knowledge engine-based enterprise information retrieval method, device, electronic device and readable storage medium. The knowledge engine-based enterprise information retrieval method first obtains the to-be-retrieved requirements input by the user, and the to-be-retrieved requirements include system retrieval requirements and policy retrieval requirements, wherein the to-be-retrieved requirements are voice retrievals. By obtaining the corresponding to-be-retrieved requirements, it is beneficial to automatically recommend the corresponding system or policy in the subsequent process. According to the system retrieval requirements and policy retrieval requirements, information crawling is performed on a preset large model knowledge engine to obtain initial system information and initial policy information. By automatically crawling information through the large model knowledge engine, not only can the corresponding information be obtained in a timely manner, but also the newer and more recent ones can be obtained. Comprehensive system and policy information; decompose the initial system information and initial policy information separately to obtain multiple sub-search items. By decomposing the information, it is convenient to manage the information for subsequent accurate recommendations; calculate the correlation of each sub-search item to obtain association analysis. Through association analysis, related systems and policies can be recommended to facilitate user retrieval and achieve easier information acquisition; make information retrieval recommendations based on association analysis to obtain retrieval results, among which the retrieval results are voice answers corresponding to voice retrieval, which can recommend information that meets the user's input requirements, facilitate retrieval and search, and the use of voice mode can also enable users to use the retrieval function easily and quickly.

[0029] It should be noted that the enterprise information retrieval method based on the knowledge engine is applied to the system and policy retrieval of the medical industry and enterprises related to the medical industry, and can also be applied to the system and policy retrieval of other industries. In the medical industry, there are many different research and medication standards for clinical trials. The enterprise information retrieval method based on the knowledge engine can help users quickly retrieve the corresponding systems and standards for easy viewing at any time.

[0030] The technical solution provided in the embodiments of the present application is further described below in conjunction with the accompanying drawings.

[0031] Reference Figure 1 , Figure 1 1 is a flow chart of a knowledge engine-based enterprise information retrieval method provided in an embodiment of the present application. The knowledge engine-based enterprise information retrieval method is applied to a knowledge engine-based enterprise information retrieval device, and the knowledge engine-based enterprise information retrieval method is executed by a processor in an electronic device or a readable storage medium. The knowledge engine-based enterprise information retrieval method includes steps S100, S200, S300, S400, and S500.

[0032] Step S100, obtaining the to-be-searched requirements input by the user, the to-be-searched requirements include system search requirements and policy search requirements, wherein the to-be-searched requirements are voice searches.

[0033] In one embodiment, the demand to be retrieved includes a system retrieval demand and a policy retrieval demand. The system retrieval demand is the retrieval demand for the relevant systems of various departments within the enterprise, and the policy retrieval demand is the demand for the relevant policies of the corresponding departments in the development of the enterprise and the demand for the relevant policies of the medical industry. First, the user logs in to the information retrieval engine, and the user can enter the demand to be retrieved. For example, the drug standard system and drug policy specifications of the clinical trial can be retrieved. The demand to be retrieved for input can also be only the system retrieval demand, or only the policy retrieval demand, or it can also include the system retrieval demand and the policy retrieval demand at the same time. According to the user's input, the user's input method can choose the voice input method or the text input method. For example, taking the voice input method as an example, the voice input button can be clicked, and the user can describe the demand by voice. Through the front-end and back-end transmission functions of the search engine, the demand to be retrieved input by the user is obtained, so as to facilitate the subsequent automatic recommendation of the corresponding system or policy.

[0034] In one embodiment, after obtaining the requirements to be retrieved input by the user, the enterprise information retrieval method based on the knowledge engine also includes but is not limited to: verifying the user's role, and when the user's role is a project team member, verifying the user's authority to obtain access rights; when the access right is a multiple access rights, determining that the requirements to be retrieved input by the user are parallel retrieval requirements.

