Intelligent management method and system for promoting business processes

Abstract

The application discloses an intelligent management method and system for promoting a business process, and the method comprises the following steps: acquiring first project data related to a first promotion project, acquiring original business information corresponding to a second business end, wherein the first project data comprises first project text information and first project annotation information; performing semantic recognition based on the first project text information to obtain first project semantic features; determining first project annotation features based on the first project annotation information; generating second project data based on the first project semantic features, the first project annotation features and the original business information, wherein a first similarity between the first project annotation information and second project annotation information in the second project data satisfies a first preset similarity condition; determining promotion business process information based on a promotion business knowledge graph and the second project data; and pushing the promotion business process information to the second business end, so that intelligent management of the promotion business process can be automatically completed, and management efficiency and accuracy can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of business management, and particularly relates to an intelligent management method and system for promoting a business process. BACKGROUND

[0002] In the actual application of business promotion, a business promotion project usually needs multiple business personnel to collaboratively process so as to improve the process efficiency of the promoted business. For example, after a business personnel negotiates a related promoted business, the business personnel usually needs to send the related data of the business to a manager, so that the manager sets corresponding process nodes / process information according to the related data of the business, and allocates the corresponding process nodes / process information to the related business personnel, thereby completing the collaborative processing.

[0003] However, the management mode of relying on the manager to set and allocate the process nodes / process information is not only inefficient, but also the management accuracy is restricted by the experience and concentration of the manager. SUMMARY

[0004] In order to solve the above technical problems, the embodiments of the present application provide an intelligent management method and system for promoting a business process, which can automatically complete the intelligent management of the promoted business process, and improve the management efficiency and accuracy.

[0005] In a first aspect, the embodiments of the present application provide an intelligent management method for promoting a business process, comprising:

[0006] obtaining first project data related to a first promotion project of a first business end, and obtaining original business information corresponding to a second business end, wherein the first project data comprises first project text information and first project annotation information of the first promotion project;

[0007] performing semantic recognition based on the first project text information to obtain first project semantic features;

[0008] determining first project annotation features based on the first project annotation information;

[0009] generating second project data based on the first project semantic features, the first project annotation features and the original business information, wherein a first similarity between the first project annotation information and second project annotation information in the second project data satisfies a first preset similarity condition;

[0010] determining promotion business process information corresponding to the second business end based on a preset promotion business knowledge graph and the second project data;

[0011] pushing the promotion business process information to the second business end.

[0012] Optionally, the first promotion item includes at least one item object, and the first item annotation information includes respective first object attributes and first object requirements of the at least one item object.

[0013] The first item annotation feature is determined based on the first item annotation information, including:

[0014] The first object attribute of each item object is subjected to feature extraction to obtain a first attribute feature of each item object.

[0015] The first object requirement of each item object is subjected to feature extraction to obtain a first requirement feature of each item object.

[0016] The first item annotation feature is determined based on the respective first attribute features and first requirement features of the at least one item object.

[0017] Optionally, the first item annotation feature is determined based on the respective first attribute features and first requirement features of the at least one item object, including:

[0018] Based on the promotion business knowledge graph, a first item business knowledge corresponding to the first item data is acquired, wherein the first item business knowledge includes typical object attribute knowledge and typical object requirement knowledge.

[0019] Features of the typical object attribute knowledge and features of the typical object requirement knowledge are respectively determined.

[0020] For each item object, a second attribute feature of the item object is obtained by performing feature interaction processing on the first attribute feature of the item object and the features of the typical object attribute knowledge, and a second requirement feature of the item object is obtained by performing feature interaction processing on the first requirement feature of the item object and the features of the typical object requirement knowledge, wherein the feature interaction processing includes cross-attention interaction.

[0021] The first item annotation feature is determined based on the respective second attribute features and second requirement features of the at least one item object.

[0022] Optionally, the second item data is generated based on the first item semantic feature, the first item annotation feature, and the original business information, including:

[0023] A plurality of business process node frameworks corresponding to the original business information are determined.

[0024] generate content respectively for the plurality of business process node frameworks based on the first project semantic features and the first project annotation features, and fill the generated content into the corresponding business process node frameworks respectively to obtain project data corresponding to each of the plurality of business process node frameworks;

[0025] determine the second project data based on the first project annotation features and the project data corresponding to each of the plurality of business process node frameworks.

[0026] Optionally, the generating content respectively for the plurality of business process node frameworks based on the first project semantic features and the first project annotation features, and filling the generated content into the corresponding business process node frameworks respectively, comprises:

[0027] for each of the plurality of business process node frameworks,

[0028] based on the first project semantic features, locate one or more semantic association regions in the business process node framework;

[0029] generate relevant content of each of the one or more semantic association regions based on the first project annotation features and the first project semantic features;

[0030] determine the generated content corresponding to the business process node framework based on the relevant content of each of the one or more semantic association regions;

[0031] fill the generated content corresponding to the business process node framework into the business process node framework.

[0032] Optionally, a second similarity between the project data corresponding to each of the business process node frameworks and the information in the first project annotation information matching the business process node framework satisfies a second preset similarity condition.

[0033] Optionally, the determining the second project data based on the first project annotation features and the project data corresponding to each of the plurality of business process node frameworks, comprises:

[0034] for the project data corresponding to each of the business process node frameworks, determine the generated content corresponding thereto, and evaluate the project data based on the first project annotation features and the generated content corresponding to the project data to obtain a corresponding evaluation result;

[0035] based on the evaluation results, determine at least part of the project data from the project data corresponding to each of the plurality of business process node frameworks;

[0036] determine the second project data based on the at least part of the project data.

[0037] Optionally, the method further comprises:

[0038] acquiring second project business knowledge corresponding to the second project data based on the promotion business knowledge graph.

[0039] determining the promotion business process information based on the second project business knowledge and the second project data.

