CN118227870A - Method, device, medium and electronic device for information push - Google Patents

Abstract

The application relates to an information pushing method, an information pushing device, a medium and electronic equipment, wherein the method comprises the following steps: acquiring behavior data of a plurality of access users on an online platform of an enterprise, wherein the behavior data carries user identity identifiers respectively corresponding to the plurality of access users; obtaining a plurality of characteristic labels based on the behavior data, and constructing a label tree according to the characteristic labels; determining a user group with an ordering intention among the plurality of access users according to the tag tree; and sending the user identity identifiers corresponding to the access users in the user group to an operation background so that the operation background can respectively push information containing the online platform service of the enterprise to the access users corresponding to the user identity identifiers. The application can rapidly identify the user group with the ordering intention among a plurality of access users, and push the information containing the online platform business of the enterprise to the user group by the operation background, thereby improving the ordering intention of the users.

Description

Method, device, medium and electronic device for information push

Technical Field

The present application relates to the technical field of data processing, and in particular to a method, device, medium and electronic device for information push.

Background technique

In daily life, as the functions of the company's online platform continue to improve, the number of users visiting it continues to increase, and the user data generated is also increasing. The company's online platform operation team needs to conduct operational interventions on the corresponding users one by one based on the user data, such as sending text messages, making phone calls, etc., in order to enable users to finally place orders.

With respect to the above-mentioned related technologies, the inventors believe that there are the following defects: operational intervention is carried out on visiting users one by one, but this method lacks specificity, resulting in poor effect in improving users' intention to place orders.

Summary of the invention

In order to enhance the user's intention to place an order, the present application provides a method, device, medium and electronic device for information push.

In a first aspect of the present application, a method for information push is provided, which specifically includes:

Obtaining behavior data of multiple visiting users on the online platform of the enterprise, wherein the behavior data carries user identity identifiers corresponding to the multiple visiting users respectively;

Obtaining a plurality of feature tags based on the behavior data, and constructing a tag tree according to the plurality of feature tags;

Determine a user group having order placement intention among the multiple visiting users according to the tag tree;

The user identity corresponding to each visiting user in the user group is sent to the operation background, so that the operation background pushes the information containing the online platform business of the enterprise to the visiting users corresponding to the user identity.

By adopting the above technical solution, after obtaining the behavioral data of clicks, exposures, registrations, etc. of multiple visiting users on the online platform, multiple dimensions of user attributes and behavioral information in the online platform are extracted from the behavioral data to obtain multiple feature tags. A label tree reflecting the user attributes of the multiple visiting users is constructed based on the multiple feature tags. Then, based on the label values of the multiple visiting users under each feature tag in the label tree, the visiting users corresponding to the required label values are screened out, and these visiting users are identified as a user group with the intention to place an order. Finally, the user identity identifiers of the visiting users in the user group are sent to the enterprise's operation background, so that the operation background can push information to the visiting users with the intention to place an order in a targeted manner to enhance the users' intention to place an order.

Optionally, obtaining the behavior data of multiple access users on the enterprise's online platform includes:

Obtain the behavioral data of multiple visiting users on the company's online platform collected at several preset tracking points;

Identify user identity identifiers corresponding to the multiple access users respectively;

The user identity is added to the behavior data of the corresponding accessing user.

By adopting the above technical solution, tracking points are set for the clicks, exposures, consultations, searches and other behaviors of visiting users on the online platform. When the visiting user has such behaviors on the online platform, the tracking points will be triggered and various behavioral data of the visiting user will be collected. At the same time, when each visiting user visits the online platform, the user identity of each visiting user, such as the phone number, will be identified, and the behavioral data will be matched with the user identity one by one. Each collected behavioral data will carry the corresponding user identity, so that each behavioral data of each visiting user can be accurately obtained.

Optionally, obtaining a plurality of feature tags based on the behavior data and constructing a tag tree according to the plurality of feature tags includes:

Inputting the behavior data as input parameters into a decision tree algorithm to obtain a plurality of feature labels;

Perform hierarchical clustering on multiple feature labels to obtain hierarchical clustering results;

A label tree of the visiting users is constructed based on the hierarchical clustering result.

By adopting the above technical solution, after obtaining the behavioral data of multiple visiting users, the behavioral data is input into the decision tree algorithm. The decision tree algorithm can accurately classify the behavioral data of multiple visiting users, thereby obtaining multiple feature labels. Then, the multiple feature labels are subjected to hierarchical clustering analysis and hierarchical decomposition to obtain a label tree containing the label value of each visiting user under the feature label, thereby better outlining the full-link user portraits of multiple visiting users, and also facilitating the subsequent identification of user groups with the intention of placing an order.

