Expert portrait construction method and device in mine field and electronic equipment

Abstract

The invention provides an expert portrait construction method and device in the mine field and electronic equipment, and the method comprises the steps: carrying out the preprocessing of expert data, and constructing a standard expert data set; obtaining a basic label, a personalized label and an associated label of each expert in the standard expert data set according to the standard expert data set, and constructing an expert label index according to the basic label, the personalized label and the associated label; searching target expert information required in the target scene according to the expert label index to obtain the target expert information; inputting the target expert information into a pre-trained mine industry large model to obtain an expert evaluation label; the expert portrait is constructed according to the target expert information and the expert evaluation label, the expert label index is constructed by performing mining analysis on the standard expert data set, the expert evaluation label is obtained based on the mine industry large model, and the expert portrait is constructed according to the target expert information and the expert evaluation label, so that the comprehensiveness and the accuracy of the expert portrait are ensured.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, device and electronic equipment for constructing expert portraits in the mining field. Background Art

[0002] In the related technologies, for the mining field, it is usually necessary to rely on manual collection and organization of expert information, and to construct expert portraits based on the manually collected and organized expert information. However, the above method is not only inefficient, but also prone to missing key expert information. At the same time, the expert information is not fully mined in depth, resulting in the constructed expert portraits being not comprehensive and accurate. How to construct a comprehensive and accurate expert portrait has become a problem that needs to be solved urgently. Summary of the invention

[0003] The purpose of this application is to solve one of the technical problems in the above technology at least to a certain extent.

[0004] The first aspect of the present application provides a method for constructing an expert portrait in the mining field, comprising: obtaining multi-source expert data in the mining field, preprocessing the expert data, and constructing a standard expert data set; according to the standard expert data set, obtaining the basic label, personalized label and associated label of each expert in the standard expert data set, and constructing an expert label index according to the basic label, the personalized label and the associated label; searching for the target expert information required in the target scenario according to the expert label index to obtain the target expert information; inputting the target expert information into a pre-trained mining industry large model to obtain expert evaluation labels; and constructing an expert portrait according to the target expert information and the expert evaluation labels.

[0005] The second aspect of the present application provides an expert portrait construction device in the mining field, including: a first acquisition module, used to acquire multi-source expert data in the mining field, pre-process the expert data, and construct a standard expert data set; a second acquisition module, used to acquire the basic label, personalized label and associated label of each expert in the standard expert data set according to the standard expert data set, and construct an expert label index according to the basic label, the personalized label and the associated label; a search module, used to search for the target expert information required in the target scenario according to the expert label index to obtain the target expert information; a third acquisition module, used to input the target expert information into a pre-trained mining industry large model to obtain expert evaluation labels; a portrait construction module, used to construct an expert portrait according to the target expert information and the expert evaluation labels.

[0006] The third aspect of the present application provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for constructing an expert portrait in the mining field provided in the first aspect of the present application.

[0007] The fourth aspect of the present application provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to execute the method for constructing an expert portrait in the mining field provided in the first aspect of the present application.

[0008] The fifth aspect of the present application provides a computer program product. When the instruction processor in the computer program product is executed, the method for constructing expert portraits in the mining field provided in the first aspect of the present application is executed.

[0009] The method and device for constructing expert portraits in the mining field provided by the present application obtain expert data from multiple sources in the mining field, pre-process the expert data, and construct a standard expert data set. According to the standard expert data set, the basic labels, personalized labels and associated labels of each expert in the standard expert data set are obtained, and an expert label index is constructed based on the basic labels, personalized labels and associated labels. According to the expert label index, the target expert information required in the target scenario is searched to obtain the target expert information, and the target expert information is input into the pre-trained mining industry big model to obtain the expert evaluation labels. According to the target expert information and the expert evaluation labels, an expert portrait is constructed. Therefore, the present application mines and analyzes the standard expert data set to construct an expert label index, obtains the expert evaluation labels based on the mining industry big model, and constructs the expert portrait according to the target expert information and the expert evaluation labels, thereby ensuring the comprehensiveness and accuracy of the expert portrait.