[0035] Specifically, the user's role is verified to see if the role is a member of the corresponding department. Only when the user is a member of the corresponding department can the corresponding system of the corresponding department be viewed. The policy information is automatically crawled and sorted and divided into corresponding departments, and can also be viewed after the role verification is passed. In the case where the user's role is a member of the project team, the user's permissions are verified to obtain access rights. This is because project team members may have different experimental projects executed in parallel, and for different project systems and standards, the user's access rights are first verified to determine whether to perform parallel retrieval. In the case where the access right has multiple access rights, it indicates that the user has retrieval requirements for multiple systems and policies, and the requirements to be retrieved entered by the user are determined to be parallel retrieval requirements. Multi-threading is used to obtain information for parallel retrieval requirements to speed up processing efficiency. Through the above-mentioned permission verification, security management and control are carried out to achieve data security.

[0036] In another embodiment, after obtaining the requirements to be retrieved input by the user, the enterprise information retrieval method based on the knowledge engine also includes but is not limited to: verifying the user's role, and when the user's role is a project team member, verifying the user's authority to obtain access rights; when the access rights are non-parallel access, determining that the requirements to be retrieved input by the user are single retrieval requirements.

[0037] Specifically, the user role is first verified to see if the role is a member of the corresponding department. Only when the user is a member of the corresponding department can he view the corresponding system and divided policies of the corresponding department. When the user's role is a member of the project team, the user's authority is verified to obtain access rights. This is because project team members may have different experimental projects executed in parallel, and the systems and standards of different projects are different. The user's access rights are first verified. When the access rights are non-parallel access, it indicates that the user does not have the retrieval needs of multiple systems and policies. The retrieval requirements input by the user are determined to be single retrieval requirements, and a single thread is used to obtain information for a single retrieval requirement to save resources. Through the above-mentioned permission verification, security management and control are carried out to achieve data security.

[0038] Step S200, crawling information on a preset large model knowledge engine according to system retrieval requirements and policy retrieval requirements to obtain initial system information and initial policy information.

[0039] In one embodiment, the preset large model knowledge engine integrates the document repository of departmental systems, the professional knowledge base of various industries, and the policy information obtained from official websites, authoritative media, professional research institutions, etc., with massive data, and uses the industry large model to mine knowledge, and uses manual screening to form crawlable data. By automatically crawling information on the large model knowledge engine, not only can the corresponding information be obtained in a timely manner, but also the newer system and policy information can be obtained, so as to control the research direction of clinical trials. First, the target and scope to be crawled are determined according to the system retrieval requirements and policy retrieval requirements. For the crawling of departmental systems, the document repository of departmental systems can be retrieved, or the relevant system information of industry-related enterprises can be retrieved from the professional knowledge base of various industries to obtain the initial system information. For the crawling of policy information, the corresponding information can be mined from various public institutions through the industry large model mining to obtain the initial policy information. By crawling information through the large model knowledge engine, not only can the corresponding information be obtained in a timely manner, but also the newer and more comprehensive system and policy information can be obtained. Among them, the industry large model is based on technologies such as P-Tuning and LoRa. On the basis of a small amount of computing power, the effect of the task is improved, and the cost of computing resources can be saved.

[0040] Step S300, decomposing the initial system information and the initial policy information respectively to obtain a plurality of sub-search items.

[0041] like Figure 2 As shown, the initial system information and the initial policy information are decomposed respectively to obtain multiple sub-search items, including but not limited to the following steps: Step S310, classify the initial system information according to the various departments in the enterprise to obtain department search sub-items.

[0042] In one embodiment, the initial system information and initial policy information crawled in step S200 are sorted so that relevant data can be retrieved more easily later. First, the initial system information is classified according to the various departments in the enterprise. Since the system of each department is different, for example, there are emergency departments, clinical research departments, etc. in the clinic, and the systems of various outpatient departments are different. The department search sub-items are obtained by classification according to the department, and the corresponding system can be divided to avoid wasting too much time in searching and finding, which facilitates the acquisition of the system.