[0040] Optionally, the method further comprises:

[0041] determining a first dynamic coefficient corresponding to the second project business knowledge by using strategy optimization based on a degree of association between the second project business knowledge and the second project data, and / or a degree of matching between the second project business knowledge and general knowledge, wherein the general knowledge is determined by the second project data, and the first dynamic coefficient comprises at least one of a first dynamic weight and a first dynamic threshold.

[0042] adjusting the second project data based on at least the second project business knowledge and the first dynamic coefficient corresponding thereto.

[0043] splitting the adjusted second project data into the promotion business process information corresponding to the second business end.

[0044] In a second aspect, an embodiment of the present application provides an intelligent management system for a promotion business process, comprising:

[0045] a data acquisition module configured to acquire first project data related to a first promotion project of a first business end and acquire original business information corresponding to a second business end, wherein the first project data comprises first project text information and first project annotation information of the first promotion project.

[0046] a semantic feature determination module configured to perform semantic recognition based on the first project text information to obtain first project semantic features.

[0047] an annotation feature determination module configured to determine first project annotation features based on the first project annotation information.

[0048] a second item data generation module configured to generate second item data based on the first item semantic feature, the first item annotation feature, and the original business information, wherein a first similarity between the first item annotation information and second item annotation information in the second item data satisfies a first preset similarity condition;

[0049] a business process determination module configured to determine promotion business process information corresponding to the second business end based on a preset promotion business knowledge graph and the second item data;

[0050] an information pushing module configured to push the promotion business process information to the second business end.

[0051] In summary, the embodiments of the present application have at least the following beneficial effects:

[0052] By using the embodiments of the present application, first item data related to a first promotion project of a first business end is obtained, and original business information corresponding to a second business end is obtained, wherein the first item data includes first item text information and first item annotation information of the first promotion project. Semantic recognition is performed based on the first item text information to obtain a first item semantic feature. A first item annotation feature is determined based on the first item annotation information. Second item data is generated based on the first item semantic feature, the first item annotation feature, and the original business information, wherein a first similarity between the first item annotation information and second item annotation information in the second item data satisfies a first preset similarity condition. Promotion business process information corresponding to the second business end is determined based on a preset promotion business knowledge graph and the second item data. The promotion business process information is pushed to the second business end, so that intelligent management of the promotion business process can be automatically completed, and management efficiency and accuracy are improved. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 is a flowchart of an intelligent management method for a promotion business process provided by the embodiments of the present application;

[0054] Figure 2 is an interaction diagram of each end for intelligent management of a promotion business process provided by the embodiments of the present application;

[0055] Figure 3 is a structural diagram of an intelligent management system for a promotion business process provided by the embodiments of the present application;

[0056] Figure 4 is a schematic diagram of a computer device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0057] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments of the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative effort are within the scope of the present application.

[0058] In the description of the present application, the terms "first", "second", "third", etc. are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second", "third", etc. can be explicitly or implicitly included one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise specified. In the description of the present application, the term "comprising" and its variants are open-ended, i.e. "including but not limited to". The term "based on" is "at least partially based on". The term "according to" is "at least partially according to". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments".

[0059] In the description of the present application, it should be noted that, unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connection" should be understood in a broad sense, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication inside two elements. For a person of ordinary skill in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0060] In the description of the present application, it should be noted that, unless otherwise defined, all technical and scientific terms used in the present application have the same meaning as understood by a person skilled in the art. The terms used in the specification of the present application are only for the purpose of describing the specific embodiments, and are not intended to limit the present application. For a person of ordinary skill in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0061] In the first aspect, see Figure 1 , a flowchart of a method for promoting business process intelligent management provided by an embodiment of the present application is shown, the method comprises S101-S106, and the details are as follows.

[0062] S101, obtain first project data related to a first promotion project of a first business end, and obtain original business information corresponding to a second business end, wherein the first project data includes first project text information and first project annotation information of the first promotion project.

[0063] In some examples, the first promotion project described above can include at least one of the following: financial investment promotion business, software operation service promotion business, marketing promotion business. Taking the marketing promotion business as an example, the goal to be achieved is usually conversion rate improvement, brand exposure, and / or user acquisition, etc.

[0064] In some examples, the first project text information described above can carry plan text content for indicating the first promotion project. For example, the plan text content can be a structured descriptive document formed in the project establishment, strategy formulation, scheme report, etc. stages corresponding to the first promotion project. Specifically, the first project text information described above includes at least one of the following corresponding to the first promotion project: marketing activity planning book, textual description in the promotion scheme slide, target and strategy paragraph in the project establishment application, execution summary in the weekly report / review report.

[0065] In some examples, the first project annotation information described above can be used to indicate annotation / comment content input by the first business end for the first promotion project. The annotation / comment content can be a user's note on the key points of the first promotion project. Specifically, the first project annotation information described above can include at least one of the following: the goal to be achieved by the first promotion project, the necessary and unchangeable data of the first promotion project, the attribute of the first promotion project, the type of the first promotion project, the time of the first promotion project, etc.

[0066] S102, performing semantic recognition based on the first project text information to obtain first project semantic features.

[0067] In this embodiment, the relevant semantics of the first project text information (characterized by the first project semantic features) can be automatically and accurately recognized by semantic recognition, thereby solving the problem of low text review efficiency and / or easy to overlook in the related art by requiring managers to manually review relevant project text information.

[0068] In some examples, the first project text information can be subjected to semantic recognition by an artificial intelligence model and / or natural language processing technology and / or large language model, etc., to obtain the first project semantic features. The artificial intelligence model described above can be obtained through reinforcement learning training, and / or the large language model described above can be obtained through reinforcement learning fine-tuning.

[0069] S103, determine a first project annotation feature based on the first project annotation information.

[0070] In some examples, the first project annotation feature can be directly extracted from the first project annotation information.