Optionally, determining a user group having an intention to place an order among the multiple visiting users according to the tag tree includes:

Obtaining tag values of the plurality of feature tags in the tag tree;

Filter out a target label value from each of the label values;

Combining the target tag values to obtain a target screening tag value;

Target behavior data is filtered out from the behavior data according to the target filtering tag value, and visiting users corresponding to the target behavior data are determined as a user group with an intention to place an order.

By adopting the above scheme, after the label tree is determined, the label values under multiple feature labels in the label tree are selected as the screening objects, and then the required target label values are screened out from the screening objects, that is, the label values required to select the user group with the intention to place an order from multiple visiting users. The screened target label values are combined by intersection and difference to obtain the target screening label value, that is, the final screening condition. Finally, according to the target screening label value, the visiting users corresponding to the final screening condition are screened out from the multiple visiting users to obtain the user group with the intention to place an order, so that the visiting users with the intention to place an order can be identified more quickly from the multiple visiting users who visit the online platform.

Optionally, filtering out a target label value from each of the label values includes:

Obtaining the identity of offline customers of the enterprise;

Determining whether there is an identity identifier of the offline customer among the user identity identifiers corresponding to the multiple visiting users;

If the identity identifier of the offline customer exists among the user identity identifiers corresponding to the multiple visiting users, searching the tag value of the feature tag corresponding to the identity identifier from the tag tree;

The tag value of the characteristic tag corresponding to the identity identifier is filtered out from the tag values, and the filtered tag value is used as the target tag value.

By adopting the above technical solution, the identity of the offline customer of the enterprise is compared with the user identity of multiple visiting users. If there is a match, it means that such offline customers have visited the online platform while already being a customer of the enterprise, which further indicates that such offline customers have a greater intention to place an order. Then, the label values of each feature label corresponding to the identity of this offline customer are found in the label tree, and the label values under multiple feature labels in the label tree are selected as the target label value, so that the user group with the intention to place an order can be more accurately selected by using this target label value combination.

Optionally, after determining a user group having order placement intention among the multiple visiting users according to the tag tree, the method further includes:

Obtaining a tag value of each visiting user in the user group under the multiple feature tags, and determining feature distribution information of the user group according to each of the tag values;

Comparing the characteristic distribution information of the user group with the characteristic distribution information of the plurality of visiting users to obtain a comparison result;

If the comparison result does not meet the preset requirements, the user group is re-determined.

By adopting the above technical solution, after determining the user group, the characteristic distribution information of the user group is determined according to the label values corresponding to multiple characteristic labels of each visiting user in the user group, and the characteristic distribution information of the user group is compared with the characteristic distribution information of all visiting users on the online platform. If the proportion of visiting users with a certain label value in the user group is less than the proportion of visiting users with a certain label value among all visiting users, the comparison result does not meet the preset requirements, indicating that the reference value of the user group with the intention to place an order is low, then the user group is re-defined so that the strategy of operating target customers can be adjusted in time.

Optionally, the method further includes:

The user identity of each access user in the user group is updated according to a preset update frequency to redefine the user group.

By adopting the above technical solution and setting a fixed update frequency, the user group selected according to a certain screening condition can automatically update the underlying user identity identification according to the update frequency. That is, the user group can be automatically defined according to the screening condition and the update frequency without manual definition, thereby greatly improving the efficiency of online platform operations.

In a second aspect of the present application, a device for information push is provided, which specifically includes:

A behavior data acquisition module, used to acquire behavior data of multiple visiting users on the online platform of the enterprise, wherein the behavior data carries user identity identifiers corresponding to the multiple visiting users respectively;

A label tree construction module, used to obtain a plurality of feature labels based on the behavior data, and to construct a label tree according to the plurality of feature labels;

A user group defining module, for determining a user group having an intention to place an order among the multiple visiting users according to the tag tree;

The information push module is used to send the user identity corresponding to each visiting user in the user group to the operation background, so that the operation background can push the information containing the online platform business of the enterprise to the visiting users corresponding to the user identity.

By adopting the above technical solution, the behavior data acquisition module obtains the behavior data of multiple visiting users on the enterprise's online platform, and the behavior data carries the user identity corresponding to the visiting user. The label tree construction module extracts multiple feature labels from the acquired behavior data and constructs a label tree based on the multiple feature labels. Then, the user group delineation module selects the user group with the intention to place an order from the multiple visiting users involved in the label tree. Finally, the information push module pushes the user group to the operation background of the online platform, so that the operation background can send push information to each visiting user in the user group according to the user identity corresponding to the visiting user, thereby increasing the user's intention to place an order.