[0043] It should be noted that the system information in each department can be repeated or different. The corresponding access rights for different system information are higher. After passing the access verification, the corresponding system can be retrieved and viewed. For the same system information, personnel from each department can access the corresponding system information through the department interface, which can speed up the retrieval efficiency.

[0044] Step S320, dividing the department search sub-items according to functions and role permissions to obtain multiple system sub-items.

[0045] In one embodiment, since different user roles have different access rights, and the systems referenced by each project in the same department may also be different, after being divided by department, the department search sub-project is divided according to function and role authority to obtain multiple system sub-projects. By dividing, it is convenient to manage information, which is conducive to subsequent rapid retrieval and recommendation.

[0046] For example, in clinical trials, functions refer to the medication standards used in project implementation, the execution standard systems for different tests, etc.; role permissions can be for the team leader and members of the project team, and the team leader has higher access rights than the members. Dividing the systems according to different projects can avoid viewing the systems of other projects, which not only ensures security but also facilitates retrieval and query. After dividing according to role permissions, access to the system is restricted. For example, members do not need to retrieve some content that can be reduced to the team leader's permissions, saving time, thereby facilitating user retrieval and enabling easier information acquisition.

[0047] Step S330, dividing the initial policy information according to the business scope of the enterprise to obtain multiple policy sub-items.

[0048] In one embodiment, in clinical trials, the initial policy information includes approval procedures, financial subsidies, taxes, etc. The business scope is obtained based on the registration information of the medical institution, and the business scope of the enterprise can be obtained by reading the enterprise information registration. A connection is established between the initial policy information and the business scope of the enterprise, and the initial policy information is divided to obtain multiple policy sub-items, thereby facilitating user retrieval so as to achieve easier information acquisition. Among them, the policy orientation can be matched with the business scope of the enterprise through a preset matching algorithm to establish an association. The preset matching algorithm can be a natural language processing algorithm, such as a word segmentation algorithm, a deep learning algorithm, etc.

[0049] It should be noted that, since policy information may be continuously updated, the initial policy information is obtained at a preset time, and a connection is established between the obtained initial policy information and the business scope. Policy information related to the initial policy information can also be updated based on the previous policy to save resources.

[0050] Step S340, grouping each system sub-item and each policy sub-item into multiple sub-search items.

[0051] In one embodiment, each system sub-item obtained in step S320 and each policy sub-item obtained in step S330 are used as sub-search items and form a project group set, which is conducive to subsequent search recommendations based on the sub-search items.

[0052] Step S400, performing correlation calculation on each sub-search item to obtain a correlation analysis.

[0053] In one embodiment, the sub-search items include multiple system sub-items and multiple policy sub-items, such as Figure 3 As shown, the correlation calculation is performed on each sub-search item to obtain the association analysis, including but not limited to the following steps: Step S410, performing feature graph processing on each system sub-item to obtain a system feature sub-graph.

[0054] In one embodiment, for each system sub-item, a graph neural network is used to perform keyword recognition to obtain system keyword features, and the system keyword features are used as nodes to extract association relationships for each system sub-item to obtain the connections between system keyword features and establish a system feature sub-graph so that the connections between sub-graphs can be calculated later and association analysis can be performed. For each system sub-item, a graph neural network can also be used to extract the relationships between each system sub-item and establish the connections between each system sub-item.

[0055] Step S420, performing feature graph processing on each policy sub-item to obtain a policy feature sub-graph.

[0056] In one embodiment, for each policy sub-item, a graph neural network is used to perform keyword recognition to obtain policy keyword features, and the policy keyword features are used as nodes to extract association relationships for each policy sub-item to obtain the connections between policy keyword features and establish a policy feature sub-graph so that the connections between sub-graphs can be subsequently calculated and association analysis can be performed. For each policy sub-item, a graph neural network can also be used to extract the relationships between each policy sub-item and establish connections between each policy sub-item.

[0057] Step S430, associating the institutional characteristic subgraph with the policy characteristic subgraph to obtain an association analysis.