[0071] In some examples, the feature extraction required in any one or more embodiments of the present application can be implemented through coding processing or through a pre-installed feature extraction model. Here, a corresponding feature extraction model can be selected according to the information modalities of the information to be extracted, for example, if the information modalities of the information to be extracted are also text modalities, the above-mentioned artificial intelligence model and / or large language model can be used as the selected feature extraction model. Of course, there can be other modalities of information, which are not specifically described in this embodiment.

[0072] S104, generate second project data based on the first project semantic feature, the first project annotation feature, and the original business information, wherein a first similarity between the first project annotation information and second project annotation information in the second project data satisfies a first preset similarity condition.

[0073] In this embodiment, by controlling the first similarity to satisfy the first preset similarity condition, it can be ensured that the generated second project data matches the original business information while the key point deviation between the generated second project data and the first project data is precisely limited (the key point is determined by the project annotation information of each other).

[0074] In some examples, the first preset similarity condition can include that the first similarity between the first project annotation information and the second project annotation information is greater than a first preset similarity threshold.

[0075] In some examples, the first preset similarity condition can include that the first similarity between the first project annotation information and the second project annotation information is greater than a first preset similarity threshold and less than a first maximum similarity threshold.

[0076] In some examples, the first preset similarity condition can include that the first similarity between the first project annotation information and the second project annotation information represents that the first project annotation information is the same as the second project annotation information.

[0077] In some examples, the second project data can be determined according to the first project semantic features, the first project annotation features and the original business information by a pre-trained project data generation model. The project data generation model can be a model trained to have a prediction capability of taking the first project semantic features, the first project annotation features and the original business information as model inputs and taking the second project data as model outputs. In the specific training, sample project semantic features, sample project annotation features and sample business information can be used as sample data (which also carries an expected corresponding project data label, which represents the corresponding project data), and a general training algorithm (such as gradient descent method) can be used to train the model, so that the trained model can have the above-mentioned capability.

[0078] In some examples, the first similarity between the first project annotation information and the second project annotation information can be calculated by: extracting features of the second project annotation information to obtain second project annotation features, and calculating a feature similarity between the second project annotation features and the first project annotation features to represent the first similarity.

[0079] In some examples, for the specific case of the second project data and the second project annotation information, reference can be made to the relevant description of the related embodiments of the first project data and the first project annotation information, which will not be repeated here.

[0080] S105, based on the preset promotion business knowledge graph and the second project data, determine the promotion business process information corresponding to the second business end.

[0081] In some examples, step S105 can include: based on the promotion business knowledge graph, obtaining second project business knowledge corresponding to the second project data, and adjusting the second project data based on the second project business knowledge and then splitting to obtain the promotion business process information corresponding to the second business end.

[0082] In some examples, the entity and the relationship related to the second project data can be directly located from the promotion business knowledge graph, and the entity and the relationship can be extracted to form corresponding second project business knowledge. The second project business knowledge can include at least one of the following: project features, historical promotion strategies, target groups, channel preferences, performance indicators, etc. The location of the entity and the relationship can be achieved by calculating the similarity.

[0083] In some examples, the promotion business knowledge graph can be pre-constructed, and can include entities and relationships corresponding to different project data.

[0084] In some examples, the extracted second project business knowledge can be directly utilized to calibrate and / or enhance the original second project data, for example, the calibration and / or enhancement can include at least one of the following: adjusting the budget allocation of the project according to historical experience, optimizing the target audience label of the promotion, correcting the priority of the promotion channel, etc.

[0085] In some examples, based on the adjusted second project data, in combination with the characteristics and capabilities of each second business end (such as channel attributes, execution rhythm, resource limitations, etc.), the overall promotion task represented by the adjusted second project data can be decomposed into a series of executable sub-processes or links, forming customized promotion business process information corresponding to each second business end. Among them, the final promotion business process information can include a clear structure process description text, which can include at least one of the following: stage tasks, responsible persons, time nodes, key actions, expected results, etc., so as to facilitate direct execution or integration by each second business end.

[0086] S106, pushing the promotion business process information to the second business end.

[0087] In some examples, each promotion business process information can be pushed into the corresponding second business end, so as to instruct the second business end to complete the multi-business-end collaborative processing related to the first promotion project according to the promotion business process information it receives. For example, the promotion business process information can be directly used to instruct the corresponding second business end to display the promotion business process information on the preset display area of the self-display module, so as to improve the pushing efficiency / display efficiency and pushing accuracy of the promotion business process information, and thus facilitate the user using the second business end to more efficiently execute the business promotion process node related to the promotion business process information.

[0088] In some examples, the above-mentioned first business end can be included in the above-mentioned second business end, in other words, the first business end itself can also participate in the multi-business-end collaborative processing related to the first promotion project.

[0089] In some examples, referring to Figure 2 The intelligent management method and / or intelligent management system described in any one embodiment of the present application can be applied to the management server / cloud 203, which is in communication connection with the first business end 201 and the second business end 202 respectively.

[0090] In an optional implementation, the first promotion project includes at least one project object, and the first project annotation information includes the first object attribute and the first object demand of each of the at least one project object.

[0091] The first item annotation feature is determined based on the first item annotation information, and the determining the first item annotation feature based on the first item annotation information comprises:

[0092] The first object attribute of each item object is subjected to feature extraction to obtain the first attribute feature of each item object;

[0093] The first object requirement of each item object is subjected to feature extraction to obtain the first requirement feature of each item object;

[0094] The first item annotation feature is determined based on the first attribute feature and the first requirement feature of each of the at least one item object.

[0095] In some examples, the determining the first item annotation feature based on the first attribute feature and the first requirement feature of each of the at least one item object can comprise:

[0096] The first attribute feature and the first requirement feature of each item object are subjected to feature fusion and / or feature interaction processing to obtain the first item annotation feature of each item object;

[0097] Alternatively,

[0098] The first attribute feature and the first requirement feature of each of the at least one item object are subjected to feature fusion to obtain the first item annotation feature.

