CN117057319A

CN117057319A - Marketing information generation method, device and equipment

Info

Publication number: CN117057319A
Application number: CN202311048684.8A
Authority: CN
Inventors: 郑丹丹; 白海波; 葛强; 李晓婵
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2023-08-18
Filing date: 2023-08-18
Publication date: 2023-11-14

Abstract

The embodiment of the specification discloses a marketing information generation method, a marketing information generation device and marketing information generation equipment. The scheme may include: in the operation process of the marketing campaign, the target prompt word can be automatically generated by utilizing a first large language model according to preset static marketing configuration information for describing the marketing object and dynamic marketing configuration information capable of changing along with the marketing result of the historical marketing information aiming at the marketing object displayed in the session group, and the first marketing information aiming at the marketing object required to be displayed at the session group can be automatically generated by utilizing a second large language model according to the target prompt word.

Description

Marketing information generation method, device and equipment

Technical Field

The present application relates to the field of internet technologies, and in particular, to a marketing information generating method, device and equipment.

Background

Marketing may refer to discovering or exploring consumer needs, letting consumers learn about products of merchants, and issuing offers to consumers, thereby facilitating the consumer's process of purchasing the products. Currently, marketing service providers manually construct and adjust artificial intelligence prompt words (AIprompt) according to marketing demands of merchants on marketing objects, so as to generate marketing information for the marketing objects according to the artificial intelligence prompt words by using an artificial intelligence algorithm, and release the marketing information for the marketing objects in a session group in social software to develop marketing activities.

Based on the method, how to improve the use convenience, the operation efficiency and the flexibility of the marketing information generation scheme becomes a technical problem to be solved urgently.

Disclosure of Invention

The marketing information generation method, the marketing information generation device and the marketing information generation equipment provided by the embodiment of the specification can improve the use convenience, the operation efficiency and the flexibility of the marketing information generation scheme.

In order to solve the above technical problems, the embodiments of the present specification are implemented as follows:

the marketing information generation method provided by the embodiment of the specification comprises the following steps:

acquiring target marketing configuration information aiming at a marketing object; the target marketing configuration information includes: the system comprises preset static marketing configuration information and dynamic marketing configuration information, wherein the preset static marketing configuration information at least comprises information for describing the marketing objects, and the dynamic marketing configuration information is determined based on marketing results of historical marketing information aiming at the marketing objects, which are displayed in a session group;

generating a target prompt word according to the target marketing configuration information by using a first large language model;

and generating first marketing information aiming at the marketing object, which is required to be displayed at the session group, according to the target prompt word by using a second large language model.

The marketing information generating device provided in the embodiment of the present specification includes:

the first acquisition module is used for acquiring target marketing configuration information aiming at a marketing object; the target marketing configuration information includes: the system comprises preset static marketing configuration information and dynamic marketing configuration information, wherein the preset static marketing configuration information at least comprises information for describing the marketing objects, and the dynamic marketing configuration information is determined based on marketing results of historical marketing information aiming at the marketing objects, which are displayed in a session group;

the first generation module is used for generating target prompt words according to the target marketing configuration information by using a first large language model;

and the second generation module is used for generating first marketing information aiming at the marketing object, which is required to be displayed at the session group, according to the target prompt word by using a second large language model.

at least one processor; the method comprises the steps of,

a memory communicatively coupled to the at least one processor; wherein,

the memory stores instructions executable by the at least one processor to enable the at least one processor to:

At least one embodiment provided in this specification enables the following benefits:

in the operation process of the marketing campaign, a target prompt word is automatically generated according to preset static marketing configuration information for describing a marketing object and dynamic marketing configuration information capable of changing along with the marketing result of the historical marketing information aiming at the marketing object displayed in a session group by using a first large language model, and the first marketing information aiming at the marketing object required to be displayed at the session group can be generated according to the target prompt word by using a second large language model. Because in the marketing information generation process, the target prompt word does not need to be constructed manually and continuously adjusted, the use convenience and the operation efficiency of the marketing information generation scheme are improved, and the human resources required to be consumed are reduced. In addition, as the target prompt word can be changed along with the marketing result of the historical marketing information aiming at the marketing object displayed in the session group, on one hand, the effectiveness of the target prompt word is improved, and the accuracy and the real-time performance of the generated marketing information are further improved, on the other hand, the target prompt word which is continuously and automatically updated can be utilized in the operation process of the marketing activity to continuously update the marketing information required to be displayed at the session group, and because manual intervention is not needed, the convenience and the flexibility of the use of the marketing information generation scheme are also improved.

Drawings

In order to more clearly illustrate the embodiments of the present description or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments described in the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is an application scenario schematic diagram of a marketing information generating method according to an embodiment of the present disclosure;

FIG. 2 is a flowchart illustrating a method for generating marketing information according to an embodiment of the present disclosure;

FIG. 3 is a schematic diagram of a lane flow corresponding to the marketing information generating method in FIG. 2 according to an embodiment of the present disclosure;

FIG. 4 is a schematic diagram of a marketing message generating device corresponding to FIG. 2 according to an embodiment of the present disclosure;

fig. 5 is a schematic structural view of a marketing information generating apparatus corresponding to fig. 2 provided in the embodiment of the present specification.

Detailed Description

For the purposes of making the objects, technical solutions and advantages of one or more embodiments of the present specification more clear, the technical solutions of one or more embodiments of the present specification will be clearly and completely described below in connection with specific embodiments of the present specification and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present specification. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without undue burden, are intended to be within the scope of one or more embodiments herein.

The following describes in detail the technical solutions provided by the embodiments of the present specification with reference to the accompanying drawings.

In the prior art, before a marketing campaign is started, a staff member at a marketing service provider is required to repeatedly adjust an artificial intelligence prompt word and marketing information generated based on the artificial intelligence prompt word according to a marketing theme, a marketing campaign and a marketing product, so as to generate marketing information required to be put into a session group. The marketing information generation scheme needs more manual participation and modification, has poor convenience in use and has lower operation efficiency. Also, only one type of marketing message is typically pre-generated for each marketing campaign, and the marketing messages that are desired to be placed into the session group are typically not altered during the marketing campaign operating cycle, so that the user is typically only able to see one type of marketing message during the marketing campaign operating cycle. Because marketing information can not be iterated and adjusted rapidly according to the conditions in the conversation group, not only are accuracy, instantaneity and flexibility poor, but also users are easy to feel tired and lose interest, and the marketing effect is influenced.

In order to solve the drawbacks of the prior art, the present solution provides the following embodiments:

Fig. 1 is an application scenario schematic diagram of a marketing information generating method according to an embodiment of the present disclosure.

As shown in fig. 1, during operation of the marketing campaign, the server side device 14 of the marketing server may obtain target marketing configuration information for the marketing object; wherein, the target marketing configuration information may include: the preset static marketing configuration information may include at least information describing the marketing object, and the dynamic marketing configuration information may be information determined based on a marketing result of the history marketing information for the marketing object shown in the session group.