[0058] In one embodiment, based on the institutional characteristic subgraph obtained in step S410 and the policy characteristic subgraph obtained in step S420, a connection is established between the characteristic graphs to obtain a correlation analysis to facilitate subsequent retrieval.

[0059] like Figure 4 As shown, the institutional feature subgraph and the policy feature subgraph are associated to obtain an association analysis, including but not limited to the following steps: Step S431, calculating the similarity between each institutional feature node in the institutional feature subgraph and the policy feature node in the policy feature subgraph to obtain the similarity.

[0060] In one embodiment, a preset keyword matching algorithm is used to calculate the similarity between each institutional feature node in the institutional feature subgraph and the policy feature node in the policy feature subgraph, and obtain the similarity, which is represented by a similarity value for subsequent correlation calculation. By performing similar correlation calculation, related institutions and policies can be recommended, which facilitates user retrieval and enables easier information acquisition. Among them, the preset keyword matching algorithm can be a longest common substring matching algorithm, or a string matching algorithm, etc.

[0061] Step S432, using a preset association rule to calculate the association between the system characteristic subgraph and the policy characteristic subgraph to obtain the association, wherein the association rule is obtained by extracting the relationship between the historical system and the historical policy.

[0062] In one embodiment, the preset association rules are obtained by extracting the relationship between historical systems and historical policies, wherein the historical system is the old system adopted by the enterprise, or the system adopted by enterprises with similar operations to the enterprise. The historical policy is the old policy previously announced, or the policy related to the policy. The historical system and historical policy are crawled from the Internet using crawler technology, and the amount of data is large. According to the acquired historical system and historical policy, the historical system and historical policy are firstly preprocessed, and the data preprocessing includes data cleaning, data conversion and standardization processing. Then, the processed historical system and historical policy are processed with feature graphs to obtain the historical system feature graph corresponding to the historical system and the historical policy feature graph corresponding to the historical policy. The initial association rules are trained using a large number of historical system feature graphs and historical policy feature graphs to obtain the trained association rules, that is, the preset association rules. The preset association rules are used to calculate the correlation between the system feature subgraph and the policy feature subgraph to obtain the correlation, which is expressed in numerical form for subsequent association analysis calculation. By performing correlation calculation, the associated system and policy can be recommended, which is convenient for user retrieval and realizes easier information acquisition. The preset association rules may be a mutual information algorithm, a covariance matrix, a deep learning model, etc. Since there are multiple system feature graphs and policy feature graphs, there are also multiple correlations obtained.

[0063] Step S433: perform weighted calculation on the similarity and the correlation to obtain correlation analysis.

[0064] In one embodiment, similarity is obtained according to step S431, and the similarity values ​​are first sorted, either in ascending order or in descending order. Exemplarily, the similarity values ​​are sorted in descending order, and the preset number of similarity values ​​arranged in front is selected to perform weighted calculation with the correlation. Among them, the preset number is the same as the number of correlations. Then the similarity and correlation are weightedly calculated, and corresponding weights can be assigned to the similarity and correlation respectively. The weights are obtained based on the experience of professionals and will not be repeated here. The proportion of similarity and correlation is adjusted by weight to obtain correlation analysis. The correlation analysis is also expressed in numerical form, and the larger the value, the stronger the correlation. So that subsequent retrieval can recommend retrieval results that meet the needs.

[0065] Step S500, performing information retrieval recommendation based on association analysis to obtain retrieval results, wherein the retrieval results are voice answers corresponding to the voice retrieval.

[0066] like Figure 5 As shown, information retrieval recommendation is performed based on association analysis to obtain retrieval results, including but not limited to the following steps: Step S510, according to the association analysis, selecting the corresponding system feature nodes whose association analysis values ​​are greater than the preset threshold as system association nodes, and selecting the corresponding policy feature nodes whose association analysis values ​​are greater than the preset threshold as policy association nodes.