[0010] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0012] Figure 1 A schematic diagram of a process of constructing an expert portrait in the mining field according to an embodiment of the present application;

[0013] Figure 2 A schematic diagram of a process of constructing an expert portrait in the mining field according to another embodiment of the present application;

[0014] Figure 3 A schematic diagram of a process of constructing an expert portrait in the mining field according to another embodiment of the present application;

[0015] Figure 4 A schematic diagram of a process of constructing an expert portrait in the mining field according to another embodiment of the present application;

[0016] Figure 5 A schematic diagram of the structure of a device for constructing an expert portrait in the mining field according to an embodiment of the present application;

[0017] Figure 6 It is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0018] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0019] The following describes the method, device, electronic device and medium for constructing an expert portrait in the mining field in an embodiment of the present application with reference to the accompanying drawings.

[0020] Figure 1 FIG. 1 is a flow chart of a method for constructing an expert portrait in the mining field according to an embodiment of the present application. Figure 1 As shown, the method includes:

[0021] S101, obtaining multi-source expert data in the mining field, preprocessing the expert data, and constructing a standard expert data set.

[0022] It should be noted that the present application does not limit the specific method of acquiring multi-source expert data in the mining field, and it can be selected according to actual conditions.

[0023] Optionally, expert data may be obtained from databases such as mining industry databases, scientific research institution databases, and university databases, and expert data may also be obtained from Internet resources such as academic papers, patents, news reports, and professional forums.

[0024] Optionally, for expert data in a database, structured expert data may be directly obtained from the database, and for expert data in Internet resources, data extraction may be performed using a machine learning algorithm (natural language processing technology).

[0025] In the embodiment of the present application, after the expert data is acquired, in order to improve the accuracy and reliability of the expert data, the expert data may be preprocessed to construct a standard expert data set.

[0026] Optionally, data cleaning and data integration operations may be performed on the expert data to obtain standard expert data, and a standard expert data set may be constructed from the standard expert data.

[0027] For example, it is possible to remove duplicate, erroneous, and irrelevant data in expert data, delete duplicate expert data through data comparison algorithms, ensure the uniqueness of data, identify and correct errors in data, such as incorrect educational background data, etc., and eliminate data irrelevant to the mining field to ensure the accuracy of standard expert data. It is possible to unify the data structure of expert data and associate expert data, that is, integrate expert data from different sources according to a unified data structure to form a unified data structure. The data structure includes at least field name, data type, etc., and integrate scattered expert data through data association technology to form complete expert data to obtain standard expert data.

[0028] In an embodiment of the present application, after the standard expert data is acquired, the standard expert data may be aggregated to obtain a standard expert data set in the mining field.

[0029] For example, the standard expert dataset includes at least the expert's basic data (for example, name, gender, age, headshot, and work unit), educational background (for example, education, major, and graduate school), work experience (academic title, position, and main responsibilities), research direction (for example, research direction and expertise in the mining field), research results (for example, published papers, projects participated in, and awards received), and public information (for example, social media speeches, speeches, industry reviews, and records of social activities).

[0030] It should be noted that the standard expert dataset can be updated in real time. For Internet resources, the standard expert dataset can be updated regularly using web crawler technology and natural language processing technology. At the same time, an audit strategy can be established to regularly audit and verify the standard expert dataset to ensure its accuracy and completeness.

[0031] S102, according to the standard expert data set, obtain the basic label, personalized label and associated label of each expert in the standard expert data set, and construct an expert label index according to the basic label, personalized label and associated label.

[0032] In an embodiment of the present application, after obtaining the standard expert data set, the standard expert data set can be mined and analyzed to obtain the basic tags, personalized tags, and associated tags of each expert.

[0033] It should be noted that the present application does not limit the specific method of constructing the expert tag index based on the basic tags, personalized tags and associated tags, and it can be selected according to the actual situation.

[0034] Optionally, the union of basic labels, personalized labels and associated labels can be used as the target labels of experts, and the target labels can be matched with expert data to ensure that each target label can accurately correspond to the corresponding expert. Database technology is used to build a label index for each expert's target label to facilitate fast retrieval and query.