In summary, the present application includes at least one of the following beneficial technical effects:

After obtaining the behavioral data of clicks, exposures, registrations, etc. of multiple visiting users on the online platform, extract user attributes of multiple dimensions and behavioral information in the online platform from the behavioral data to obtain multiple feature tags, and construct a label tree reflecting the user attributes of the multiple visiting users based on the multiple feature tags. Then, based on the label values of the multiple visiting users under each feature tag in the label tree, filter out the visiting users corresponding to the required label values and determine them as the user group with the intention to place an order. Finally, send the user identity identifiers of the visiting users in the user group to the enterprise's operation background, so that the operation background can push text messages to the visiting users with the intention to place an order in a targeted manner to enhance the users' intention to place an order.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the architecture of an information push system provided in an embodiment of the present application;

FIG2 is a flow chart of a method for information push provided in an embodiment of the present application;

FIG3 is a schematic diagram of the relationship between a feature tag and a tag value provided in an embodiment of the present application;

FIG4 is a flow chart of another method for information push provided in an embodiment of the present application;

FIG5 is a flow chart of another method for information push provided in an embodiment of the present application;

FIG6 is a schematic diagram of a target tag value combination provided in an embodiment of the present application;

FIG7 is a schematic diagram of another target tag value combination provided in an embodiment of the present application;

FIG8 is a schematic diagram of feature distribution information provided in an embodiment of the present application;

FIG9 is a schematic diagram of the structure of an information push device provided in an embodiment of the present application;

FIG. 10 is a schematic diagram of the structure of another information push device provided in an embodiment of the present application.

Explanation of the accompanying drawings: 11. Behavior data acquisition module; 12. Tag tree construction module; 13. User group identification module; 14. Information push module.

Detailed ways

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

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

In the description of the embodiments of the present application, the term "and/or" is only a kind of association relationship describing the associated objects, indicating that there may be three kinds of relationships, for example, A and/or B, which can represent: A exists alone, B exists alone, and A and B exist at the same time. In addition, unless otherwise specified, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" can explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

Referring to FIG. 1 , an embodiment of the present application discloses a flow chart of a system for information push, which specifically includes a terminal for accessing a user, a server, and an operation background, wherein the terminal for accessing the user is a smart phone.

Specifically, the visiting user accesses the online platform through the terminal, and the server obtains the behavior data of multiple visiting users by tracking the user behavior in the online platform, analyzes the behavior data to extract multiple feature tags, and builds a tag tree based on the feature tags. Then, based on the tag tree, the user group with the intention to place an order is screened out from the multiple visiting users, and the user identity and corresponding tag value of each visiting user in the user group are sent to the operation background, so that the operation background pushes information containing the company's online platform business to each visiting user in the user group.

Referring to FIG. 2 , an embodiment of the present application discloses a flowchart of a method for information push, which can be implemented by a computer program or run on an information push device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application, specifically including:

S101: Obtaining behavior data of multiple visiting users on an online platform of an enterprise, wherein the behavior data carries user identity identifiers corresponding to the multiple visiting users.

Specifically, whether the visiting user accesses the company's online platform through a PC or a mobile terminal, a variety of user behaviors will be generated during the visit. These user behaviors are integrated into data, that is, the behavioral data corresponding to the visiting user is formed. Among them, the company's online platform can be a website, or an app product or a small program. User behavior includes a series of behaviors such as exposure, clicks, consultations, searches, registration and login. Exposure refers to the business information of the company's online platform, and the number of times it is seen by the visiting user in the search results list and the list page for browsing by category. Concepts such as clicks, consultations, searches, and registration and login are relatively common and will not be repeated here. The behavioral data corresponding to user behavior mainly includes the basic information of the visiting user (gender, age, occupation, etc.), the keywords searched by the visiting user, the online time of the visiting user, the content browsed by the user for a long time, the activity of the visiting user within 7 days, etc.

The method of obtaining the behavior data of multiple visiting users on the online platform of the enterprise is to perform tracking analysis on the specific user behaviors of the visiting users on the online platform. When the visiting user generates a specific user behavior, the tracking point will be triggered to capture the user behavior of the visiting user and obtain the behavior data of the visiting user. In addition, each collected behavior data will carry the user identity corresponding to the visiting user.

It should be noted that each access user corresponds to a unique user identity, which is the access user's phone number in the embodiment of the present application, and may also be an email address, ID card information, etc. The user identity can be used to quickly identify whether the access user of the enterprise's online platform is the same person as the enterprise's offline customer.

S102: Obtain multiple feature tags based on the behavior data, and construct a tag tree according to the multiple feature tags.