The service side device 14 of the marketing service provider can also generate a target prompt word according to the target marketing configuration information by using the first large language model; and generating first marketing information aiming at the marketing object, which is required to be displayed at the session group, according to the target prompt words by using a second large language model. Subsequently, the first marketing message generated by the server side device 14 of the marketing server may be sent to the terminal devices (e.g., device 11, device 12, device 13, etc.) of the group members in the session group, so that the group members may utilize the personal terminal devices to view the first marketing message.

Next, a marketing information generation method provided for the embodiments of the specification will be specifically described with reference to the accompanying drawings:

fig. 2 is a flowchart of a marketing information generating method according to an embodiment of the present disclosure. From the program perspective, the execution subject of the process may be a service-side device of the marketing service provider, or an application program carried at the service-side device. As shown in fig. 2, the process may include the steps of:

step 202: acquiring target marketing configuration information aiming at a marketing object; the target marketing configuration information includes: the system comprises preset static marketing configuration information and dynamic marketing configuration information, wherein the preset static marketing configuration information at least comprises information for describing the marketing objects, and the dynamic marketing configuration information is determined based on marketing results of historical marketing information aiming at the marketing objects, which are displayed in a session group.

In the embodiment of the present disclosure, the marketing object may be a core object of a marketing campaign, which may specifically be a main object of a merchant production and marketing work, and in practical application, the marketing object may be a physical product, or may be a virtual product such as a service, an event, an experience, a property, a data asset, and the like, which is not limited specifically.

In the embodiment of the present disclosure, the marketing campaign for the marketing object may be configured according to the actual needs of the merchant to which the marketing object belongs, so as to obtain the target marketing configuration information for the marketing object. The target marketing configuration information may generally include preset static marketing configuration information that does not change during the operation of the marketing campaign, for example, text or image information for describing the marketing object, and may further include: marketing purposes, marketing style (discounts, offers, issued coupons), marketing information morphology category (plain text messages, picture messages, graphic messages, link messages, coupons, merchandise messages, applet messages), expression setting information, marketing title information, marketing campaign duration range, and the like. This is not particularly limited.

In order to enable the generated marketing message to be quickly iterated and adjusted according to the conditions in the session group, the target marketing configuration message may also include dynamic marketing configuration messages, such as a marketing message style type and a marketing message delivery time, which can be changed according to the marketing results of the historical marketing messages of the marketing objects shown in the session group. In addition, the dynamic marketing configuration information may further include holiday setting information having timeliness, etc., which is not particularly limited.

The session group can be an instant messaging mode supporting multi-person communication provided by an application program, belongs to a chat social service, and generally comprises a plurality of group members, wherein the group members can be mutually connected and communicated through various text messages, voice messages, pictures or video and other media messages, and can perform operations such as multi-person real-time chat, group voice, video conversation and the like, so that the method is convenient and quick. Because the session group is widely applied to social media marketing, merchants can carry out brand promotion, marketing activities, customer services and the like through group messages, and the viscosity of users and the purchase conversion rate are improved. In particular, the historical marketing information for the marketing object that is presented in the conversation group may generally be marketing information that is generated using the method of FIG. 2 and that has been published into the conversation group.

In practical application, the session group may also have an electronic coupon related to the marketing object, so that the group member can get the electronic coupon from the session group and use the electronic coupon to purchase the marketing object. Alternatively, the session group may also have a purchase channel associated with the marketing object such that the group member can utilize the purchase channel to purchase the marketing object. Based on the above, after the historical marketing information aiming at the marketing object is displayed in the session group, the acquisition and use condition of the electronic coupons by the subsequent group members can be detected, or the use condition of the purchase channels by the subsequent group members can be detected to be used as the marketing result of the historical marketing information aiming at the marketing object displayed in the session group, so that the method is convenient and quick and has good accuracy.

If the marketing result indicates that the historical marketing information aiming at the marketing object displayed in the session group is not obviously improved in the use possibility of the electronic coupons and the purchasing channels, the dynamic marketing configuration information in the target marketing configuration information needs to be adjusted, so that the marketing information put into the session group can be quickly iterated and adjusted along with the situation in the session group, thereby being beneficial to improving the accuracy, the instantaneity and the flexibility of the marketing information generation scheme, and avoiding the group members from producing tiredness and losing interest so as to improve the marketing effect.

In practical application, the application program to which the session group belongs and the verification program of the electronic coupon may be the same application program or different application programs; and, the application to which the session group belongs and the application to which the purchase channel belongs may be the same application or different applications; this is not particularly limited. The execution subject of the method in fig. 2 may be a server device of an application program to which the session group belongs, or may be a device capable of being communicatively connected to a server device of an application program to which the session group belongs, which is not particularly limited.

For example, when the session group is a group at an electronic transaction platform, the electronic coupon may be a virtual coupon issued at the electronic transaction platform, and the purchase channel may be a purchase channel at the electronic transaction platform. Alternatively, when the session group is a group of payment applications, the electronic coupon may be a virtual coupon issued by an electronic transaction applet installed at the payment application, and the purchase channel may be a purchase channel at the electronic transaction applet. Alternatively, when the session group is a group of social applications, the electronic coupon may be a virtual coupon issued by an application other than the social application, and the purchase channel may be a purchase channel at an application other than the social application.

Step 204: and generating target prompt words according to the target marketing configuration information by using the first large language model.

In the embodiment of the present specification, the large language model (Large Language Model, abbreviated LLM) may be a deep learning-based natural language processing model that can learn the grammar and semantics of natural language so that human-readable text can be generated. The large language model can process various natural language tasks such as text classification, question-answering, dialogue, natural language generation, machine translation and the like.

Prompt may refer to an Artificial Intelligence (AI) prompter, a method of using natural language to direct or motivate an artificial intelligence model to accomplish a particular task. Briefly, promt is a text Prompt that directs the AI to perform specific tasks, such as generating text, translating, answering questions, etc. At present, the application scene of the Prompt is also very rich. In the chat robot field, promt can be used to generate more humanized, intelligent conversations; in the search engine field, promt can be used to optimize the quality of search results; in the intelligent customer service field, promt can be used for improving the response speed and the resolution of customer service. Even in the art, promt can be used to create artwork, such as poetry, painting, and the like. Because of the specific requirements and context of tasks that the design of promt requires to take into account, it is often necessary for a staff member with a high skill level and expertise to manually generate promt.

In the embodiment of the present disclosure, when the target prompt word required for generating the marketing information needs to be constructed, the staff with higher technical level and expertise is not required to manually generate the target prompt word, but the target marketing configuration information obtained in step 202 is input into the first large language model, so that the natural language generating capability of the first large language model is utilized to automatically generate the target prompt word.