[0067] In one embodiment, since the association analysis is expressed in numerical form, i.e., the association analysis value, according to the association analysis obtained in step S400, the corresponding system feature node whose association analysis value is greater than the preset threshold is selected as the system association node, indicating that the system information corresponding to the system association node meets the user's search requirements, and the corresponding policy feature node whose association analysis value is greater than the preset threshold is selected as the policy association node, indicating that the policy information corresponding to the policy association node meets the user's search requirements, so as to make subsequent recommendations. Among them, the preset threshold is set by professionals based on experience, and will not be elaborated here.

[0068] It should be noted that, when there are multiple correlation analysis values ​​that are all greater than a preset threshold, the institutional characteristic node and policy characteristic node corresponding to the maximum correlation analysis value among the correlation analysis values ​​are selected.

[0069] In another embodiment, when the association analysis value is less than or equal to a preset threshold, the association analysis values ​​are sorted, exemplarily in descending order, and then the institutional characteristic nodes corresponding to the top N association analysis values ​​in the association analysis value are selected as institutional association nodes, and the corresponding policy characteristic nodes are selected as policy association nodes.

[0070] Step S520, connecting the system corresponding to the system-related node and the policy corresponding to the policy-related node, and recommending the connected results to obtain the search results.

[0071] In one embodiment, according to the system-related nodes and policy-related nodes obtained in step S510, the system corresponding to the system-related nodes and the policy corresponding to the policy-related nodes are string-connected, that is, the system and the policy are combined into one record, and the connected results are recommended to obtain search results, which can recommend information that meets the user's input requirements, making it easy to search and find.

[0072] It should be noted that, in the case of parallel search requirements, the search results corresponding to each search requirement are displayed in a table form for the convenience of users. In the case of a single search requirement, it can also be displayed in a table form, with only one record in the table.

[0073] In one embodiment, information retrieval recommendation is performed based on association analysis to obtain retrieval results, including but not limited to: responding to the user's retrieval filtering operation, filtering the association analysis to obtain a filtered association analysis; matching the filtered association analysis with the user's operation records within the enterprise to generate retrieval results.

[0074] In some possible embodiments of the present application, the user can also selectively search for systems and policies, and can also impose conditional restrictions on the recommended search results. In response to the user's search and filtering operation, the association analysis is filtered to obtain a filtered association analysis. Exemplarily, the user's search and filtering operation is to only view the search system or only view the search policy, then the system is filtered out or the policy is filtered out, and the filtered association analysis is a system analysis or a policy analysis. Since the user has operation records within the enterprise, the user's operation records within the enterprise can be read, and the operation records can be matched with the filtered association analysis, and the filtered association analysis can be matched with the user's operation records within the enterprise. A partial character matching algorithm can be used to generate a search result that better meets the user's search needs.

[0075] It should be noted that the search results are presented to the user in the form of voice and displayed in the form of dialogue on the front-end page, allowing users to use the search function easily and quickly.

[0076] like Figure 6As shown, the embodiment of the present application provides an enterprise information retrieval device 100 based on a knowledge engine. The enterprise information retrieval device 100 based on the knowledge engine obtains the to-be-retrieved requirements input by the user through a data acquisition module 110. The to-be-retrieved requirements include system retrieval requirements and policy retrieval requirements, wherein the to-be-retrieved requirements are voice retrievals. By obtaining the corresponding to-be-retrieved requirements, it is convenient to automatically recommend the corresponding system or policy in the subsequent process. Then, the data search module 120 crawls information on a preset large model knowledge engine according to the system retrieval requirements and the policy retrieval requirements to obtain initial system information and initial policy information. By automatically crawling information through the large model knowledge engine, not only can the corresponding information be obtained in a timely manner, but also relatively new and comprehensive system and policy information can be obtained. ; Then use the data decomposition module 130 to decompose the initial system information and the initial policy information respectively to obtain multiple sub-search items. By decomposing the information, it is convenient to manage the information for subsequent accurate recommendations; use the correlation calculation module 140 to calculate the correlation of each sub-search item to obtain an association analysis. By performing the association analysis, it is possible to recommend related systems and policies, which facilitates user retrieval and enables easier information acquisition; use the retrieval recommendation module 150 to perform information retrieval recommendations based on the association analysis to obtain retrieval results, wherein the retrieval results are voice answers corresponding to voice retrieval, which can recommend information that meets the user's input requirements, facilitate retrieval and search, and the use of voice mode can also enable users to use the retrieval function easily and quickly.

