Lottery drawing method based on multi-dimensional portrait label of platform user

Abstract

The invention relates to the field of lottery drawing, in particular to a lottery drawing method based on platform user multi-dimensional portrait labels. The method comprises two parts of early-stage data preparation and later-stage lottery drawing. According to the data preparation part, platform user multi-dimensional portrait labels are established, lottery drawing rules are formulated according to label grouping, and corresponding prizes and winning probabilities are configured; and in the lottery drawing part, a winning rule is matched through the lottery drawing user group, and then a winning result is calculated. According to the method, a multi-dimensional and refined portrait label is created for each user. Personalized lottery drawing rules are made according to the portrait labels of the users, and prize types, prize numbers and corresponding prize winning probabilities are configured. The lottery drawing method is more diversified, the user experience is improved, and the credibility and attraction of the activity are also enhanced.

Description

Technical Field

[0001] The present invention relates to the field of lottery, and in particular to a lottery method based on multi-dimensional portrait tags of platform users. Background Art

[0002] During promotional activities held by the platform, in order to enhance the user's sense of participation and the fun of the activities, a lucky draw session is usually introduced. This measure can not only effectively improve user stickiness, but also stimulate user enthusiasm through the reward mechanism. However, how to reasonably set the lucky draw probability has become a crucial task in platform operations. The following are several common ways to set the lucky draw probability.

[0003] 1. Equal probability:

[0004] In this setting, each prize has an equal chance of being drawn. This is usually suitable for situations where the prizes are of similar value or you want to give participants a sense of fairness. The disadvantage of this lottery method is obvious. If the prizes have a large difference in value, equal probability may cause users to feel unfair because users who get prizes of lower value may be disappointed.

[0005] 2. Setting of graded probability:

[0006] Prizes are divided into different levels according to their value or rarity, and different winning probabilities are set for each level of prizes. Generally speaking, prizes with higher value have lower winning probabilities.

[0007] For example, the probability of winning the first prize can be set at 1%, the second prize at 5%, the third prize at 10%, and so on. The disadvantages are also obvious. Since the probability of winning high-value prizes is low, users may feel hopeless about winning, thus reducing their enthusiasm for participation. At the same time, if the setting is too biased, it may cause users to question the fairness of the event.

[0008] 3. Progressive probability setting:

[0009] As the number of draws increases, the probability of winning gradually increases. This method can motivate participants to participate in the draw multiple times and increase the stickiness of the activity. For example, the initial probability of winning is 5%, but after each failure, the probability of winning the next time will increase slightly. It seems reasonable, but there are actually problems. As the number of draws increases, the increase in the probability of winning may cause the system burden to increase, especially when the number of users is large. In addition, if the initial probability is set too low, users may lose interest before the probability increases to an attractive level.

[0010] 4. Guaranteed winning settings:

[0011] Set the probability of winning one or more prizes to 100% to ensure that each participant can get at least one prize. This is often used to increase user participation and satisfaction. However, this method requires a large amount of prize reserves, which is costly. If the prize reserves are insufficient, the event may end early or cause user dissatisfaction. At the same time, if the guaranteed winning prize is low in value, it may not effectively motivate users to participate.

[0012] 5. Conditional probability setting:

[0013] The probability of winning is adjusted according to certain specific conditions of the participants (such as consumption amount, membership level, etc.). This method can achieve incentives or feedback for specific user groups.

[0014] For example, a premium member's chance of winning a prize may be twice as high as that of a regular member. This approach may trigger a sense of unfairness among users, especially when the conditions are significantly different (such as premium members vs. regular members). In addition, if the conditions are set too complex or opaque, it may cause users to feel confused or dissatisfied.

[0015] 6. Time period probability setting:

[0016] Set different winning probabilities in different time periods. This can be used to guide users to participate in activities during specific time periods, thereby balancing traffic or increasing the exposure of activities. For example, lower the winning probability during the peak period of the activity, and increase the winning probability during the trough period. The time period probability setting may cause users to participate in activities during non-peak hours, thereby reducing the overall popularity and exposure of the activity. At the same time, if the time period setting is unreasonable or the publicity is not in place, it may cause users to miss the high probability period, and then become dissatisfied with the activity. Summary of the invention

[0017] In view of the problems existing in the background technology, a lottery method based on multi-dimensional portrait tags of platform users is proposed. Each user creates a multi-dimensional and refined portrait tag, formulates personalized lottery rules according to the user's portrait tag, and configures the prize type, quantity and corresponding winning probability. This lottery method is more diversified, improves the user experience, and enhances the credibility and attractiveness of the event.

[0018] The present invention proposes a lottery method based on multi-dimensional portrait tags of platform users, including two parts: early data preparation and later lottery. In the data preparation part, multi-dimensional portrait tags of platform users are established, and lottery rules are formulated according to tag grouping, and corresponding prizes and winning probabilities are configured. In the lottery part, winning rules are matched by lottery user groups, and then the winning results are calculated.

[0019] Preferably, the method for establishing a multi-dimensional portrait label of a platform user is as follows:

[0020] Data collection: Identify the types of user online behavior data that need to be captured, and collect the corresponding types of user behavior data on the network platform;

[0021] Data processing: The processing methods include cleaning, deduplication, conversion, formatting and standardization;

[0022] Data analysis: After data processing, use machine learning algorithms and data mining techniques to conduct in-depth analysis of the data;

[0023] Portrait construction: Based on business needs and data characteristics, design a reasonable portrait label system to cover multiple feature dimensions of users; select appropriate machine learning algorithms based on the label system and data characteristics; use machine learning algorithms to train preprocessed data to generate user portrait models;

[0024] Label generation: Based on the trained model, generate multi-dimensional portrait labels for each user.