In practical application, the first large language model can be trained by using a small number of training samples (including a static marketing configuration information sample, a dynamic marketing configuration information sample and a manually constructed prompting word sample) in advance, so that the accuracy of target prompting words generated by using the first large language model is improved. Or, large language models with good self-adaptation capability such as LLaMA models, chatGPT (Chat Generative Pre-trained Transformer) and the like can be utilized to generate target prompt words, and parameters of the models can be quickly and automatically adjusted to cope with different contexts without retraining the large language models, so that the model training process can be reduced, and convenience and rapidness are realized.

Step 206: and generating first marketing information aiming at the marketing object, which is required to be displayed at the session group, according to the target prompt word by using a second large language model.

In this embodiment of the present disclosure, the second large language model may be a large language model for executing a natural language generating task, and after the target prompt word obtained in step 204 is input into the second large language model, the natural language generating capability of the second large language model may be utilized to automatically generate the first marketing information for the marketing object in text form, where the first marketing information may be a marketing document. In practical applications, the second large language model and the first large language model may be different models, which is favorable for improving the accuracy of the generated target prompt word and the first marketing information. Of course, the second large language model and the first large language model may be the same model, which is not particularly limited.

In practical application, the second large language model can be trained by using a small amount of training samples (including the prompt word sample and the marketing document information generated by using the first large language model) in advance, so that the accuracy of the first marketing information generated by using the second large language model is improved. Or, the first marketing information can also be directly generated by using a LLaMA model, a ChatGPT (full name: chat Generative Pre-trained Transformer) and other large language models with better self-adaption capability, so that the model training process can be reduced, and the method is convenient and quick.

The method in fig. 2 does not need to manually construct and continuously adjust the target prompt words in the marketing information generation process, so that the convenience in use and the high efficiency in operation of the marketing information generation scheme are improved, and the human resources required to be consumed are reduced. In addition, as the target prompt word can be changed along with the marketing result of the historical marketing information aiming at the marketing object displayed in the session group, on one hand, the effectiveness of the target prompt word is improved, and the accuracy and the real-time performance of the generated marketing information are further improved, on the other hand, the target prompt word which is continuously and automatically updated can be utilized in the operation process of the marketing activity to continuously update the marketing information required to be displayed at the session group, and because manual intervention is not needed, the convenience and the flexibility of the use of the marketing information generation scheme are also improved.

Based on the method in fig. 2, the examples of the present specification also provide some specific embodiments of the method, as described below.

In the embodiment of the present disclosure, in order to ensure the normal operation of the solution, before the marketing campaign is started, the management user needs to preset the static marketing configuration information and the initial dynamic marketing configuration information for the marketing object in advance. In addition, the preset static marketing configuration information is not changed in the subsequent marketing campaign operation process, and the initial dynamic marketing configuration information may be automatically adjusted according to the marketing result of the marketing information displayed at the session group.

Based on this, step 202: the obtaining target marketing configuration information for the marketing object may specifically include:

acquiring the preset static marketing configuration information and initial dynamic marketing configuration information which are set by a management user aiming at the marketing object; or,

acquiring the preset static marketing configuration information set by a management user aiming at the marketing object; the method comprises the steps of,

obtaining a marketing result of the history marketing information aiming at the marketing object, which is recently shown in the session group; the historical marketing information is marketing information aiming at the marketing object according to the generation of the historical prompt words by utilizing the second large language model; the historical prompt words are generated by the first large language model according to the preset static marketing configuration information and the historical dynamic marketing configuration information.

And if the marketing result does not meet the preset marketing requirement, adjusting the historical dynamic marketing configuration information to obtain the target dynamic marketing configuration information aiming at the marketing object.

In the embodiment of the present disclosure, before the marketing campaign is started, or when the marketing campaign is just started and no marketing information for the marketing object is put in the session group, at this time, the marketing result of the historical marketing information for the marketing object shown in the session group cannot be obtained, so that no adjustment is required for the initial dynamic marketing configuration information. At this time, in order to ensure the comprehensiveness of the acquired target marketing configuration information for the marketing object, preset static marketing configuration information set by the management user for the marketing object may be determined as the target marketing configuration information together with the initial dynamic marketing configuration information.

After the marketing campaign is started and the historical marketing information for the marketing target generated based on the scheme principle in fig. 2 is already put in the session group, it is generally required to identify whether the validity and accuracy of the last-put historical marketing information are poor, and if so, it is required to adjust the dynamic marketing configuration information used when the last-put historical marketing information is generated. If not, the dynamic marketing configuration information used in the process of generating the last released historical marketing information is not required to be adjusted, so that the marketing information required to be released at the conversation group is not required to be changed; or, the dynamic marketing configuration information used in generating the last released historical marketing information may be continuously adjusted, so as to change the marketing information required to be released at the session group, so as to promote the interest of the user, which is not limited specifically.

In practical application, if the marketing result of the last-displayed historical marketing information for the marketing object in the session group does not meet the preset marketing requirement, the effectiveness and the accuracy of the last-released historical marketing information can be poor. If the marketing result of the last-displayed historical marketing information aiming at the marketing object in the session group meets the preset marketing requirement, the last-released historical marketing information can be indicated to have better effectiveness and accuracy.

The types of the marketing results may be various, and the preset marketing requirements may also be generally set by the management user according to the types of the marketing results and the actual requirements, which is not limited in particular. This is illustrated for ease of understanding. For example, the marketing results may include: the target business index is used for reflecting the processing condition of the group members in the session group on the electronic coupons of the marketing objects after the historical marketing information aiming at the marketing objects is displayed in the session group, wherein the target business index can comprise the electronic coupon retrieval rate, the electronic coupon verification rate and the like, and the preset marketing requirement can be that the electronic coupon retrieval rate and/or the electronic coupon verification rate is greater than or equal to a preset threshold value and the like. Alternatively, the marketing result may further include: after the historical marketing information for the marketing object is displayed in the session group, the purchasing rate, the repurchase rate and the like of the group members in the session group for the marketing object are reflected, and the preset marketing requirement can be that the purchasing rate and/or the repurchase rate is greater than or equal to a preset threshold value and the like.

In the embodiment of the present disclosure, the dynamic marketing configuration information may be various, and the manner of adjusting the dynamic marketing configuration information according to the marketing result of the historical marketing information may also be various, so as to facilitate understanding, and explain the adjustment principle of the dynamic marketing configuration information.

Specifically, when the dynamic marketing configuration information includes: when the marketing information style is of the type, the adjusting the historical dynamic marketing configuration information to obtain the target dynamic marketing configuration information for the marketing object may include:

and determining a second marketing information style except the first marketing information style corresponding to the historical dynamic marketing configuration information from the preset marketing information style types to obtain the target dynamic marketing configuration information aiming at the marketing object.

In the embodiment of the present disclosure, the preset style of the marketing message may be used to represent the style that the marketing message can have generated by using the scheme in fig. 2. For example, the preset marketing message style types may include, but are not limited to: poetry, network language, looks, small, two-person, children's interest, serious, etc. In practical applications, the preset style of marketing information may be all the styles supported by the method main body in fig. 2, or may be a part of style selected by the management user from all the styles supported by the method main body in fig. 2, so as to ensure the willingness of the user, which is not limited specifically.