Specifically, a label refers to the abstract extraction and summary of a common feature of multiple visiting users. Feature labels are user behavior data labels, which are differentiated summary words obtained by analyzing the behavior data generated by visiting users on the online platform. Feature labels include three types: user labels, enterprise labels, and content labels. User labels are highly refined feature identifiers obtained by analyzing the user information of visiting users, and are also key factors in forming user portraits. Among them, user portraits refer to labeled user models abstracted from user attributes, user preferences, etc. Enterprise labels are feature identifiers of the enterprise information to which the visiting user belongs. Content labels are feature descriptions of text, graphics, and other content in the form of keywords or phrases. The three types of feature labels cover basic user attributes, channel sources, behavioral preferences, conversion information, and detailed data labels.

The label tree is a tree-structured data set obtained by hierarchically classifying the behavior data corresponding to the feature labels. After obtaining the behavior data of multiple visiting users, the decision tree algorithm can be used to extract multiple feature labels from the numerous behavior data. Each visiting user corresponds to at least one feature label. Then, the hierarchical clustering algorithm is used to hierarchically decompose the collection of behavior data according to the feature labels, and finally obtain the label tree of multiple visiting users.

S103: Determine a user group with order placement intention among multiple visiting users according to the tag tree.

Specifically, each feature tag corresponds to a different tag value. Among them, the tag value is an attribute of the tag, which is used to represent the specific tag content. As shown in Figure 3, a schematic diagram of the relationship between a feature tag and a tag value provided in an embodiment of the present application is shown. For example, feature tag A is "Number of logins in the past 30 days", then the tag values under this feature tag are: 0 times, 1-5 times, 6-10 times and more than 10 times. For another example, if feature tag B is "User life cycle", then the tag values under feature tag B are: New users, active users, dormant users and lost users.

After the label tree is determined, all label values under multiple feature labels in the label tree are selected, and then external input filtering conditions are received, such as exact match or fuzzy match, greater than or less than, included or not included, etc. After a series of label values are filtered out, the filtered label values are combined in the intersection and difference method to obtain a label value combination that is easier to place an order. Finally, the visiting users who meet the requirements are filtered out according to the label value combination and determined as the user group with the intention of placing an order.

For example, if the screening conditions are that the number of logins in the past 30 days is greater than 5 times, and the user life cycle does not include lost users, the filtered label values are 6-10 times, more than 10 times, new users, active users, and dormant users. If it is determined that the number of logins in 30 days must be 6-10 times and the user must be a new user, then the user group that meets both of these conditions will be screened out from multiple visiting users.

In one practicable manner, the user identity of each access user in the user group is updated according to a preset update frequency to redefine the user group.

Specifically, because the number of users visiting the online platform varies every day, and the user behaviors within the online platform are also different, the visiting users in the defined user group are not fixed. If promotional marketing is conducted on the group of people selected based on a certain screening condition to increase order intention, the operator needs to manually select all of them every day, and it cannot be updated automatically, resulting in reduced operational efficiency. The preset update frequency automatically and regularly re-defines the user group according to the update frequency, and updates the user identity identifiers in the user group, which can effectively save operating costs. It should be noted that the update frequency can be 1 day or 12 hours.

S104: Send the user identity corresponding to each visiting user in the user group to the operation background, so that the operation background pushes information including the online platform business of the enterprise to the visiting users corresponding to the user identity.

Specifically, the operation backend is an operating platform used by internal personnel of an enterprise, serving the operation activities of online platforms (websites, apps or mini-programs). After the user group with the intention to place an order is determined, each visiting user in the user group is used as the target user for operation, and the user identity corresponding to each visiting user in the user group is sent to the operation backend, so that the operation backend sends information containing the online platform business of the enterprise to the visiting users with the intention to place an order according to each user identity, so that after quickly identifying the visiting users with the intention to place an order, the user's intention to place an order can be improved through corresponding operation intervention, and even the visiting users can be prompted to place an order.

In one feasible method, an externally input user identity, such as a telephone number, is received. Based on the input user identity, each feature tag corresponding to the user identity and the tag value corresponding to each feature tag can be found from the constructed tag tree, and the user portrait composed of multiple tag values can be displayed, so that personnel can easily query the full-link user portrait of each visiting user, and then formulate customized operation strategies.

Referring to FIG4 , the present application embodiment discloses a flowchart of another method for information push, which can be implemented by a computer program or run on an information push device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application, specifically including:

S201: Obtaining behavioral data of multiple visiting users on the enterprise's online platform collected at a number of preset embedding points.

Specifically, the embedding point is to implant statistical codes at the corresponding positions of the front-end to record the user's behavior. In an embodiment of the present application, the embedding point is set for each behavior of the visiting user in the online platform of the enterprise through the Sensors platform. In other embodiments, the embedding point can also be set through the GrowingIO platform. For example, the embedding point is set for the visiting user's click behavior through the Sensors platform. When the user clicks, refreshes, touches, and other active operations on the company's online platform through a computer or smart phone, the embedding point will be triggered and the recorded behavior data will be sent to the server. For another example, the exposure is embedded. When the data of the online platform with the embedding point setting is displayed, the behavior data is collected immediately. Specifically, when a page on the online platform is being browsed by the visiting user, that is, the page is exposed, the embedding point is triggered to collect the visiting user's behavior data on the exposure.