[0077] It should be noted that the data acquisition module 110 is connected to the data search module 120, the data search module 120 is connected to the data decomposition module 130, the data decomposition module 130 is connected to the correlation calculation module 140, and the correlation calculation module 140 is connected to the retrieval recommendation module 150. The above-mentioned knowledge engine-based enterprise information retrieval method is applied to a knowledge engine-based enterprise information retrieval device 100. The knowledge engine-based enterprise information retrieval device 100 obtains the requirements to be retrieved input by the user, and the requirements to be retrieved include system retrieval requirements and policy retrieval requirements, wherein the requirements to be retrieved are voice retrievals. By obtaining the corresponding requirements to be retrieved, it is beneficial to automatically recommend the corresponding system or policy in the subsequent process; according to the system retrieval requirements and policy retrieval requirements, information crawling is performed on the preset large model knowledge engine to obtain initial system information and initial policy information. By automatically crawling information through the large model knowledge engine, not only can the corresponding information be obtained in a timely manner, but also the newer and relatively newer ones can be obtained. Comprehensive system and policy information; decompose the initial system information and initial policy information separately to obtain multiple sub-search items. By decomposing the information, it is convenient to manage the information for subsequent accurate recommendations; calculate the correlation of each sub-search item to obtain association analysis. Through association analysis, related systems and policies can be recommended to facilitate user retrieval and achieve easier information acquisition; make information retrieval recommendations based on association analysis to obtain retrieval results, among which the retrieval results are voice answers corresponding to voice retrieval, which can recommend information that meets the user's input requirements, facilitate retrieval and search, and the use of voice mode can also enable users to use the retrieval function easily and quickly.

[0078] It should also be noted that: when the device provided in the above embodiment realizes its function, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0079] The present application also discloses an electronic device. Figure 7 , Figure 7 The electronic device 500 may include: at least one processor 501 , at least one network interface 504 , a user interface 503 , a memory 505 , and at least one communication bus 502 .

[0080] The communication bus 502 is used to realize the connection and communication between these components.

[0081] The user interface 503 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 503 may also include a standard wired interface and a wireless interface.

[0082] The network interface 504 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0083] Among them, the processor 501 may include one or more processing cores. The processor 501 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 505, and calling data stored in the memory 505. Optionally, the processor 501 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 501 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 501, and it can be implemented separately through a chip.

[0084] Among them, the memory 505 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 505 includes a non-transitory computer-readable storage medium. The memory 505 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 505 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 505 may also be optionally at least one storage device located away from the aforementioned processor 501. Refer to Figure 7, the memory 505 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of an enterprise information retrieval method based on a knowledge engine.

[0085] exist Figure 7 In the electronic device 500 shown, the user interface 503 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 501 can be used to call the application program stored in the memory 505 for an enterprise information retrieval method based on a knowledge engine, and when executed by one or more processors 501, the electronic device 500 executes one or more methods in the above-mentioned embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0086] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0087] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0088] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0089] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0090] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0091] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, those skilled in the art will easily think of other embodiments of the present disclosure.

[0092] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A knowledge engine-based enterprise information retrieval method, characterized in that: The method comprises: Acquire the to-be-searched requirements input by the user, wherein the to-be-searched requirements include system search requirements and policy search requirements, wherein the to-be-searched requirements are voice search; According to the system retrieval requirements and policy retrieval requirements, information is crawled on a preset large model knowledge engine to obtain initial system information and initial policy information; Decomposing the initial system information and the initial policy information respectively to obtain a plurality of sub-search items; Calculate the correlation of each of the sub-search items to obtain a correlation analysis; An information retrieval recommendation is performed based on the association analysis to obtain a retrieval result, wherein the retrieval result is a voice answer corresponding to the voice retrieval.