When the marketing result of the last displayed historical marketing information does not meet the preset marketing requirement, the marketing effect of the first marketing information style corresponding to the historical marketing information is not good, so that the preset marketing information style except the first marketing information style is required to be used as the second marketing information style. Subsequently, the marketing information of the second marketing information style aiming at the marketing object can be generated by using the scheme in fig. 2 and is distributed to the session group, so that users in the session group can view the marketing information of different styles, the interest of the users is favorably improved, and the marketing effect is further improved.

In the embodiment of the present disclosure, in the operation process of a marketing campaign, after a history marketing message for a marketing object is put in a session group, a marketing result of the history marketing message may be obtained, and because the marketing result may include a history business index focused by a merchant, a correspondence between the history business index and a preset marketing message style type to which the history marketing message belongs may be established. So as to accurately adjust the style of the marketing message based on the corresponding relation.

Based on this, the determining, from the preset marketing information style types, the second marketing information style type other than the first marketing information style type corresponding to the historical dynamic marketing configuration information may specifically include:

determining the second marketing information style type from the preset marketing information style types of the corresponding previous preset number with the largest historical service index according to the corresponding relation between the preset marketing information style type and the historical service index; or,

and determining the second marketing information style type from the preset marketing information style types without the corresponding historical service indexes according to the corresponding relation between the preset marketing information style type and the historical service indexes.

And, the obtained marketing result of the historical marketing information for the marketing object, which is displayed in the conversation group last time, may include: and the target business index is used for reflecting the processing condition of the group members in the session group on the electronic coupons of the marketing objects after the historical marketing information aiming at the marketing objects is displayed in the session group. After obtaining the marketing result of the history marketing information for the marketing object, which is displayed in the conversation group last time, the method can further comprise:

Storing a corresponding relation between a target business index contained in the marketing result and the first marketing information style type; wherein the historical business index is consistent with the category of the target business index.

In the embodiment of the present disclosure, during the operation of the marketing campaign, if a marketing message of a certain preset marketing message style type has not been issued in the session group, there will be no specific numerical value of the historical business index having a corresponding relationship with the preset marketing message style type. If a certain marketing message of a preset marketing message style is issued in the session group, a specific numerical value of a historical business index with a corresponding relation with the preset marketing message style exists. Based on the method, when the second marketing information style is required to be determined, the preset marketing information style with larger specific values of the corresponding historical service indexes can be preferentially used as the second marketing information style, so that the marketing effect of the subsequently generated marketing information can be guaranteed; or, the preset marketing information style which does not have the specific value of the corresponding historical service index can be used as the second marketing information style, so that the marketing information of the brand new style can be generated conveniently, and the novelty of the marketing information is improved.

In the embodiments of the present description, the a/B test (also referred to as ABtest) may refer to a way to randomly display to a user a variety of schemes, and statistical analysis determines what schemes are better for a given conversion target. In order to improve the efficiency of determining the preset marketing information style with good marketing results, multiple second marketing information style types can be selected so as to simultaneously put marketing information of different style types into different group members.

Based on this, the number of categories of the second marketing message style type included in the target dynamic marketing configuration message may be greater than 1.

Correspondingly, step 204: generating a target prompt word according to the target marketing configuration information by using a first large language model, wherein the method specifically comprises the following steps:

and generating the target prompt word corresponding to any one of the second marketing information style types according to the target marketing configuration information carrying any one of the second marketing information style types by using the first large language model to obtain a target prompt word set.

Step 206: utilizing a second large language model, generating first marketing information aiming at the marketing object, which is required to be displayed at the session group, according to the target prompt word, specifically comprises the following steps:

Generating various first marketing information aiming at the marketing object according to various target prompt words in the target prompt word set by utilizing the second large language model; and the style types of the second marketing information corresponding to the different first marketing information are different.

In addition, the method in fig. 2 may further include:

for any one of the first marketing messages, sending a group message carrying the first marketing message to terminal equipment of part of group members in the session group; wherein the terminal device of any one of the group members receives only one of the first marketing messages.

In this embodiment of the present disclosure, when the target dynamic marketing configuration information includes multiple second marketing information style types, a corresponding target prompt word and a corresponding first marketing information may be generated for each of the second marketing information style types. Subsequently, various first marketing messages need to be distributed to different group members, so that each group member can view one first marketing message, and the first marketing messages viewed by the different group members can be different. And then, separately counting marketing results of the group users receiving the first marketing information, and obtaining the historical business indexes corresponding to the second marketing information style types. Is convenient and quick.

In the embodiment of the present disclosure, since the time of delivering the marketing message also affects the marketing result, for example, the marketing result of the marketing message delivered late at night is generally poor, there is a need for adaptively adjusting the time of delivering the marketing message according to the actual situation of the session group.

Specifically, the dynamic marketing configuration information may further include: marketing information delivery time.

Correspondingly, the adjusting the historical dynamic marketing configuration information to obtain the target dynamic marketing configuration information for the marketing object may include:

and determining second marketing information throwing time except the first marketing information throwing time corresponding to the historical dynamic marketing configuration information, and obtaining target dynamic marketing configuration information aiming at the marketing object. The second marketing information release time is the release time of the second marketing information generated in the step 206 and the second marketing information generated in the subsequent embodiment.

In the embodiment of the present specification, since the marketing result may include: a target business index for reflecting the processing condition of the electronic coupons of the marketing objects by group members in the session group after the historical marketing information of the marketing objects is displayed in the session group; therefore, the second marketing information delivery time except the first marketing information delivery time corresponding to the historical dynamic marketing configuration information can be determined based on the corresponding relation between each appointed marketing information delivery time and the historical business index, and accuracy is good.

Based on this, the determining the second marketing information delivery time except the first marketing information delivery time corresponding to the historical dynamic marketing configuration information may specifically include:

determining the second designated marketing information delivery time from the designated marketing information delivery time of the preset number with the maximum corresponding historical service indexes according to the corresponding relation between the designated marketing information delivery time and the historical service indexes; wherein the historical service index is consistent with the type of the target service index; or,

and determining the second designated marketing information delivery time from the marketing information delivery time without the corresponding historical service index according to the corresponding relation between the designated marketing information delivery time and the historical service index.

In addition, the method in fig. 2 may further include:

and storing the corresponding relation between the target business index and the first marketing information throwing time.

In the embodiment of the present disclosure, the principle adopted when determining the second marketing information delivery time except the first marketing information delivery time corresponding to the historical dynamic marketing configuration information and the principle adopted when determining the second marketing information style except the first marketing information style corresponding to the historical dynamic marketing configuration information may be identical, and will not be described in detail.

In the embodiment of the specification, the accurate index condition and feedback can be mastered by utilizing the data analysis and algorithm optimization technology, the marketing information throwing time can be adjusted in real time, the problem that the traditional scheme depends on personal experience of staff to fix the marketing information throwing time period is avoided, the differentiated management of different areas and different groups is realized, and the marketing effect and the conversion rate are improved.