[0035] S103, searching for target expert information required in the target scenario according to the expert tag index to obtain the target expert information.

[0036] It should be noted that the present disclosure does not limit the setting of the target scene, and it can be set according to actual conditions.

[0037] Optionally, the target scenarios include but are not limited to project review, project cooperation, technical training, talent introduction, R&D innovation and other scenarios.

[0038] In the disclosed embodiment, the corresponding target expert tag can be determined according to the actual needs of the target scene, and the target expert tag can be used as a search condition to search in the expert tag index through the target expert tag to obtain the target expert information that meets the search condition.

[0039] S104, input the target expert information into the pre-trained mining industry large model to obtain the expert evaluation label.

[0040] In an embodiment of the present application, sample data in the mining field can be obtained, and the pre-trained base large model can be incrementally trained based on the sample data to obtain the trained initial large model, obtain fine-tuning tasks and fine-tuning training data, and fine-tune the initial large model based on the fine-tuning tasks and fine-tuning training data to obtain a large model for the mining industry.

[0041] Optionally, in order to make the big model better applied in the mining field, a base big model that performs well in the general field can be selected, such as BERT, GPT, etc., and a masked language model (MLM) is used to perform incremental pre-training on sample data in the mining field to obtain an initial big model that has been trained. On the basis of the initial big model, an output layer for a specific fine-tuning task is added, that is, expert evaluation labels are inferred based on expert information. After determining the fine-tuning task, the fine-tuning training data corresponding to the fine-tuning task can be determined, wherein the fine-tuning training data includes expert information and evaluation labels corresponding to the expert information, so as to fine-tune the initial big model according to the fine-tuning task and the fine-tuning training data, and iterate repeatedly until the performance of the fine-tuned big model meets the training end conditions, that is, meets the performance requirements, so as to obtain a big model for the mining industry.

[0042] It should be noted that in order to avoid catastrophic forgetting of general domain knowledge, the fine-tuning training data can include data in the mining field and data in the general field. For example, the ratio of data in the mining field to data in the general field is set to 1:5, and the initial large model is fine-tuned according to the fine-tuning task and fine-tuning training data to obtain the large model of the mining industry.

[0043] In the disclosed embodiment, after the mining industry big model is acquired, the target expert information can be input into the pre-trained mining industry big model, and the expert evaluation label output by the mining industry big model can be obtained.

[0044] For example, if the target scenario is a project review scenario, the expert evaluation labels can be the number of project reviews, the number of times serving as the head of the review expert group, the number of outstanding review experts, the review pass rate, and the review fee, etc.; if the target scenario is a project cooperation scenario, the expert evaluation labels can be the research direction fit, professional ability level, number of cooperation, acceptance results of previous cooperation projects, etc.; if the target scenario is a technical training scenario, the expert evaluation labels can be the training course satisfaction, training effect evaluation, number of trainings, training fields, etc.; if the target scenario is a talent introduction scenario, the expert evaluation labels can be educational background, work experience, scientific research results, industry influence, and contribution to the development of the enterprise after introduction, etc.; if the target scenario is an R&D and innovation scenario, the expert evaluation labels can be the number of patent applications and authorizations, the conversion rate of R&D results, etc.

[0045] S105, constructing an expert portrait based on the target expert information and expert evaluation labels.

[0046] In an embodiment of the present application, the target expert information can be classified into types to obtain the types of the target expert information. The first target expert information corresponding to each type can be obtained from the target expert information. The first target expert information is used as an expert label. Based on the expert label and the expert evaluation label, the expert portrait information is determined. According to the expert portrait information, an expert portrait is constructed.

[0047] Optionally, expert labels and expert evaluation labels can be integrated to obtain the integrated expert's target label, and the expert's portrait information can be obtained based on the expert's target label. The expert portrait can be constructed through visualization technology, where the expert portrait includes expert labels, capability radar charts, cooperation networks, and evaluation labels. Through the expert cooperation network, information such as cooperation experts, cooperation fields, and cooperation results can be directly queried.