[0099] In some examples, the item object can be used to indicate a basic element or component of a promotion / marketing activity corresponding to the first promotion item, the first object attribute can be used to indicate inherent features or description information of the corresponding item object itself, and the first object requirement can be used to indicate specific requirements or expectations for the corresponding item object in the promotion item.

[0100] Taking a marketing promotion business as an example, common item objects can include at least one of the following: a target customer group (such as divided according to age, occupation, social identity, gender, and / or region), a promotion channel (such as social media advertising, search engine marketing, and / or KOL (Key Opinion Leader) cooperation, etc.), promotion content (such as promotional videos, graphic materials, and / or coupons, etc.), a marketing tool, a time node (such as a preheating period, a burst period, and / or a tailing period).

[0101] Taking a marketing promotion business as an example, the first object attribute can include at least one of the following: an attribute of a target customer group (such as age, gender, region, and / or consumption habit), an attribute of a promotion channel (such as a covered user amount, a click rate, and / or a delivery cost), an attribute of promotion content (such as a content type (video / graphic), a time length, and / or a theme style), and an attribute of a time node (such as a start time, a duration, and / or a key milestone).

[0102] Taking a marketing promotion business as an example, the first object requirement can include at least one of the following: a requirement of a target customer group, a requirement of a promotion channel (e.g., an ROI (Return On Investment) is not lower than a corresponding preset threshold, such as 200%, and a CTR (Click-Through-Rate) is higher than a corresponding preset threshold, such as an industry average level), a requirement of promotion content, and a requirement of a time node (e.g., completing a preheating launch before a certain key time node).

[0103] In an optional implementation, the determining of the first project annotation feature based on the respective first attribute feature and the first requirement feature of the at least one project object includes:

[0104] obtaining first project business knowledge corresponding to the first project data based on the promotion business knowledge graph, wherein the first project business knowledge includes typical object attribute knowledge and typical object requirement knowledge;

[0105] respectively determining a feature of the typical object attribute knowledge and a feature of the typical object requirement knowledge;

[0106] For each project object, performing feature interaction processing on the first attribute feature of the project object and the feature of the typical object attribute knowledge to obtain a second attribute feature of the project object, and performing feature interaction processing on the first requirement feature of the project object and the feature of the typical object requirement knowledge to obtain a second requirement feature of the project object, wherein the feature interaction processing includes cross-attention interaction.

[0107] determining the first project annotation feature based on the respective second attribute feature and the second requirement feature of the at least one project object.

[0108] In this embodiment, each first attribute feature and the feature of the typical object attribute knowledge can generate a new second attribute feature through feature interaction, so as to enrich the information amount contained in the second attribute feature, and make the information amount contained match the typical object related to the first promotion project, thereby improving the information amount of the first project annotation feature while preventing it from deviating too much from the typical object related to the first promotion project.

[0109] In some examples, an entity and a relationship related to the first project data can be located directly from the promotion business knowledge graph, and the entity and the relationship can be extracted to form the corresponding first project business knowledge. The location of the entity and the relationship can be achieved by calculating a similarity.

[0110] In some examples, the typical object attribute knowledge can be used to indicate a typical object attribute related to the first promotion item, for example, the typical object attribute can be determined according to the item type of the first promotion item, and exemplary meanings of the typical object attribute can refer to the above description of the first object attribute, except that the specific attribute values are adjusted according to the specific circumstances of the first promotion item, which will not be repeated here.

[0111] In some examples, the typical object requirement knowledge can be used to indicate a typical object requirement related to the first promotion item, for example, the typical object requirement can be determined according to the item type of the first promotion item, and exemplary meanings of the typical object requirement can refer to the above description of the first object requirement, except that the specific requirement values are adjusted according to the specific circumstances of the first promotion item, which will not be repeated here.

[0112] In an optional implementation, the generating the second item data based on the first item semantic feature, the first item annotation feature and the original business information comprises:

[0113] determining a plurality of business process node frameworks corresponding to the original business information;

[0114] generating content for the plurality of business process node frameworks respectively based on the first item semantic feature and the first item annotation feature, and filling the generated content into the corresponding business process node frameworks respectively to obtain the item data corresponding to the plurality of business process node frameworks respectively;

[0115] determining the second item data based on the first item annotation feature and the item data corresponding to the plurality of business process node frameworks respectively.

[0116] In some examples, the above business process node framework can be used to indicate a business process node template, that is, the business process node framework can be represented as a framework / template of some to-be-filled text content.

[0117] In some examples, at least part of the plurality of business process node frameworks can be pre-configured for the original business information of different second business ends.

[0118] In some examples, at least part of the plurality of business process node frameworks can be obtained by non-directional revision of the pre-configured business process node frameworks based on the original business information, so as to enrich the number and / or format type of the business process node frameworks.

[0119] In some examples, the content generation model can be pre-trained to generate content for each of the plurality of business process node frameworks based on the first project semantic features and the first project annotation features. The content generation model can be trained to have the ability to take the first project semantic features and the first project annotation features as input and generate the content as output. In the training process, sample project semantic features and sample project annotation features can be used as sample data, which also carries the expected corresponding content label representing the expected content. A general training algorithm, such as gradient descent, can be used to train the model, so that the trained model can have the above-mentioned ability.

[0120] In some examples, the project data determination model can be pre-trained to determine the second project data based on the first project annotation features and the project data corresponding to each of the plurality of business process node frameworks. The project data determination model can be trained to have the ability to take the first project annotation features and the project data corresponding to each of the plurality of business process node frameworks as input and generate the second project data as output. In the training process, sample project annotation features and the project data corresponding to each of the plurality of business process node frameworks can be used as sample data, which also carries the expected corresponding project data label representing the expected project data. A general training algorithm, such as gradient descent, can be used to train the model, so that the trained model can have the above-mentioned ability.