S202: Identify user identity tags corresponding to multiple access users.

S203: Add the user identity to the behavior data of the corresponding accessing user.

Specifically, a tracking point is set for the registration and login behavior of the visiting user. When the visiting user performs real-name registration and login operations on the online platform, the tracking point is triggered to capture the user's behavior in real time and collect the corresponding behavior data of the visiting user, that is, the user identity (telephone number) of the visiting user, so as to identify the user identity corresponding to each visiting user. For example, a tracking point is set on the registration and login interface of the online platform. When the visiting user clicks the "Register" or "Login" button, the tracking point is triggered to obtain the user identity of the visiting user. After the visiting user logs in to the online platform, various behaviors (clicks, exposure, searches, etc.) generated on the online platform are triggered by the corresponding behavior tracking points. The corresponding user identity will be added to the collected behavior data of each visiting user, so that each behavior data can be corresponding to the user identity.

S204: Input the behavior data as input parameters into the decision tree algorithm to obtain multiple feature labels.

Specifically, the decision tree is a basic classification and regression method. The decision tree model has a tree structure. In the classification problem, it represents the process of classifying instances based on features. It can be considered as a collection of if-then rules, or it can be considered as a conditional probability distribution defined in the feature space and the class space. Its main advantages are that the model is readable and the classification speed is fast. When learning, the training data is used to establish a decision tree model based on the principle of minimizing the loss function. When predicting, the new data is classified using the decision tree model. This is a prior art and will not be described here. The behavior data of multiple visiting users is input into the decision tree algorithm so that the behavior data is analyzed and classified, and the behavior data is classified into multiple dimensions, that is, multiple feature labels.

It should be noted that in other embodiments, the behavior data may also be classified according to preset rules. For example, the rules for user activity level are defined. If the number of logins is a times a day, the corresponding user activity level is A; if the number of logins is b times a day, the corresponding user activity level is B. Thus, characteristic labels of user activity levels are attached to multiple visiting users. In this way, multiple characteristic labels corresponding to the behavior data are further determined.

S205: Perform hierarchical clustering on the multiple feature labels to obtain a hierarchical clustering result.

S206: Construct a tag tree of visiting users based on the hierarchical clustering result.

Specifically, hierarchical clustering is to perform hierarchical decomposition on the set of each given data object. According to the decomposition strategy adopted by the hierarchical decomposition, hierarchical clustering can be divided into agglomerative and divisive hierarchical clustering. Each feature label corresponds to a part of the behavior data. In the embodiment of the present application, hierarchical clustering is used to perform hierarchical decomposition on multiple feature labels. In fact, hierarchical decomposition is performed on the behavior data of multiple visiting users to obtain hierarchical clustering results, that is, the results of hierarchical decomposition. Then, a label tree is generated according to the hierarchical clustering results, wherein each layer of branches in the label tree corresponds to a feature label, and the label value under each feature label corresponds to the leaf node on each layer of branches. This is a prior art and will not be repeated here.

S207: Determine a user group with order placement intention among multiple visiting users according to the tag tree.

S208: Send the user identity corresponding to each visiting user in the user group to the operation background, so that the operation background pushes information including the online platform business of the enterprise to the visiting users corresponding to the user identity.

For details, please refer to steps S103-S104, which will not be described in detail here.

Referring to FIG5 , the present application embodiment discloses a flowchart of another method for information push, which can be implemented by a computer program or run on an information push device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application, specifically including:

S301: Obtaining behavior data of multiple visiting users on the enterprise's online platform, where the behavior data carries user identity identifiers corresponding to the multiple visiting users.

S302: Obtain multiple feature tags based on the behavior data, and construct a tag tree according to the multiple feature tags.

For details, please refer to steps S101-S102, which will not be described in detail here.

S303: Obtain tag values of multiple feature tags in the tag tree.

Specifically, after the tag tree is constructed, the method of obtaining DOM elements (JavaScript, JS) is used to first obtain multiple feature tags in the tag tree, and then the tag value under each feature tag is obtained through value. Among them, JS is a lightweight, interpreted or just-in-time compiled programming language with function priority. For example, the tag tree has two feature tags, user preference and gender. Through this method, the tag values under the "user preference" feature tag are obtained as: preference A, preference B; the tag values under the "gender" feature tag are: male, female.

S304: Filter out a target label value from each label value.