[0048] The method for constructing an expert portrait in the mining field proposed in the present application obtains multi-source expert data in the mining field, pre-processes the expert data, constructs a standard expert data set, obtains the basic label, personalized label and associated label of each expert in the standard expert data set according to the standard expert data set, and constructs an expert label index according to the basic label, personalized label and associated label, searches for the target expert information required in the target scenario according to the expert label index, obtains the target expert information, inputs the target expert information into the pre-trained mining industry big model, obtains the expert evaluation label, and constructs the expert portrait according to the target expert information and the expert evaluation label. Therefore, the present application constructs the expert label index by mining and analyzing the standard expert data set, and by obtaining the mining industry big model, the needs of the mining field can be better utilized. Based on the mining industry big model, the expert evaluation label is obtained, and the multi-angle evaluation of the expert is realized, the limitation of a single dimension is avoided, and the objectivity and credibility of the expert evaluation are improved. The expert portrait is constructed according to the target expert information and the expert evaluation label, and the comprehensiveness and accuracy of the expert portrait are ensured. The method is applicable to a wide range of scenarios and has certain practical value.

[0049] In the above embodiment, the specific process of obtaining the basic labels of each expert in the standard expert dataset according to the standard expert dataset can be combined with Figure 2 Further understanding, Figure 2 FIG. 1 is a flow chart of a method for constructing an expert portrait in the mining field according to another embodiment of the present application. Figure 2 As shown, the method includes:

[0050] S201, obtaining the type of the basic label of the preset expert, and obtaining the basic label of each expert from the standard expert data set according to the type of the basic label.

[0051] It should be noted that the present application does not limit the types of basic tags for experts, and can be set according to actual conditions. Optionally, the types of basic tags can include description tags and achievement tags.

[0052] Among them, descriptive tags include but are not limited to the expert's name, gender, age, professional title, work unit, research direction, work experience, research field, profile photo, etc.

[0053] Among them, achievement labels include but are not limited to experts’ published papers, scientific research results, scientific research projects, and achievement transformation.

[0054] In the embodiment of the present application, after the type of the basic label is obtained, the basic label of each expert can be obtained from the standard expert data set according to the type of the basic label.

[0055] S202, extract keywords from the standard expert data set to obtain keywords for each expert.

[0056] Optionally, text-based standard expert data in a standard expert data set may be obtained, and a pre-trained term frequency-inverse document frequency (TF-IDF) model may be used to perform keyword extraction on the text-based standard expert data to obtain keywords for each expert.

[0057] For example, for any document d of an expert, determine the term frequency TF of each word t in document d, determine the inverse document frequency IDF for each word t in document d, multiply the term frequency TF and the inverse document frequency IDF for each word t, and obtain the product TF-IDF value between the term frequency TF and the inverse document frequency IDF of each word t. The TF-IDF values ​​can be sorted in descending order, and the words corresponding to the top N TF-IDF values ​​are selected as keywords, where N is an integer, and this application does not limit the setting of N.

[0058] S203, determining the extraction label of each expert based on the keywords.

[0059] Among them, the extracted labels are obtained by data mining and analysis of standard expert data sets, that is, they are automatically extracted from text-based standard expert data and are a supplement to the basic labels.

[0060] In the embodiment of the present application, after the keywords are obtained, the keywords are used as extraction labels for each expert.

[0061] S204, taking the union of the basic label and the extracted label as the basic label for each expert.

[0062] In an embodiment of the present application, after the basic labels and extracted labels are obtained, in order to prevent repeated labels, the union of the basic labels and the extracted labels can be used as the basic labels for each expert.

[0063] In the above embodiment, the specific process of obtaining the personalized label of each expert in the standard expert dataset according to the standard expert dataset can be combined with Figure 3 Further understanding, Figure 3 FIG. 1 is a flow chart of a method for constructing an expert portrait in the mining field according to another embodiment of the present application. Figure 3 As shown, the method includes:

[0064] S301, performing vector conversion on the keywords of each expert to obtain keyword vectors.