[0121] In an optional implementation, the content generation based on the first project semantic features and the first project annotation features for each of the plurality of business process node frameworks, and filling the generated content into the corresponding business process node framework, comprises:

[0122] For each of the plurality of business process node frameworks,

[0123] Based on the first project semantic features, one or more semantic association regions are located in the business process node framework;

[0124] Based on the first project annotation features and the first project semantic features, relevant content for each of the one or more semantic association regions is generated;

[0125] Based on the relevant content for each of the one or more semantic association regions, the generated content corresponding to the business process node framework is determined;

[0126] The generated content corresponding to the business process node framework is filled into the business process node framework.

[0127] In some examples, the first item annotation feature and the first item semantic feature can be directly input into a large language model for content generation to generate the relevant content for each of the one or more semantic association regions.

[0128] In some examples, the relevant content can be determined based on the first item annotation feature and the first item semantic feature by a pre-trained relevant content generation model. The relevant content generation model can be trained to have the ability to take the first item annotation feature and the first item semantic feature as input and generate the relevant content as output. During training, sample item annotation features and sample item semantic features can be used as sample data, which also carries the expected corresponding relevant content label representing the expected generated relevant content. A general training algorithm such as gradient descent can be used to train the model so that the trained model has the above-mentioned ability.

[0129] In an optional implementation, a second similarity between the project data corresponding to each of the business process node frameworks and the information in the first item annotation information matching the business process node framework satisfies a second preset similarity condition.

[0130] In some examples, the second preset similarity condition can include that the second similarity between the two is greater than a second preset similarity threshold. Optionally, the second preset similarity threshold can be less than or equal to the first preset similarity threshold, or can be irrelevant to the first preset similarity threshold.

[0131] In some examples, the second preset similarity condition can include that the second similarity between the two is greater than a second preset similarity threshold and less than a second maximum similarity threshold.

[0132] In an optional implementation, the determining of the second project data based on the first item annotation feature and the project data corresponding to each of the plurality of business process node frameworks includes:

[0133] For the project data corresponding to each of the business process node frameworks, the generated content corresponding thereto is determined therefrom, and the project data is evaluated based on the first item annotation feature and the generated content corresponding to the project data to obtain a corresponding evaluation result.

[0134] Based on the evaluation results, at least part of the project data corresponding to each of the plurality of business process node frameworks is determined.

[0135] The second project data is determined based on the at least part of the project data.

[0136] In the embodiment, since each business process node framework is an initial framework corresponding to original business information, not all data in the project data corresponding to each business process node framework is necessary to be added to the second project data, and therefore, the project data corresponding to each business process node framework can be further evaluated respectively, so as to screen the project data corresponding to each business process node framework through the evaluation results, and select at least part of the project data meeting the requirements to determine the second project data.

[0137] In some examples, the at least part of the project data can be directly composed into the second project data.

[0138] In some examples, the evaluation result for indicating whether the generated content and the first project annotation feature match and / or are similar and / or relevant to each other can be determined by calculating the matching degree and / or similarity and / or relevance between the generated content and the first project annotation feature corresponding to the project data.

[0139] In some examples, the at least part of the project data can be the project data with the evaluation result data value higher than a preset evaluation threshold, and / or can also be at least one project data with the largest evaluation result data value.

[0140] In some examples, the generated content corresponding to each project data can be determined from each project data.

[0141] In an optional implementation, the determining of the promotion business process information corresponding to the second business end based on the preset promotion business knowledge graph and the second project data comprises:

[0142] obtaining second project business knowledge corresponding to the second project data based on the promotion business knowledge graph;

[0143] determining the promotion business process information based on the second project business knowledge and the second project data.

[0144] In some examples, the second project data can be directly adjusted according to the second project business knowledge, and the adjusted second project data can be split into the promotion business process information corresponding to the second business end.

[0145] In an optional implementation, the determining of the promotion business process information based on the second project business knowledge and the second project data comprises:

[0146] determine the first dynamic coefficient corresponding to the second project business knowledge by strategy optimization based on the degree of association between the second project business knowledge and the second project data, and / or the degree of matching between the second project business knowledge and general knowledge, wherein the general knowledge is determined by the second project data, and the first dynamic coefficient includes at least one of the following: a first dynamic weight, a first dynamic threshold value;

[0147] adjust the second project data based on at least the second project business knowledge and the first dynamic coefficient corresponding thereto;

[0148] split the adjusted second project data into promotion business process information corresponding to the second business end.

[0149] In some examples, the above-mentioned strategy optimization can be implemented by a strategy optimization model constructed based on a rule engine, a learning model or a hybrid system, which can generate the final dynamic coefficient through the degree of association and / or the degree of matching. In some specific examples, the degree of association and / or the degree of matching can be input into the above-mentioned strategy optimization model to obtain the first dynamic coefficient output by the strategy optimization model. Generally speaking, the strategy optimization model can be implemented based on a reinforcement learning algorithm, for example, first define the three elements of reinforcement learning, take an agent as the strategy optimization module, which is responsible for deciding how much dynamic coefficient to allocate to the knowledge, take the promotion project system as the environment, which can be used to input project data, knowledge, and output execution results, and take the final effect of the project (such as GMV improvement, new customer number, ROI) as the reward. Then, the agent can be configured with a corresponding reinforcement learning algorithm, such as DQN, PPO, A2C, etc., so that the agent uses the reinforcement learning algorithm to continuously optimize the strategy, thereby completing the training of the strategy optimization model.