Obtain the identity of offline customers of the enterprise;

Determine whether there is an offline customer's identity identifier among the user identity identifiers corresponding to the multiple access users;

If the user identities corresponding to multiple access users include an offline customer's identity, then the tag value of the feature tag corresponding to the identity is searched from the tag tree;

The label value of the characteristic label corresponding to the identity identifier is filtered out from each label value, and the filtered label value is used as the target label value.

Specifically, after determining the corresponding tag values under multiple feature tags in the tag tree, the company's offline customer identity, that is, the offline customer's phone number, is obtained through the company's Customer Relationship Management (CRM) system. Among them, the CRM system refers to an information system that uses software, hardware and network technology to establish a customer information collection, management, analysis and utilization system for the company. With customer data management as the core, it records various interactive behaviors between the company and customers during the marketing and sales process, as well as the status of various related activities, and provides various data models to support subsequent analysis and decision-making.

After obtaining the identity of the company's offline customers from the CRM system, the identity of the offline customers is compared with the user identity of multiple visiting users of the company's online platform one by one. If the comparison is consistent, it means that there are offline customers of the company among the multiple visiting users of the online platform, and further indicates that the offline customers have a strong intention to place an order. Therefore, the label values of the offline customers under multiple feature labels in the label tree are filtered out as the target label values. For example, the feature labels involved in the behavioral data of offline customers visiting the online platform are: user preference and activity. The label values under the feature labels are preference B and activity exceeds C, so preference B and activity exceeds C are used as target label values.

It should be noted that in other embodiments, externally input label values may also be accepted. These externally input label values may be label values that operators determine based on experience to be easier to order, and these may be directly used as target label values.

S305: Combine the target tag values to obtain a target screening tag value.

S306: Filter out target behavior data from the behavior data according to the target filtering tag value, and determine the visiting users corresponding to the target behavior data as the user group with the intention to place an order.

Specifically, after the target tag value is determined, the intersection and/or union and/or difference between the behavior data corresponding to each target tag value is taken to obtain the target screening tag value, that is, the final screening condition for defining the user group. As shown in Figure 6, for example, if the target tag value is activity exceeding C and preference B, if you want to filter out visiting users who meet both activity exceeding C and preference B, then the two target tag values of activity exceeding C and preference B are processed by intersection; as shown in Figure 7, if you want to filter out visiting users who meet activity exceeding C and preference B, then the two target tag values are processed by union.

After the target filtering tag value is determined, the target behavior data that meets the target filtering tag value will be filtered out from the behavior data of multiple visiting users, because the target behavior data all carry the user identity of the corresponding visiting user, and then the filtered user identity is determined, and finally the visiting users corresponding to each user identity are determined as the user group with the intention to place an order.

S307: Obtain the tag value of each visiting user in the user group under multiple feature tags, and determine the feature distribution information of the user group according to each tag value.

Specifically, after determining the user group with the intention to place an order according to the target filtering label value, each visiting user in the user group corresponds to a series of behavioral data, corresponding to label values under multiple feature labels, not just the target filtering label value. According to the label values of each visiting user, the feature distribution information of each visiting user in the user group is displayed in the form of an icon. For example, the visiting users with the label value of "prefer B" account for 3%, and the visiting users with the label value of "activeness exceeds A" account for 5%. The distribution of visiting users with label values under each feature label can be better presented through the feature distribution information of the user group.

S308: Compare the characteristic distribution information of the user group with the characteristic distribution information of multiple visiting users to obtain a comparison result.

S309: If the comparison result does not meet the preset requirements, redefine the user group.

Specifically, after the characteristic distribution information of the user group is determined, the characteristic distribution information of all visiting users on the online platform is determined in the same way, and the characteristic distribution information of the user group is compared with the characteristic distribution information of all visiting users. For example, as shown in FIG8, a schematic diagram of characteristic distribution information provided by an embodiment of the present application is shown. According to the target screening label value ("income exceeds D" and "preference B"), the number of visiting users with the label value of "active on the platform in the past 7 days" in the user group screened out accounts for 3% of the number of visiting users in the user group, while the number of visiting users with the label value of "active on the platform in the past 7 days" among all visiting users on the online platform accounts for 5% of all visiting users on the online platform. 3% is less than 5% (i.e., the comparison result), indicating that there are many inactive visiting users in the user group circled this time, and the characteristic distribution of the user group is not as good as the characteristic distribution of all visiting users. There may be fewer visiting users in the user group who intend to place an order, which does not meet the requirements for the circle of visiting users who intend to place an order. Therefore, it is necessary to redo the user group and redefine the target customers for operation.

S310: Send the user identity corresponding to each visiting user in the user group to the operation background, so that the operation background pushes information including the online platform business of the enterprise to the visiting users corresponding to the user identity.

For details, please refer to step S104, which will not be described in detail here.