[0065] It should be noted that the specific method for obtaining the keywords of each expert will not be described here in detail, and reference may be made to the above embodiment.

[0066] Optionally, a pre-trained Word2Vec model may be used to convert each expert's keywords into vector representations to obtain keyword vectors.

[0067] S302, clustering and classifying the keyword vectors, determining a target keyword vector from the keyword vectors, and using the target keyword corresponding to the target keyword vector as a personalized label for each expert.

[0068] Optionally, the K-Means algorithm can be used to cluster and classify keyword vectors to identify the correlation and hierarchy between keyword vectors, select the high-frequency keyword vector in each cluster as the target keyword vector, and use the target keyword corresponding to the target keyword vector as the personalized label of each expert.

[0069] Furthermore, after the target keywords are obtained, in order to improve the accuracy of obtaining personalized tags, operations such as verification and optimization can be performed on the target keywords to obtain the final target keywords as the final personalized tags of each expert.

[0070] Among them, the personalized labels of experts are aimed at deeply exploring the personalized characteristics and professional influence of experts. For example: personalized labels include but are not limited to emotional tendency labels, academic influence labels and industry influence labels.

[0071] For example, we can extract the expert's language style and attitude tendencies through his or her social media speeches, public speeches or peer reviews, and construct a sentiment tendency label; construct an academic influence label through the number of citations of the expert's published papers, the impact factor of the journal in which the paper was published, whether he or she is an editorial board member of the academic journal, academic peer reviews, etc.; construct an industry influence label through the expert's social position, social welfare participation, social evaluation, social activity activity, etc.

[0072] In the above embodiment, the specific process of obtaining the associated labels of each expert in the standard expert dataset according to the standard expert dataset can be combined with Figure 4 Further understanding, Figure 4 FIG. 1 is a flow chart of a method for constructing an expert portrait in the mining field according to another embodiment of the present application. Figure 4 As shown, the method includes:

[0073] S401, extracting features from a standard expert data set to obtain correlation feature vectors between experts in the standard expert data set.

[0074] For example, the associated feature vector includes, but is not limited to: expert name, cooperation project, cooperation field, cooperation results, etc.

[0075] S402, performing association rule mining on the association feature vector using the Apriori association rule mining model to obtain a target association rule.

[0076] Optionally, the Apriori association rule mining model can be used to mine association rules for the association feature vectors, where the association rule can be understood as a cooperation model. For any association rule X, the support support(X) can be expressed as:

[0077]

[0078] Where X is an item set, which can be a single expert or a combination of multiple experts. Generate association rules for each frequent item set and calculate the confidence of each association rule. The confidence (X→Y) can be expressed as:

[0079]

[0080] Among them, X and Y are two item sets.

[0081] Optionally, target association rules that meet the conditions are screened out through minimum support (min_support) and minimum confidence (min_confidence) thresholds to generate frequent item sets, that is, to obtain a set of all possible cooperation modes.

[0082] S403, obtaining the associated label of each expert according to the target association rule.

[0083] In the embodiment of the present application, after the target association rule is obtained, the association label of each expert can be obtained according to the target association rule.

[0084] The expert portrait construction method in the mining field proposed in the present application obtains the basic labels, personalized labels and associated labels of each expert in the standard expert data set, and constructs an expert label index based on the basic labels, personalized labels and associated labels. The target expert information required in the target scenario is searched according to the expert label index to obtain the target expert information. Therefore, the present invention uses a variety of technologies such as machine learning and natural language processing to efficiently and accurately obtain expert labels from multi-source expert data, and classifies and indexes expert information by constructing an expert label index system to achieve efficient management and rapid retrieval of expert information. The required expert information can be easily found through labels, thereby improving work efficiency.