[0150] In some examples, the method can further include determining a second dynamic coefficient corresponding to the general knowledge by strategy optimization based on the degree of association between the second project business knowledge and the second project data, and / or the degree of matching between the second project business knowledge and the general knowledge, wherein the second dynamic coefficient includes at least one of the following: a second dynamic weight, a second dynamic threshold value.

[0151] Following the above example, the above-mentioned adjustment of the second project data based on at least the second project business knowledge and the first dynamic coefficient corresponding thereto can include adjustment of the second project data based on the second project business knowledge and the first dynamic coefficient corresponding thereto, and the general knowledge and the second dynamic coefficient corresponding thereto. Here, in specific implementation, it can be any one or a combination of multiple of the following cases:

[0152] In the case that the dynamic coefficient comprises a dynamic weight, an adjustment ratio corresponding to the second project business knowledge can be determined based on the first dynamic weight, and the second project data is adjusted based on the adjustment ratio. The adjustment ratio corresponding to the second project business knowledge refers to a ratio of a data amount involved in adjustment of the second project data by the second project business knowledge to a total data amount of the second project data. In other words, a part of data of the second project data that needs to be adjusted by the second project business knowledge can be found and adjusted, and the data amount of the part of data is limited by the first dynamic weight. In addition, an adjustment ratio corresponding to the general knowledge can be determined based on the second dynamic weight, and the second project data is adjusted based on the adjustment ratio. The specific principle can be referred to the description of the adjustment ratio corresponding to the second project business knowledge, which will not be described herein.

[0153] In the case that the dynamic coefficient comprises a dynamic threshold, the project business knowledge with an adjustment probability higher than the first dynamic threshold can be determined from the second project business knowledge based on the first dynamic threshold, and the second project data is adjusted based on the project business knowledge with the adjustment probability higher than the first dynamic threshold. The adjustment probability of the project business knowledge can be determined by the degree of association between the corresponding project business knowledge and the second project data. In addition, the general knowledge with an adjustment probability higher than the second dynamic threshold can be determined from the general knowledge based on the second dynamic threshold, and the second project data is adjusted based on the general knowledge with the adjustment probability higher than the second dynamic threshold. The adjustment probability of the general knowledge can be determined by the matching degree between the corresponding general knowledge and the second project business knowledge.

[0154] It should be noted that in the case that the first dynamic coefficient comprises the first dynamic weight, the second dynamic coefficient can comprise or not comprise the second dynamic weight, and the two cases do not have a certain relationship. Similarly, in the case that the first dynamic coefficient comprises the first dynamic threshold, the second dynamic coefficient can comprise or not comprise the second dynamic threshold.

[0155] In a second aspect, correspondingly, the embodiments of the present application further provide an intelligent management system for promoting a business process, which can realize all processes of the intelligent management method for promoting a business process provided by the embodiments.

[0156] Referring to Figure 3 , a structure schematic diagram of an intelligent management system for promoting a business process provided by the embodiments of the present application is shown. The intelligent management system for promoting a business process comprises:

[0157] The data acquisition module 301 is configured to acquire first project data related to a first promotion project of a first business end and acquire original business information corresponding to a second business end, wherein the first project data includes first project text information and first project annotation information of the first promotion project.

[0158] The semantic feature determination module 302 is configured to perform semantic recognition based on the first project text information to obtain first project semantic features.

[0159] The annotation feature determination module 303 is configured to determine first project annotation features based on the first project annotation information.

[0160] The second project data generation module 304 is configured to generate second project data based on the first project semantic features, the first project annotation features and the original business information, wherein a first similarity between the first project annotation information and second project annotation information in the second project data satisfies a first preset similarity condition.

[0161] The business process determination module 305 is configured to determine promotion business process information corresponding to the second business end based on a preset promotion business knowledge graph and the second project data.

[0162] The information pushing module 306 is configured to push the promotion business process information to the second business end.

[0163] In an optional implementation, the first promotion project includes at least one project object, and the first project annotation information includes first object attributes and first object requirements of the at least one project object.

[0164] The first project annotation features are determined based on the first object attributes and the first object requirements of the at least one project object.

[0165] The first object attributes of each project object are subjected to feature extraction to obtain first attribute features of each project object.

[0166] The first object requirements of each project object are subjected to feature extraction to obtain first requirement features of each project object.

[0167] The first project annotation features are determined based on the first attribute features and the first requirement features of the at least one project object.

[0168] In an optional implementation, the first project annotation features are determined based on the first attribute features and the first requirement features of the at least one project object, including:

[0169] Based on the promotion business knowledge graph, first project business knowledge corresponding to the first project data is acquired, wherein the first project business knowledge comprises typical object attribute knowledge and typical object demand knowledge;

[0170] Features of the typical object attribute knowledge and features of the typical object demand knowledge are respectively determined;

[0171] For each project object, first attribute features thereof are subjected to feature interaction processing with features of the typical object attribute knowledge to obtain second attribute features of the project object, and first demand features thereof are subjected to feature interaction processing with features of the typical object demand knowledge to obtain second demand features of the project object, wherein the feature interaction processing comprises cross-attention interaction;

[0172] Based on the second attribute features and the second demand features of the at least one project object respectively, the first project annotation features are determined.

[0173] In an optional implementation, the generating of the second project data based on the first project semantic features, the first project annotation features and the original business information comprises:

[0174] A plurality of business process node frameworks corresponding to the original business information are determined;

[0175] Based on the first project semantic features and the first project annotation features, content generation is respectively performed for the plurality of business process node frameworks, and the generated content is respectively filled into corresponding business process node frameworks to obtain project data corresponding to the plurality of business process node frameworks respectively;

[0176] Based on the first project annotation features and the project data corresponding to the plurality of business process node frameworks respectively, the second project data is determined.