The implementation principle of the information push method of the embodiment of the present application is: the behavior data of multiple visiting users on the online platform of the enterprise is obtained through the embedding point, and each behavior data obtained will carry the user identity corresponding to the visiting user. A decision tree algorithm is used to extract multiple feature tags from the behavior data, and a tag tree is constructed based on the multiple feature tags. Then, a user group with order intention is determined among the multiple visiting users based on the tag tree, and finally each visiting user in the user group is sent to the operation background of the enterprise, so that the operation background can send information containing the online platform business of the enterprise to the visiting user, thereby improving the user's order intention.

The following are device embodiments of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

Please refer to Figure 9, which is a schematic diagram of the structure of an information push device provided in an embodiment of the present application. The device applied to information push can be implemented as all or part of the device through software, hardware, or a combination of both. The device 1 includes a behavior data acquisition module 11, a tag tree construction module 12, a user group delineation module 13, and an information push module 14.

The behavior data acquisition module 11 is used to acquire the behavior data of multiple visiting users on the online platform of the enterprise, and the behavior data carries the user identity identifiers corresponding to the multiple visiting users respectively;

A label tree construction module 12 is used to obtain multiple feature labels based on the behavior data and construct a label tree according to the multiple feature labels;

A user group identification module 13, used to identify a user group with order intention among multiple visiting users according to the tag tree;

The information push module 14 is used to send the user identity corresponding to each visiting user in the user group to the operation background, so that the operation background pushes the information containing the online platform business of the enterprise to the visiting users corresponding to the user identity.

Optionally, the behavior data acquisition module 11 is specifically used for:

Obtain the behavioral data of multiple visiting users on the company's online platform collected at a number of preset tracking points;

Identify user identities corresponding to multiple access users;

Add the user identity to the corresponding access user's behavior data.

Optionally, the tag tree construction module 12 is specifically used for:

The behavioral data is input as input parameters to the decision tree algorithm to obtain multiple feature labels;

Perform hierarchical clustering on multiple feature labels to obtain hierarchical clustering results;

Construct a tag tree of visiting users based on the hierarchical clustering results.

Optionally, the user group defining module 13 is specifically used for:

Get the label values of multiple feature labels in the label tree;

Filter the target label value from each label value;

Combine the target label values to obtain the target filter label value;

Target behavior data is filtered out from the behavior data according to the target filtering tag value, and visiting users corresponding to the target behavior data are determined as the user group with the intention to place an order.

Optionally, the user group defining module 13 is further used for:

Obtain the identity of offline customers of the enterprise;

Determine whether there is an offline customer's identity identifier among the user identity identifiers corresponding to the multiple access users;

If the user identities corresponding to multiple access users include an offline customer's identity, then the tag value of the feature tag corresponding to the identity is searched from the tag tree;

The label value of the characteristic label corresponding to the identity identifier is filtered out from each label value, and the filtered label value is used as the target label value.

Optionally, as shown in FIG10 , the device 1 further includes a feature distribution analysis module 15, which is specifically used for:

Obtain the label value of each visiting user in the user group under multiple feature labels, and determine the feature distribution information of the user group according to each label value;

Comparing the characteristic distribution information of the user group with the characteristic distribution information of multiple visiting users to obtain a comparison result;

If the comparison result does not meet the preset requirements, the user group is re-determined.

Optionally, the device 1 further comprises:

The automatic updating module 16 is used to update the user identity of each access user in the user group according to a preset updating frequency to redefine the user group.

It should be noted that, when executing the information push method, the device for information push provided in the above embodiment only uses the division of the above functional modules as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device for information push provided in the above embodiment and the method embodiment for information push belong to the same concept, and the implementation process thereof is detailed in the method embodiment, which will not be repeated here.

An embodiment of the present application further discloses a computer-readable storage medium, and the computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, a method for pushing information in the above embodiment is adopted.

Among them, the computer program can be stored in a computer-readable medium, the computer program includes computer program code, the computer program code can be in the form of source code, object code, executable file or certain middleware, etc. The computer-readable medium includes any entity or device that can carry the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the computer-readable medium includes but is not limited to the above-mentioned components.

Among them, through this computer-readable storage medium, a method for information push of the above embodiment is stored in a computer-readable storage medium, and is loaded and executed on a processor to facilitate the storage and application of the above method.

An embodiment of the present application further discloses an electronic device, wherein a computer program is stored in a computer-readable storage medium, and when the computer program is loaded and executed by a processor, the above-mentioned information push method is adopted.

The electronic device may be a desktop computer, a laptop computer, a cloud server or other electronic device, and the electronic device includes but is not limited to a processor and a memory. For example, the electronic device may also include input and output devices, a network access device, and a bus.