[0085] Figure 5 FIG. 1 is a schematic diagram of a structure of a device for constructing an expert portrait in the mining field according to an embodiment of the present application. Figure 5 As shown, the expert portrait construction device 500 in the mining field includes a first acquisition module 51, a second acquisition module 52, a search module 53, a third acquisition module 54 and a portrait construction module 55, wherein:

[0086] A first acquisition module 51 is used to acquire multi-source expert data in the mining field, pre-process the expert data, and construct a standard expert data set;

[0087] A second acquisition module 52 is used to acquire the basic label, personalized label and associated label of each expert in the standard expert data set according to the standard expert data set, and construct an expert label index according to the basic label, the personalized label and the associated label;

[0088] A search module 53, configured to search for target expert information required in a target scenario according to the expert tag index to obtain the target expert information;

[0089] The third acquisition module 54 is used to input the target expert information into the pre-trained mining industry large model to obtain the expert evaluation label;

[0090] The portrait construction module 55 is used to construct an expert portrait according to the target expert information and the expert evaluation label.

[0091] The expert portrait construction device in the mining field provided in the second aspect of the present application also has the following technical features, including:

[0092] According to one embodiment of the present application, the second acquisition module 52 is used to: obtain the type of basic labels of preset experts, and obtain the basic labels of each expert from the standard expert data set according to the type of basic labels; perform keyword extraction on the standard expert data set to obtain keywords of each expert; determine the extracted labels of each expert based on the keywords; and use the union of the basic labels and the extracted labels as the basic labels of each expert.

[0093] According to one embodiment of the present application, the second acquisition module 52 is used to: perform vector conversion on the keywords of each expert to obtain a keyword vector; cluster and classify the keyword vectors, determine a target keyword vector from the keyword vectors, and use the target keyword corresponding to the target keyword vector as a personalized label for each expert.

[0094] According to one embodiment of the present application, the second acquisition module 52 is used to: obtain text-type standard expert data in the standard expert data set, and perform keyword extraction on the text-type standard expert data using a pre-trained word frequency inverse document frequency model to obtain keywords for each expert.

[0095] According to one embodiment of the present application, the second acquisition module 52 is used to: perform feature extraction on the standard expert data set to obtain the association feature vectors between experts in the standard expert data set; perform association rule mining on the association feature vectors by the Apriori association rule mining model to obtain target association rules; and obtain the association labels of each expert according to the target association rules.

[0096] According to one embodiment of the present application, the training process of the mining industry model includes: obtaining sample data in the mining field, and performing incremental training on a pre-trained base large model based on the sample data to obtain a trained initial large model; obtaining fine-tuning tasks and fine-tuning training data, and fine-tuning the initial large model based on the fine-tuning tasks and the fine-tuning training data to obtain the mining industry large model.

[0097] According to one embodiment of the present application, the portrait construction module 55 is used to: classify the target expert information into types to obtain the types of the target expert information; obtain the first target expert information corresponding to each type from the target expert information, and use the first target expert information as an expert label; determine the portrait information of the expert based on the expert label and the expert evaluation label; and construct the expert portrait based on the expert portrait information.

[0098] The expert portrait construction device proposed in the present application obtains multi-source expert data in the mining field, pre-processes the expert data, constructs a standard expert data set, obtains the basic labels, personalized labels and associated labels of each expert in the standard expert data set according to the standard expert data set, and constructs an expert label index according to the basic labels, personalized labels and associated labels, searches for the target expert information required in the target scenario according to the expert label index, obtains the target expert information, inputs the target expert information into the pre-trained mining industry big model, obtains the expert evaluation labels, and constructs the expert portrait according to the target expert information and the expert evaluation labels. Therefore, the present application mines and analyzes the standard expert data set to construct an expert label index, obtains the expert evaluation labels based on the mining industry big model, and constructs the expert portrait according to the target expert information and the expert evaluation labels, thereby ensuring the comprehensiveness and accuracy of the expert portrait.

[0099] To achieve the above embodiments, the present application also provides an electronic device, a computer-readable storage medium and a computer program product.

[0100] Figure 6 A block diagram of an electronic device according to an embodiment of the present application, such as Figure 6 As shown, the device 1000 includes a memory 101, a processor 102, and a computer program stored in the memory 101 and executable on the processor 102. When the processor 102 executes the program instructions, the execution is realized. Figures 1 to 5 Embodiment of the invention is a method for constructing an expert portrait in the mining field.