[0177] In an optional implementation, the content generation based on the first project semantic features and the first project annotation features for the plurality of business process node frameworks respectively and the filling of the generated content into corresponding business process node frameworks respectively comprise:

[0178] For each business process node framework in the plurality of business process node frameworks,

[0179] Based on the first project semantic features, one or more semantic association regions are located in the business process node framework;

[0180] Based on the first project annotation features and the first project semantic features, relevant content of the one or more semantic association regions respectively is generated;

[0181] determine the generated content corresponding to the business process node framework based on the relevant content of each of the one or more semantic association areas;

[0182] fill the generated content corresponding to the business process node framework into the business process node framework.

[0183] In an optional implementation, a second similarity between the project data corresponding to each of the business process node frameworks and the information in the first project annotation information matching the business process node framework satisfies a second preset similarity condition.

[0184] In an optional implementation, the determining the second project data based on the first project annotation feature and the project data corresponding to each of the plurality of business process node frameworks comprises:

[0185] For the project data corresponding to each of the business process node frameworks, the generated content corresponding thereto is determined therefrom, and the project data is evaluated based on the first project annotation feature and the generated content corresponding to the project data to obtain a corresponding evaluation result;

[0186] Based on each of the evaluation results, at least part of the project data is determined from the project data corresponding to each of the plurality of business process node frameworks;

[0187] The second project data is determined based on the at least part of the project data.

[0188] In an optional implementation, the determining the promotion business process information corresponding to the second business end based on the preset promotion business knowledge graph and the second project data comprises:

[0189] Based on the promotion business knowledge graph, second project business knowledge corresponding to the second project data is obtained;

[0190] The promotion business process information is determined based on the second project business knowledge and the second project data.

[0191] In an optional implementation, the determining the promotion business process information based on the second project business knowledge and the second project data comprises:

[0192] determine the first dynamic coefficient corresponding to the second project business knowledge by strategy optimization based on a degree of association between the second project business knowledge and the second project data, and / or a degree of matching between the second project business knowledge and general knowledge, wherein the general knowledge is determined by the second project data, and the first dynamic coefficient comprises at least one of a first dynamic weight and a first dynamic threshold;

[0193] adjust the second project data based on at least the second project business knowledge and the first dynamic coefficient corresponding thereto;

[0194] split the adjusted second project data into promotion business process information corresponding to the second business end.

[0195] In a third aspect, an embodiment of the present application provides a non-transitory computer-readable storage medium, having stored thereon a computer program, which, when executed by a processor, implements the steps of any of the above methods.

[0196] In a fourth aspect, an embodiment of the present application provides a computer program product, comprising computer instructions, which, when executed by a processor, implement the steps of any of the above methods.

[0197] In a fifth aspect, an embodiment of the present application provides a computer device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the steps of any of the above methods when executing the computer program.

[0198] Referring to Figure 4 The computer device of the embodiment comprises a processor 401, a memory 402, and a computer program stored in the memory 402 and executable on the processor 401, such as an intelligent management program for promotion business process. The processor 401 implements the steps in each of the above embodiments of the intelligent management method for promotion business process when executing the computer program, such as steps S101-S106 shown in the figure. Figure 1

[0199] For example, the computer program can be divided into one or more modules / units, which are stored in the memory 402 and executed by the processor 401 to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the computer device.

[0200] ​The computer device can be a desktop computer, a notebook computer, a palm computer, a cloud server, and the like. The computer device can include, but is not limited to, a processor 401 and a memory 402. Those skilled in the art can understand that the schematic diagram is only an example of the computer device, and does not constitute a limitation on the computer device, and can include more or fewer components than the diagram, or combine certain components, or different components, for example, the computer device can also include an input / output device, a network access device, a bus, and the like.

[0201] The processor 401 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic components, discrete hardware components, and the like. The general-purpose processor can be a microprocessor or the processor 401 can also be any conventional processor and the like, and the processor 401 is the control center of the computer device, which connects various parts of the computer device through various interfaces and lines.

[0202] The memory 402 can be used to store computer programs and / or modules, and the processor 401 realizes various functions of the computer device by running or executing computer programs and / or modules stored in the memory 402, and calling data stored in the memory 402. The memory 402 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required for a function (such as a sound playing function, an image playing function, and the like), and the like; and the data storage area can store data created according to use of the mobile phone (such as audio data, a phone book, and the like), and the like. In addition, the memory 402 can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0203] The modules / units integrated in the computer device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on this understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a non-transitory computer readable storage medium. When the computer program is executed by the processor 401, the steps of the above-mentioned various method embodiments can be realized. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal, and software distribution medium, etc. that can carry the computer program code.

[0204] In summary, the embodiments of the present application have at least the following beneficial effects:

[0205] By adopting the embodiments of the present application, first project data related to a first promotion project of a first business end is acquired, and original business information corresponding to a second business end is acquired, wherein the first project data includes first project text information and first project annotation information of the first promotion project; semantic recognition is performed based on the first project text information to obtain first project semantic features; first project annotation features are determined based on the first project annotation information; second project data is generated based on the first project semantic features, the first project annotation features, and the original business information, wherein a first similarity between the first project annotation information and second project annotation information in the second project data satisfies a first preset similarity condition; promotion business process information corresponding to the second business end is determined based on a preset promotion business knowledge graph and the second project data; and the promotion business process information is pushed to the second business end, so that intelligent management of the promotion business process can be automatically completed, and management efficiency and accuracy can be improved.

[0206] Those skilled in the art can clearly understand the application by the description of the above embodiments that the application can be implemented by means of software and necessary hardware platform, and of course, can also be implemented by hardware. Based on such understanding, all or part of the technical solutions of the application that make contributions to the background art can be embodied in the form of a software product. The computer software product can be stored in a storage medium, such as a ROM (Read-Only Memory), a RAM (Random Access Memory), a magnetic disk, an optical disk, and the like, and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in various embodiments or some parts of the embodiments of the application.