In the embodiment of the present disclosure, when the operation time range of the marketing campaign approaches or includes a holiday, the marketing configuration information set by the management user for the marketing target may further include preset holiday information. Since the holiday is indicated to have been spent when the current time is later than the holiday date, there is no need to continue to generate marketing information related to the holiday, and based on this, the holiday information can be used as dynamic marketing configuration information. Implementations of adjusting for holiday information contained in dynamic marketing configuration information are also presented herein.

Specifically, step 202: acquiring target marketing configuration information for a marketing object may include:

and acquiring preset holiday information set by the management user for the marketing object.

And judging whether the current time is later than the holiday ending time corresponding to the preset holiday information.

If yes, the target marketing configuration information is forbidden to contain the preset holiday information.

If not, the target marketing configuration information contains the preset holiday information.

In the embodiment of the present disclosure, the types of the forms of the marketing information for the marketing object may be various, and for example, may include a text type, a picture type, a graphics type, and the like. The first marketing message generated using the second large language model in step 206 of FIG. 2 may be a marketing document message that belongs to a textual class of marketing messages. When the management user indicates that the marketing information of the picture class or the picture-text class also needs to be generated, an implementation manner of generating the second marketing information of the picture class or the picture-text class aiming at the marketing object is also provided.

Specifically, the preset static marketing configuration information may include: carrying the target image of the marketing object. Correspondingly, the method described in fig. 2 may further comprise:

and carrying out image main body extraction processing on the target image to obtain a marketing object image.

Generating second marketing information aiming at the marketing object to be displayed at the session group according to the marketing object image and the first marketing information by utilizing a multi-mode large model; the second marketing information is at least one of static marketing image information, dynamic marketing image information and marketing video information.

In the present embodiment, the multi-modal large model refers to a machine learning model that can process a plurality of information from different modalities (e.g., image, voice, text, etc.). The multi-modal large model can be trained in advance by using training samples (for example, a marketing object image sample and a marketing document sample which are used as input information, and a static marketing image information sample, a dynamic marketing image information sample and/or a marketing video information sample which are used as output labels), so that after the multi-modal large model receives the marketing object image and the first marketing information, second marketing information which meets the requirements and aims at a marketing object can be generated. In practical applications, the second marketing message may be a static, dynamic or video form marketing poster. The second marketing information may include, in addition to the marketing object image and the first marketing information, poster background information or other poster elements that are automatically generated by the multi-mode big model based on the marketing object image and the first marketing information, which is not limited in detail.

In practical applications, the marketing document sent as the group message to the session group is not generally consistent with the marketing document to be added to the marketing poster, so, in order to improve the accuracy of the issued marketing message, the first marketing message generated in step 206 may specifically include: first marketing document information in the form of a poster message and second marketing document information in the form of a group message.

Correspondingly, the generating, by using the multi-modal large model, the second marketing information for the marketing object to be displayed at the session group according to the marketing object image and the first marketing information may specifically include:

and generating the second marketing information aiming at the marketing object, which is required to be displayed at the session group, according to the marketing object image and the first marketing document information by utilizing the multi-mode large model.

In addition, the method in fig. 2 may further include:

and sending the group message carrying the second marketing document information to the terminal equipment of the group members in the session group. The method comprises the steps of,

and sending the group message carrying the second marketing information to the terminal equipment of the group members in the session group.

In the embodiment of the present disclosure, a single second large language model may be used to generate the first marketing document information and the second marketing document information, or 2 second large language models may be used to generate the first marketing document information and the second marketing document information, which is not limited in particular.

And when the forms of the second marketing information are also multiple, the same multi-mode big model can be used for generating the second marketing information of each form, or a plurality of multi-mode big models can be used for generating the second marketing information of each form respectively, which is not particularly limited.

In the embodiment of the specification, the second marketing text information in text form and the second marketing information in picture/moving picture/video form are respectively sent to the terminal equipment of the group member, so that the type of the marketing information which can be seen by the group member is improved, and the user experience is improved.

In practical application, if the marketing information form category set by the management user only includes the picture/moving picture/video form, the second marketing information in the picture/moving picture/video form may be sent to the terminal device of the group member only, and the second marketing information in the text form may not be sent to the terminal device of the group member. Or if the marketing information form types set by the management user only comprise plain text messages, the second marketing information in the form of pictures/moving pictures/videos does not need to be generated, so that the second marketing information in the form of text is only sent to the terminal equipment of the group member, and the flexibility is good.

In the embodiment of the specification, since the marketing server also needs to automatically reply to the session message issued by the group member in the session group, intelligent operation and marketing are realized. Based on this, the present embodiment of the present disclosure also provides an implementation of automatically adjusting the content and form of the group message according to the behavior and feedback of the group members to automatically reply to the session information at the session group.

Specifically, the method in fig. 2 may further include:

and acquiring target session information issued by the target group member in the session group.

And generating target reply information aiming at the target session information according to the target session information and at least part of the target marketing configuration information by using a third large language model.

And sending a fifth group message carrying the target reply information to terminal equipment of at least part of group members including the target group member in the session group.

In this embodiment of the present disclosure, the third large language model may be a large language model for executing question-answering and dialogue tasks, and after the target session information and at least part of the target marketing configuration information are input into the third large language model, the natural language generating capability of the third large language model may be utilized to automatically generate the target reply information for the target session information in text form, so as to assist in implementing intelligent customer service, service consultation and business handling functions.

Wherein the at least part of the target marketing configuration information may include: the marketing information style in the dynamic marketing configuration information can enable the third large language model to generate reply information of different styles along with continuous change of the marketing information style in the dynamic marketing configuration information, and intelligent sending of thousands of faces and capability of adjusting reply content are achieved. The at least part of the target marketing configuration information may also include other kinds of marketing configuration information, which is not particularly limited.

In practical applications, the third largest language model may also have a knowledge base (knowledgeg), or the third largest language model may be trained in advance using information in the knowledge base (knowledgeg). The third biggest language model can be provided with the ability to connect, throw and make stems by allowing the knowledge base (knowledgee) to contain laugh points, vowels or information specifically related to meaning or irony. And the user experience is improved.

In the embodiment of the present specification, an expression image (emoticon) may refer to a manner of representing emotion by using pictures. The expression image is a popular culture formed after social software is active. Based on this, when the preset static marketing configuration information set by the management user includes information for indicating to use the expression image related to the marketing object, it is further required to generate the expression image related to the marketing object, and reply the session information of the group member by using the expression image, so as to improve the user experience.

Specifically, if the preset static marketing configuration information includes: at least one of a target image carrying the marketing object and a target description text for the marketing object, and information for indicating use of an expression image related to the marketing object; the target description text can be used for introducing names, categories, shapes, colors, brands, characteristics and the like of marketing objects. The method of fig. 2 may further include:

And generating an expression image related to the marketing object according to at least one of the target image and the target description text by using an image generation model. The expression image may be at least one of a still image and a moving image.