Among them, the processor can adopt a central processing unit (CPU). Of course, according to actual usage, other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. can also be adopted. The general-purpose processor can adopt a microprocessor or any conventional processor, etc., and this application does not impose any restrictions on this.

Among them, the memory can be an internal storage unit of the electronic device, such as a hard disk or memory of the electronic device, or it can be an external storage device of the electronic device, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD) or flash memory card (FC) equipped on the electronic device, and the memory can also be a combination of an internal storage unit and an external storage device of the electronic device. The memory is used to store computer programs and other programs and data required by the electronic device. The memory can also be used to temporarily store data that has been output or is to be output, and this application does not impose any restrictions on this.

Among them, through this electronic device, a method for pushing information in the above embodiment is stored in the memory of the electronic device, and is loaded and executed on the processor of the electronic device for easy use.

The above is only an exemplary embodiment of the present disclosure, and the scope of the present disclosure cannot be limited thereto. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the specification and practicing the disclosure here, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the technical field not recorded in the present disclosure. The description and examples are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims (10)

1. A method for information push, characterized in that the method comprises: Obtaining behavior data of multiple visiting users on the online platform of the enterprise, wherein the behavior data carries user identity identifiers corresponding to the multiple visiting users respectively; Obtaining a plurality of feature tags based on the behavior data, and constructing a tag tree according to the plurality of feature tags; Determine a user group having order placement intention among the multiple visiting users according to the tag tree; The user identity corresponding to each visiting user in the user group is sent to the operation background, so that the operation background pushes the information containing the online platform business of the enterprise to the visiting users corresponding to the user identity. 2. The information push method according to claim 1, characterized in that the step of obtaining the behavior data of multiple visiting users on the enterprise's online platform comprises: Obtain the behavioral data of multiple visiting users on the company's online platform collected at several preset tracking points; Identify user identity identifiers corresponding to the multiple access users respectively; The user identity is added to the behavior data of the corresponding accessing user. 3. The information push method according to claim 1, characterized in that the step of obtaining a plurality of feature tags based on the behavior data and constructing a tag tree according to the plurality of feature tags comprises: Inputting the behavior data as input parameters into a decision tree algorithm to obtain a plurality of feature labels; Perform hierarchical clustering on multiple feature labels to obtain hierarchical clustering results; A label tree of the visiting users is constructed based on the hierarchical clustering result. 4. The information push method according to claim 1, characterized in that the step of determining a user group having order intention among the multiple visiting users according to the tag tree comprises: Obtaining label values of the plurality of feature labels in the label tree; Filter out a target label value from each of the label values; Combining the target tag values to obtain a target screening tag value; Target behavior data is filtered out from the behavior data according to the target filtering tag value, and visiting users corresponding to the target behavior data are determined as a user group with an intention to place an order. 5. The information push method according to claim 4, wherein the step of filtering out a target label value from each of the label values comprises: Obtaining the identity of offline customers of the enterprise; Determining whether there is an identity identifier of the offline customer among the user identity identifiers corresponding to the multiple visiting users; If the identity identifier of the offline customer exists among the user identity identifiers corresponding to the multiple visiting users, searching the tag value of the feature tag corresponding to the identity identifier from the tag tree; The tag value of the characteristic tag corresponding to the identity identifier is filtered out from the tag values, and the filtered tag value is used as the target tag value. 6. The information push method according to claim 1, characterized in that after determining the user group with order intention among the multiple visiting users according to the tag tree, it also includes: Obtaining a tag value of each visiting user in the user group under the multiple feature tags, and determining feature distribution information of the user group according to each of the tag values; Comparing the characteristic distribution information of the user group with the characteristic distribution information of the plurality of visiting users to obtain a comparison result; If the comparison result does not meet the preset requirements, the user group is re-determined. 7. The information push method according to claim 1, characterized in that the method further comprises: The user identity of each access user in the user group is updated according to a preset update frequency to redefine the user group. 8. An information push device, characterized by comprising: A behavior data acquisition module (11) is used to acquire behavior data of multiple visiting users on the online platform of the enterprise, wherein the behavior data carries user identity identifiers corresponding to the multiple visiting users respectively; A label tree construction module (12), used to obtain a plurality of feature labels based on the behavior data, and to construct a label tree according to the plurality of feature labels; A user group defining module (13), used for determining a user group having an intention to place an order among the plurality of visiting users according to the tag tree; The information push module (14) is used to send the user identity corresponding to each visiting user in the user group to the operation background, so that the operation background pushes the information containing the online platform business of the enterprise to the visiting users corresponding to the user identity. 9. A computer-readable storage medium storing a computer program, wherein when the computer program is loaded and executed by a processor, the method according to any one of claims 1 to 7 is adopted. 10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor adopts the method according to any one of claims 1 to 7 when loading and executing the computer program.