2. The method according to claim 1, characterized in that The sub-search items include multiple system sub-items and multiple policy sub-items; The performing correlation calculation on each of the sub-search items to obtain the association analysis includes: Performing feature graph processing on each of the system sub-items to obtain a system feature sub-graph; Performing feature graph processing on each of the policy sub-items to obtain a policy feature sub-graph; The institutional characteristic subgraph and the policy characteristic subgraph are associated to obtain the association analysis.

3. The method according to claim 2, characterized in that The step of associating the system feature subgraph with the policy feature subgraph to obtain the association analysis includes: Calculate the similarity between each institutional feature node in the institutional feature subgraph and the policy feature node in the policy feature subgraph to obtain a similarity; Calculating the correlation between the system characteristic subgraph and the policy characteristic subgraph by using a preset association rule to obtain the correlation, wherein the association rule is obtained by extracting the relationship between the historical system and the historical policy; The similarity and the correlation are weightedly calculated to obtain the correlation analysis.

4. The method according to claim 1, characterized in that The initial system information and the initial policy information are decomposed respectively to obtain a plurality of sub-search items, including: Classify the initial system information according to various departments in the enterprise to obtain department search sub-items; Divide the department search sub-items according to functions and role permissions to obtain multiple system sub-items; Dividing the initial policy information according to the business scope of the enterprise to obtain multiple policy sub-items; The various system sub-items and the various policy sub-items are combined into a plurality of the sub-search items.

5. The method according to claim 1, characterized in that The information retrieval recommendation is performed according to the association analysis to obtain the retrieval results, including: According to the association analysis, the corresponding institutional feature nodes whose association analysis values ​​are greater than a preset threshold are selected as institutional association nodes, and the corresponding policy feature nodes whose association analysis values ​​are greater than a preset threshold are selected as policy association nodes; The system corresponding to the system association node and the policy corresponding to the policy association node are connected, and the connected results are recommended to obtain the search results.

6. The method according to claim 1, characterized in that The information retrieval recommendation is performed according to the association analysis to obtain the retrieval results, including: In response to the search and screening operation of the user, screening the association analysis to obtain a screened association analysis; The filtered association analysis is matched with the user's operation records within the enterprise to generate the search result.

7. The method according to claim 1, characterized in that After obtaining the to-be-searched requirement input by the user, the method further includes: Verify the role of the user, and if the role of the user is a project team member, verify the authority of the user to obtain access rights; In a case where the access right has multiple access rights, determining that the to-be-searched requirement input by the user is a parallel search requirement; In a case where the access permission is non-parallel access, it is determined that the to-be-retrieved requirement input by the user is a single retrieval requirement.

8. An enterprise information retrieval device based on a knowledge engine, characterized in that: The device comprises: A data acquisition module (110) is used to acquire the to-be-searched requirements input by the user, wherein the to-be-searched requirements include system search requirements and policy search requirements, wherein the to-be-searched requirements are voice searches; A data search module (120) is used to crawl information on a preset large model knowledge engine according to the system search requirements and policy search requirements to obtain initial system information and initial policy information; A data decomposition module (130) is used to decompose the initial system information and the initial policy information respectively to obtain a plurality of sub-search items; A correlation calculation module (140) is used to perform correlation calculation on each of the sub-search items to obtain a correlation analysis; A retrieval recommendation module (150) is used to perform information retrieval recommendation based on the association analysis to obtain a retrieval result, wherein the retrieval result is a voice answer corresponding to the voice retrieval.

9. An electronic device, characterized in that: The electronic device (500) comprises a processor (501), a memory (505), a user interface (503), a communication bus (502) and a network interface (504), wherein the processor (501), the memory (505), the user interface (503) and the network interface (504) are respectively connected to the communication bus (502), the memory (505) is used to store instructions, the user interface (503) and the network interface (504) are used to communicate with other devices, and the processor (501) is used to execute the instructions stored in the memory (505) so that the electronic device (500) executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is performed.

Images