Correspondingly, the generating, by using the third large language model, target reply information for the target session information according to the target session information and at least part of the target marketing configuration information may specifically include:

and generating initial reply information aiming at the target session information according to the target session information and at least part of the target marketing configuration information by using the third large language model.

And adding at least one expression image in the initial reply information to obtain target reply information aiming at the target session information.

In the embodiment of the present disclosure, the image generation model may be a model for generating an image, which is generally set up by using a deep learning algorithm, for example, the image generation model may be a graphical model, a meridional model, or a multi-model, which is not limited in particular. In practical applications, training samples (for example, a marketing object image sample and/or a corresponding description text sample may be used as input data, and an expression image sample of a corresponding marketing object may be used as label data) may be used in advance to train the image generation model, so that the image generation model may generate an expression image meeting requirements. The expression image should generally include an image of the marketing object, thereby being beneficial to improving user experience.

In the embodiment of the specification, the generated target reply information aiming at the target session information can comprise the expression image related to the marketing object, so that the user experience is improved, and the marketing effect is further improved.

The first large language model, the second large language model, the third large language model, the multi-modal model and the image generation model mentioned in the embodiments of the present disclosure may be deployed at the execution subject of the method in fig. 2, or may be deployed at another platform (for example, an AIPaaS artificial intelligence platform, i.e., a service) capable of being communicatively connected to the execution subject of the method in fig. 2, so long as the execution subject of the method in fig. 2 can generate the information required by using the models mentioned above, which is not limited in particular.

Fig. 3 is a schematic flow chart of a lane corresponding to the marketing information generating method in fig. 2 according to the embodiment of the present disclosure. As shown in fig. 3, the marketing information processing procedure may involve an execution subject of a service end device of a marketing server, a terminal device of a group member, or the like.

In the marketing information generation stage, the service end equipment of the marketing server can acquire the marketing result of the history marketing information aiming at the marketing object, which is displayed in the conversation group last time. If the marketing result does not meet the preset marketing requirement, determining a second marketing information style type except the first marketing information style type corresponding to the historical marketing information from the preset marketing information style types with the largest preset quantity of the corresponding historical service indexes according to the corresponding relation between the preset marketing information style type and the historical service indexes; and determining second marketing information throwing time except the first marketing information throwing time corresponding to the historical marketing information. After the preset static marketing configuration information set by the management user for the marketing object is obtained, the target marketing configuration information containing the preset static marketing configuration information, the second marketing information style, the second marketing information release time and other dynamic marketing configuration information is obtained.

The server device of the marketing service provider may further generate a target prompt word according to the target marketing configuration information by using a first large language model, and generate first marketing information aiming at the marketing object, which needs to be displayed at the session group, according to the target prompt word by using a second large language model, where the first marketing information may include: the first marketing document information in the form of a poster message and the second marketing document information in the form of a group message.

Based on the method, the server side equipment of the marketing server can also perform image main body extraction processing on the target image carrying the marketing object to obtain the marketing object image, and then generate second marketing information aiming at the image type of the marketing object according to the marketing object image and the first marketing information in the form of the poster text in the first marketing information by utilizing the multi-mode large model.

In the marketing information release stage, the server side equipment of the marketing server can send the second marketing information of the image type of the marketing object to the terminal equipment of the group member according to the second marketing information release time, so that the terminal equipment can receive and display the second marketing information of the image type of the marketing object to the group member.

And the service end equipment of the marketing server can also send the second marketing text information in the form of group information in the first marketing information aiming at the marketing object to the terminal equipment of the group member according to the second marketing information release time, so that the terminal equipment can receive and display the second marketing text information in the form of group information aiming at the marketing object to the group member.

Based on the same thought, the embodiment of the specification also provides a device corresponding to the method. Fig. 4 is a schematic structural diagram of a marketing information generating device corresponding to fig. 2 according to an embodiment of the present disclosure. As shown in fig. 4, the apparatus may include:

a first obtaining module 402, configured to obtain target marketing configuration information for a marketing target; the target marketing configuration information includes: the system comprises preset static marketing configuration information and dynamic marketing configuration information, wherein the preset static marketing configuration information at least comprises information for describing the marketing objects, and the dynamic marketing configuration information is determined based on marketing results of historical marketing information aiming at the marketing objects, which are displayed in a session group.

The first generating module 404 is configured to generate, according to the target marketing configuration information, a target prompt word using a first large language model.

And a second generating module 406, configured to generate, according to the target prompt word, first marketing information for the marketing object, which needs to be displayed at the session group, using a second large language model.

The present description example also provides some specific embodiments of the device based on the device of fig. 4, which is described below.

Optionally, the first obtaining module 402 may include:

the first acquisition unit is used for acquiring the preset static marketing configuration information and the initial dynamic marketing configuration information which are set by the management user aiming at the marketing object; or,

the second acquisition unit is used for acquiring the preset static marketing configuration information set by the management user aiming at the marketing object; obtaining a marketing result of the history marketing information aiming at the marketing object, which is displayed in the conversation group last time; the historical marketing information is marketing information aiming at the marketing object according to the generation of the historical prompt words by utilizing the second large language model; the historical prompt words are generated by the first large language model according to the preset static marketing configuration information and the historical dynamic marketing configuration information.

And the adjustment unit is used for adjusting the historical dynamic marketing configuration information to obtain the target dynamic marketing configuration information aiming at the marketing object if the marketing result does not meet the preset marketing requirement.

Optionally, the dynamic marketing configuration information may include: marketing information style type.

Correspondingly, the adjusting unit may specifically be configured to: and determining a second marketing information style except the first marketing information style corresponding to the historical dynamic marketing configuration information from the preset marketing information style types to obtain the target dynamic marketing configuration information aiming at the marketing object.

Optionally, the marketing result may include: and the target business index is used for reflecting the processing condition of the group members in the session group on the electronic coupons of the marketing objects after the historical marketing information aiming at the marketing objects is displayed in the session group.

Correspondingly, the apparatus in fig. 4 may further include:

and the first storage module is used for storing the corresponding relation between the target business index and the first marketing information style type.

The adjusting unit may specifically include:

The first determining subunit is configured to determine, according to a correspondence between the preset marketing information style and the historical service index, the second marketing information style from the preset marketing information style of a previous preset number with the largest corresponding historical service index; wherein the historical service index is consistent with the type of the target service index; or,

and the second determining subunit is used for determining the second marketing information style type from the preset marketing information style types without the corresponding historical service indexes according to the corresponding relation between the preset marketing information style type and the historical service indexes.

Optionally, the number of categories of the second marketing information style type included in the target dynamic marketing configuration information may be greater than 1.

Correspondingly, the first generating module 404 may specifically be configured to:

Correspondingly, the second generating module 406 may specifically be configured to:

The apparatus in fig. 4 may further include:

the first sending module is used for sending a group message carrying any one of the first marketing messages to terminal equipment of part of group members in the session group aiming at any one of the first marketing messages; wherein the terminal device of any one of the group members receives only one of the first marketing messages.