[0101] In order to implement the above embodiments, the present application also provides a non-transitory computer-readable storage medium storing computer instructions, the computer instructions are used to enable the computer to execute Figures 1 to 4 A method for constructing an expert portrait in the mining field of an embodiment.

[0102] In order to implement the above embodiment, the present application also provides a computer program product. When the instruction processor in the computer program product executes, Figures 1 to 4 A method for constructing an expert portrait in the mining field of an embodiment.

[0103] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0104] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0105] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0106] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0107] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0108] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0109] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0110] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for constructing an expert portrait in the mining field, characterized in that: The method comprises: Acquire multi-source expert data in the mining field, pre-process the expert data, and construct a standard expert data set; According to the standard expert data set, basic tags, personalized tags and associated tags of each expert in the standard expert data set are obtained, and an expert tag index is constructed according to the basic tags, the personalized tags and the associated tags; Searching for target expert information required in a target scenario according to the expert tag index to obtain the target expert information; Input the target expert information into the pre-trained mining industry large model to obtain expert evaluation labels; An expert portrait is constructed based on the target expert information and the expert evaluation label.

2. The method according to claim 1, characterized in that The step of obtaining the basic label of each expert in the standard expert dataset according to the standard expert dataset includes: Obtaining the type of the preset basic label of the expert, and obtaining the basic label of each expert from the standard expert data set according to the type of the basic label; Extracting keywords from the standard expert data set to obtain keywords for each expert; Determine an extraction label for each expert according to the keywords; The union of the basic label and the extracted label is used as the basic label for each expert.

3. The method according to claim 2, characterized in that The step of obtaining, according to the standard expert dataset, a personalized label for each expert in the standard expert dataset comprises: Performing vector conversion on the keywords of each expert to obtain a keyword vector; The keyword vectors are clustered and classified, a target keyword vector is determined from the keyword vectors, and a target keyword corresponding to the target keyword vector is used as a personalized label for each expert.

4. The method according to claim 2, characterized in that: The step of extracting keywords from the standard expert data set to obtain keywords of each expert includes: The text-based standard expert data in the standard expert data set is obtained, and keywords are extracted from the text-based standard expert data using a pre-trained word frequency inverse document frequency model to obtain keywords for each expert.

5. The method according to claim 1, characterized in that The step of obtaining, according to the standard expert dataset, an associated label of each expert in the standard expert dataset comprises: Performing feature extraction on the standard expert data set to obtain correlation feature vectors between experts in the standard expert data set; Performing association rule mining on the association feature vector using an Apriori association rule mining model to obtain a target association rule; According to the target association rule, the association label of each expert is obtained.

6. The method according to claim 1, characterized in that The training process of the mining industry model includes: Acquire sample data in the mining field, and perform incremental training on the pre-trained base large model according to the sample data to obtain a trained initial large model; A fine-tuning task and fine-tuning training data are obtained, and the initial large model is fine-tuned according to the fine-tuning task and the fine-tuning training data to obtain the mining industry large model.

7. The method according to claim 1, characterized in that The step of constructing an expert portrait according to the target expert information and the expert evaluation label includes: Classify the target expert information into types to obtain the types of the target expert information; From the target expert information, obtain first target expert information corresponding to each type, and use the first target expert information as an expert label; Determining the expert's portrait information based on the expert label and the expert evaluation label; The expert portrait is constructed according to the expert portrait information.

8. A device for constructing expert portraits in the mining field, characterized in that: The device comprises: A first acquisition module is used to acquire multi-source expert data in the mining field, pre-process the expert data, and construct a standard expert data set; A second acquisition module is used to acquire the basic label, personalized label and associated label of each expert in the standard expert data set according to the standard expert data set, and construct an expert label index according to the basic label, the personalized label and the associated label; A search module, used to search for target expert information required in a target scenario according to the expert tag index to obtain the target expert information; The third acquisition module is used to input the target expert information into the pre-trained mining industry large model to obtain the expert evaluation label; The portrait construction module is used to construct an expert portrait based on the target expert information and the expert evaluation label.

9. An electronic device, characterized in that: including a processor and a memory; The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images