[0207] The above is the preferred embodiment of the application. It should be pointed out that, for those skilled in the art, without departing from the principles of the application, a number of improvements and refinements can be made, which are also considered within the protection scope of the application.

Claims

1. An intelligent management method for promoting business processes, characterized in that, include: Obtain first project data related to the first promotion project of the first business end, and obtain original business information corresponding to the second business end, wherein the first project data includes the first project text information and the first project annotation information of the first promotion project; Semantic recognition is performed based on the textual information of the first item to obtain the semantic features of the first item; The first project annotation features are determined based on the first project annotation information; Based on the semantic features of the first project, the annotation features of the first project, and the original business information, second project data is generated, wherein the first similarity between the annotation information of the first project and the annotation information of the second project in the second project data satisfies a first preset similarity condition. Based on the preset promotion business knowledge graph and the second project data, determine the promotion business process information corresponding to the second business terminal; The promotion business process information is pushed to the second business terminal.

2. The method according to claim 1, characterized in that, The first promotion project includes at least one project object, and the first project annotation information includes the first object attributes and first object requirements of each of the at least one project object; Determining the first item annotation feature based on the first item annotation information includes: Feature extraction is performed on the first object attribute of each project object to obtain the first attribute feature of each project object; For each project object, the first object requirement feature is extracted to obtain the first requirement feature of each project object. The first project annotation feature is determined based on the first attribute feature and the first requirement feature of each of the at least one project object.

3. The method according to claim 2, characterized in that, The step of determining the first project annotation feature based on the first attribute feature and first requirement feature of each of the at least one project object includes: Based on the promotion business knowledge graph, first project business knowledge corresponding to the first project data is obtained, wherein the first project business knowledge includes typical object attribute knowledge and typical object requirement knowledge. Determine the characteristics of the typical object attribute knowledge and the characteristics of the typical object requirement knowledge respectively; For each project object, feature interaction processing is performed on its first attribute feature and the feature of the typical object attribute knowledge to obtain the second attribute feature of the project object. In addition, feature interaction processing is performed on the first requirement feature of the project object and the feature of the typical object requirement knowledge to obtain the second requirement feature of the project object. The feature interaction processing includes cross-attention interaction. The first project annotation feature is determined based on the second attribute features and second requirement features of each of the at least one project object.

4. The method according to claim 1, characterized in that, The step of generating second project data based on the semantic features of the first project, the annotation features of the first project, and the original business information includes: Determine the framework of multiple business process nodes corresponding to the original business information; Based on the semantic features and annotation features of the first project, content is generated for each of the multiple business process node frameworks, and the generated content is filled into the corresponding business process node frameworks to obtain the project data corresponding to each of the multiple business process node frameworks. Based on the first project annotation features and the project data corresponding to each of the multiple business process node frameworks, the second project data is determined.

5. The method according to claim 4, characterized in that, The step of generating content for each of the multiple business process node frameworks based on the semantic features and annotation features of the first project, and then filling the generated content into the corresponding business process node frameworks, includes: For each of the multiple business process node frameworks, Based on the semantic features of the first project, one or more semantically related regions are located in the business process node framework. Based on the annotation features and semantic features of the first project, generate relevant content for each of the one or more semantically related regions; Based on the relevant content of each of the one or more semantically related regions, determine the generated content corresponding to the business process node framework; The generated content corresponding to the business process node framework is then filled into the business process node framework.

6. The method according to claim 4, characterized in that, The project data corresponding to each business process node framework, and the information in the first project annotation information that matches the business process node framework, satisfy the second preset similarity condition between the two.

7. The method according to claim 4, characterized in that, The step of determining the second project data based on the first project annotation features and the project data corresponding to each of the multiple business process node frameworks includes: For each business process node framework, the corresponding project data is determined, and the generated content is evaluated based on the first project annotation features and the generated content corresponding to the project data to obtain the corresponding evaluation result. Based on the evaluation results, at least a portion of the project data is determined from the project data corresponding to each of the multiple business process node frameworks. The second project data is determined based on the at least partial project data.

8. The method according to claim 1, characterized in that, The step of determining the promotion business process information corresponding to the second business terminal based on the preset promotion business knowledge graph and the second project data includes: Based on the promotion business knowledge graph, obtain the second project business knowledge corresponding to the second project data; Based on the business knowledge and data of the second project, the promotion business process information is determined.

9. The method according to claim 8, characterized in that, The step of determining the promotion business process information based on the business knowledge and data of the second project includes: Based on the degree of correlation between the second project business knowledge and the second project data, and / or the degree of matching between the second project business knowledge and general knowledge, a first dynamic coefficient corresponding to the second project business knowledge is determined by strategy optimization, wherein the general knowledge is determined by the second project data, and the first dynamic coefficient includes at least one of the following: a first dynamic weight, a first dynamic threshold; The data for the second project shall be adjusted based at least on the business knowledge of the second project and its corresponding first dynamic coefficient; The adjusted second project data is split into promotion business process information corresponding to the second business terminal.

10. An intelligent management system for promoting business processes, characterized in that, include: The data acquisition module is used to acquire first project data related to the first promotion project of the first business end, and to acquire original business information corresponding to the second business end. The first project data includes the first project text information and the first project annotation information of the first promotion project. The semantic feature determination module is used to perform semantic recognition based on the text information of the first item to obtain the semantic features of the first item. An annotation feature determination module is used to determine the first item annotation feature based on the first item annotation information; The second project data generation module is used to generate second project data based on the semantic features of the first project, the annotation features of the first project, and the original business information, wherein the first similarity between the annotation information of the first project and the annotation information of the second project in the second project data satisfies the first preset similarity condition. The business process determination module is used to determine the promotion business process information corresponding to the second business terminal based on the preset promotion business knowledge graph and the second project data. The information push module is used to push the promotion business process information to the second business terminal.

Images