Optionally, the dynamic marketing configuration information may include: marketing information delivery time.

The adjusting unit may specifically be configured to: and determining second marketing information throwing time except the first marketing information throwing time corresponding to the historical dynamic marketing configuration information, and obtaining target dynamic marketing configuration information aiming at the marketing object.

Correspondingly, the apparatus in fig. 4 may further include:

and the second storage module is used for storing the corresponding relation between the target business index and the first marketing information throwing time.

The adjusting unit may specifically include:

the third determining subunit is configured to determine, according to a correspondence between a specified marketing information delivery time and a historical service index, the second specified marketing information delivery time from a preset number of specified marketing information delivery times with the largest corresponding historical service index; wherein the historical service index is consistent with the type of the target service index; or,

and the fourth determining subunit is used for determining the second designated marketing information throwing time from the marketing information throwing time without the corresponding historical service index according to the corresponding relation between the designated marketing information throwing time and the historical service index.

Optionally, the preset static marketing configuration information may include: carrying the target image of the marketing object.

Correspondingly, the apparatus in fig. 4 may further include:

and the image main body extraction module is used for carrying out image main body extraction processing on the target image to obtain a marketing object image.

The third generation module is used for generating second marketing information aiming at the marketing objects, which is required to be displayed at the session group, according to the marketing object image and the first marketing information by utilizing a multi-mode big model; the second marketing information is at least one of static marketing image information, dynamic marketing image information and marketing video information.

Optionally, the first marketing message may specifically include: first marketing document information in the form of a poster message and second marketing document information in the form of a group message.

Correspondingly, the third generating module may be specifically configured to: and generating the second marketing information aiming at the marketing object, which is required to be displayed at the session group, according to the marketing object image and the first marketing document information by utilizing the multi-mode large model.

The apparatus of fig. 4 may further include:

and the second sending module is used for sending the group message carrying the second marketing document information to the terminal equipment of the group member in the session group.

And the third sending module is used for sending the group message carrying the second marketing information to the terminal equipment of the group members in the session group.

Optionally, the apparatus in fig. 4 may further include:

and the second acquisition module is used for acquiring target session information released by the target group member in the session group.

And the fourth generation module is used for generating target reply information aiming at the target session information according to the target session information and at least part of the target marketing configuration information by using a third large language model.

And a fourth sending module, configured to send a group message carrying the target reply message to terminal devices of at least part of group members including the target group member in the session group.

Optionally, the preset static marketing configuration information may include: at least one of a target image carrying the marketing object and a target descriptive text for the marketing object, and information for indicating use of an expression image related to the marketing object.

Correspondingly, the apparatus in fig. 4 may further include:

and a fifth generation module, configured to generate an expression image related to the marketing object according to at least one of the target image and the target description text by using an image generation model.

The fourth generating module may specifically include:

And the initial reply information generating unit is used for generating initial reply information aiming at the target session information according to the target session information and at least part of the target marketing configuration information by utilizing the third large language model.

The target reply information generation unit is used for adding at least one expression image into the initial reply information to obtain target reply information aiming at the target session information.

Based on the same thought, the embodiment of the specification also provides equipment corresponding to the method.

Fig. 5 is a schematic structural view of a marketing information generating apparatus corresponding to fig. 2 provided in the embodiment of the present specification. As shown in fig. 5, the apparatus 500 may include:

at least one processor 510; the method comprises the steps of,

a memory 530 communicatively coupled to the at least one processor; wherein,

the memory 530 stores instructions 520 executable by the at least one processor 510, the instructions being executable by the at least one processor 510 to enable the at least one processor 510 to:

acquiring target marketing configuration information aiming at a marketing object; the target marketing configuration information includes: the system comprises preset static marketing configuration information and dynamic marketing configuration information, wherein the preset static marketing configuration information at least comprises information for describing the marketing objects, and the dynamic marketing configuration information is determined based on marketing results of historical marketing information aiming at the marketing objects, which are displayed in a session group.

And generating target prompt words according to the target marketing configuration information by using the first large language model.

In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for the apparatus shown in fig. 5, the description is relatively simple as it is substantially similar to the method embodiment, with reference to the partial description of the method embodiment being relevant.

In the 90 s of the 20 th century, improvements to one technology could clearly be distinguished as improvements in hardware (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) or software (improvements to the process flow). However, with the development of technology, many improvements of the current method flows can be regarded as direct improvements of hardware circuit structures. Designers almost always obtain corresponding hardware circuit structures by programming improved method flows into hardware circuits. Therefore, an improvement of a method flow cannot be said to be realized by a hardware entity module. For example, a programmable logic device (Programmable Logic Device, PLD) (e.g., field programmable gate array (FieldProgrammable Gate Array, FPGA)) is an integrated circuit whose logic function is determined by the programming of the device by a user. A designer programs to "integrate" a digital system onto a PLD without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Moreover, nowadays, instead of manually manufacturing integrated circuit chips, such programming is mostly implemented by using "logic compiler" software, which is similar to the software compiler used in program development and writing, and the original code before the compiling is also written in a specific programming language, which is called hardware description language (Hardware Description Language, HDL), but not just one of the hdds, but a plurality of kinds, such as ABEL (AdvancedBoolean Expression Language), AHDL (Altera Hardware Description Language), confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (JavaHardware Description Language), lava, lola, myHDL, PALASM, RHDL (Ruby Hardware Description Language), etc., VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog are currently most commonly used. It will also be apparent to those skilled in the art that a hardware circuit implementing the logic method flow can be readily obtained by merely slightly programming the method flow into an integrated circuit using several of the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer readable medium storing computer readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable logic controllers, and embedded microcontrollers, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, atmel AT91SAM, microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic of the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller in a pure computer readable program code, it is well possible to implement the same functionality by logically programming the method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Such a controller may thus be regarded as a kind of hardware component, and means for performing various functions included therein may also be regarded as structures within the hardware component. Or even means for achieving the various functions may be regarded as either software modules implementing the methods or structures within hardware components.

The system, apparatus, module or unit set forth in the above embodiments may be implemented in particular by a computer chip or entity, or by a product having a certain function. One typical implementation is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being functionally divided into various units, respectively. Of course, the functions of each element may be implemented in the same piece or pieces of software and/or hardware when implementing the present application.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.

The application may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The application may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.

Claims

1. A marketing information generation method, comprising:

2. The method of claim 1, wherein the obtaining the target marketing configuration information for the marketing object comprises:

obtaining a marketing result of the history marketing information aiming at the marketing object, which is recently shown in the session group; the historical marketing information is marketing information aiming at the marketing object according to the generation of the historical prompt words by utilizing the second large language model; the historical prompting words are prompting words generated by the first large language model according to the preset static marketing configuration information and the historical dynamic marketing configuration information;

3. The method of claim 2, the dynamic marketing configuration information comprising: marketing information style;

the step of adjusting the historical dynamic marketing configuration information to obtain target dynamic marketing configuration information aiming at the marketing object specifically comprises the following steps:

4. The method of claim 3, the marketing result comprising: a target business index for reflecting the processing condition of the electronic coupons of the marketing objects by group members in the session group after the historical marketing information of the marketing objects is displayed in the session group; the method further comprises the steps of:

storing a corresponding relation between the target business index and the first marketing information style type;

and determining a second marketing information style type except the first marketing information style type corresponding to the historical dynamic marketing configuration information from preset marketing information style types, wherein the method specifically comprises the following steps of:

determining the second marketing information style type from the preset marketing information style types of the corresponding previous preset number with the largest historical service index according to the corresponding relation between the preset marketing information style type and the historical service index; wherein the historical service index is consistent with the type of the target service index; or,

5. The method of claim 3, wherein the number of categories of the second marketing message style type contained in the target dynamic marketing configuration message is greater than 1;

the generating, by using the first large language model, a target prompt word according to the target marketing configuration information specifically includes:

generating the target prompt words corresponding to any one of the second marketing information style types according to the target marketing configuration information carrying any one of the second marketing information style types by using the first large language model to obtain a target prompt word set;

the generating, by using the second large language model, the first marketing information aiming at the marketing object, which needs to be displayed at the session group, according to the target prompt word specifically includes:

6. The method of claim 5, further comprising:

7. The method of claim 2, the dynamic marketing configuration information comprising: marketing information delivery time;

and determining second marketing information throwing time except the first marketing information throwing time corresponding to the historical dynamic marketing configuration information, and obtaining target dynamic marketing configuration information aiming at the marketing object.

8. The method of claim 7, the marketing result comprising: a target business index for reflecting the processing condition of the electronic coupons of the marketing objects by group members in the session group after the historical marketing information of the marketing objects is displayed in the session group; the method further comprises the steps of:

Storing a corresponding relation between the target business index and the first marketing information throwing time;

the determining of the second marketing information delivery time except the first marketing information delivery time corresponding to the historical dynamic marketing configuration information specifically comprises the following steps:

9. The method of any one of claims 1-8, wherein the preset static marketing configuration information comprises: a target image carrying the marketing object; the method further comprises the steps of:

performing image main body extraction processing on the target image to obtain a marketing object image;

10. The method of claim 9, the first marketing message comprising: first marketing document information in the form of a poster document;

the generating, by using the multi-modal big model, a second marketing message for the marketing object to be displayed at the session group according to the marketing object image and the first marketing message, specifically includes:

11. The method of claim 10, the first marketing message further comprising: second marketing document information in the form of a group message; the method further comprises the steps of:

sending a group message carrying the second marketing document information to the terminal equipment of the group member in the session group;

12. The method of claim 1, the method further comprising:

acquiring target session information issued by target group members in the session group;

generating target reply information aiming at the target session information according to the target session information and at least part of the target marketing configuration information by using a third large language model;

13. The method of claim 12, wherein the preset static marketing configuration information comprises: at least one of a target image carrying the marketing object and a target descriptive text for the marketing object, and information for indicating use of an expression image related to the marketing object; the method further comprises the steps of:

generating an expression image related to the marketing object according to at least one of the target image and the target description text by using an image generation model;

The generating, by using a third large language model, target reply information for the target session information according to the target session information and at least part of the target marketing configuration information specifically includes:

generating initial reply information for the target session information according to the target session information and at least part of the target marketing configuration information by using the third large language model;

14. A marketing information generating apparatus comprising:

15. The apparatus of claim 14, the first acquisition module comprising:

the second acquisition unit is used for acquiring the preset static marketing configuration information set by the management user aiming at the marketing object; obtaining a marketing result of the history marketing information aiming at the marketing object, which is displayed in the conversation group last time; the historical marketing information is marketing information aiming at the marketing object according to the generation of the historical prompt words by utilizing the second large language model; the historical prompting words are prompting words generated by the first large language model according to the preset static marketing configuration information and the historical dynamic marketing configuration information;

16. The apparatus of claim 15, the dynamic marketing configuration information comprising: marketing information style;

the adjusting unit is specifically configured to:

17. The apparatus of claim 16, the marketing result comprising: a target business index for reflecting the processing condition of the electronic coupons of the marketing objects by group members in the session group after the historical marketing information of the marketing objects is displayed in the session group; the apparatus further comprises:

the first storage module is used for storing the corresponding relation between the target business index and the first marketing information style type;

the adjusting unit specifically comprises:

18. The apparatus of claim 16, wherein the number of categories of the second marketing message style type contained in the target dynamic marketing configuration message is greater than 1;

the first generation module is specifically configured to:

the second generating module is specifically configured to:

generating various first marketing information aiming at the marketing object according to various target prompt words in the target prompt word set by utilizing the second large language model; wherein the style types of the second marketing information corresponding to the different first marketing information are different;

the device further comprises:

19. The apparatus of claim 14, the dynamic marketing configuration information comprising: marketing information delivery time;

the adjusting unit is specifically configured to:

20. The apparatus of claim 19, the marketing result comprising: a target business index for reflecting the processing condition of the electronic coupons of the marketing objects by group members in the session group after the historical marketing information of the marketing objects is displayed in the session group; the apparatus further comprises:

the second storage module is used for storing the corresponding relation between the target business index and the first marketing information throwing time;

The adjusting unit specifically comprises:

21. The apparatus of any one of claims 14-20, wherein the preset static marketing configuration information comprises: a target image carrying the marketing object; the apparatus further comprises:

the image main body extraction module is used for carrying out image main body extraction processing on the target image to obtain a marketing object image;

22. The apparatus of claim 21, the first marketing message comprising: first marketing document information in the form of a poster message and second marketing document information in the form of a group message;

the third generating module is specifically configured to:

generating the second marketing information aiming at the marketing object to be displayed at the session group according to the marketing object image and the first marketing document information by utilizing the multi-mode large model;

the apparatus further comprises:

a second sending module, configured to send a group message carrying the second marketing document information to a terminal device of a group member in the session group;

23. The apparatus of claim 14, the apparatus further comprising:

the second acquisition module is used for acquiring target session information released by the target group member in the session group;

a fourth generation module, configured to generate target reply information for the target session information according to the target session information and at least part of the target marketing configuration information by using a third large language model;

24. The apparatus of claim 23, wherein the preset static marketing configuration information comprises: at least one of a target image carrying the marketing object and a target descriptive text for the marketing object, and information for indicating use of an expression image related to the marketing object; the apparatus further comprises:

a fifth generation module, configured to generate an expression image related to the marketing object according to at least one of the target image and the target description text by using an image generation model;

the fourth generation module includes:

an initial reply information generating unit, configured to generate initial reply information for the target session information according to the target session information and at least part of the target marketing configuration information by using the third large language model;

25. A marketing information generating device comprising:

at least one processor; the method comprises the steps of,

a memory communicatively coupled to the at least one processor; wherein,