[0025] Preferably, the data collection method includes using web page analysis tools, embedding technology, server logs, client data collection or external data sources.

[0026] Preferably, data is collected from multiple channels, and the types of data collected include static attributes, dynamic attributes, and psychological attributes.

[0027] Preferably, after data processing, data from different channels are integrated to form a unified user data view, and key user features are extracted from the original data, which will be used for subsequent portrait construction.

[0028] Preferably, based on the user's static attributes, dynamic attributes, and psychological attributes, combined with the user's activity, historical purchase records, and multi-dimensional data of user feedback, the commonalities and differences of each user's label are determined, and finally the user portrait labels are intelligently grouped.

[0029] Preferably, before formulating the lottery rules, it is necessary to define the grouping criteria of user portrait tags, then design the lottery levels, and design lottery activities of different levels according to the grouping criteria. Each level sets different prizes, winning probabilities and lottery conditions for users in a specific group, and then formulate specific rules for the lottery activities of each level, including participation conditions and lottery methods.

[0030] Preferably, different winning probabilities are set according to the grouping situation, and prizes matching the portrait tags are set according to the activity purpose and budget.

[0031] Preferably, the activity data is monitored by a data analysis tool when executing the sweepstakes activity.

[0032] Preferably, when a user triggers a lottery draw, the user's probability of winning is calculated by real-time analysis of the user's tag and the lottery rules to which it belongs based on probability theory and statistical methods.

[0033] Compared with the prior art, the present invention has the following beneficial technical effects:

[0034] The present invention creates multi-dimensional and refined portrait tags for each user through in-depth analysis of user behavior data. These tags can accurately reflect the user's personal characteristics, interest preferences and consumption habits, and provide data support for subsequent personalized lottery settings. Personalized lottery rules are formulated according to the user's portrait tags. This means that different user groups will enjoy different lottery treatments according to their characteristics and behaviors, which greatly improves the pertinence and fairness of the lottery. The prize type and quantity, as well as the corresponding winning probability, are flexibly configured according to the lottery rules. This flexibility makes the lottery activities more diverse and attractive, and can also better control the activity costs and risks. When a user participates in the lottery, the system can analyze the user's tags in real time, determine their lottery level and winning probability, and immediately trigger the reward distribution process after the user wins the prize. This real-time performance not only improves the user experience, but also enhances the credibility and attractiveness of the activity. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A flow chart of the method for the data preparation part of the present invention;

[0036] Figure 2 This is a flow chart of the method for the lottery drawing part of the present invention. DETAILED DESCRIPTION

[0037] like Figure 1-Figure 2 As shown, the lottery method based on multi-dimensional portrait tags of platform users proposed in the present invention includes two parts: early data preparation and later lottery drawing; in the data preparation part, by establishing multi-dimensional portrait tags of platform users, and then formulating lottery rules according to tag grouping, configuring corresponding prizes and winning probabilities; in the lottery part, matching the winning rules by lottery user groups, and then calculating the winning results.

[0038] In a further embodiment, the method for establishing a multi-dimensional portrait label of a platform user is as follows:

[0039] Data collection: Identify the types of user online behavior data that need to be captured, and collect the corresponding types of user behavior data on the network platform, such as browsing history, purchase history, search keywords, and other behavior data;

[0040] Data processing: The collected raw data often has problems such as duplication, errors or omissions, and needs to be cleaned to ensure the accuracy and completeness of the data; the processing methods include cleaning, deduplication, conversion, formatting and standardization;

[0041] Data analysis: After data processing, use machine learning algorithms and data mining techniques to conduct in-depth analysis of the data; common analysis methods include behavioral event analysis, user path analysis, page click analysis, etc.

[0042] Portrait construction: According to business needs and data characteristics, design a reasonable portrait label system to cover multiple feature dimensions such as basic user attributes (such as age, gender, region, etc.), behavioral characteristics, interest preferences (such as shopping preferences, active time periods, price sensitivity), etc.; select appropriate machine learning algorithms based on the label system and data characteristics; common algorithms include clustering algorithms (such as K-means), classification algorithms (such as decision trees, support vector machines, logistic regression, etc.), deep learning algorithms, etc.; use machine learning algorithms to train preprocessed data to generate user portrait models. During the training process, it is necessary to continuously optimize model parameters to improve the accuracy and generalization ability of the model;

[0043] Label generation: Based on the trained model, generate multi-dimensional portrait labels for each user.

[0044] By deeply analyzing user behavior data and giving users refined tags, the present invention can more accurately identify users' interests and preferences, thereby setting prizes. This ensures that the lottery is closely related to the actual needs of users, and increases users' willingness and satisfaction in participating in the lottery.

[0045] In further embodiments, the data collection method includes using web analytics tools, embedding technology, server logs, client data collection, or external data sources;

[0046] Web analytics tools: Use web analytics tools such as Google Analytics and Adobe Analytics. These tools can automatically collect user behavior data on web pages, including page views, dwell time, bounce rate, etc.

[0047] Tracking technology: Tracking technology embeds specific codes or scripts in applications or web pages to track specific user behaviors. Tracking can be divided into page tracking, event tracking, etc., which can capture detailed behaviors such as user clicks, swipes, and form filling. Common tracking tools include Amplitude, Mixpanel, etc.

[0048] Server log: Analyze the log files generated by the Web server, which record the interaction between the server and the client, including user requests, responses, error messages, etc. By parsing these logs, you can obtain data such as the user's access path and stay time.

[0049] Client data collection: Data is collected directly on the user's device, such as obtaining real-time user behavior data in the application through SDK, API, etc. This method can provide more accurate and rich user behavior data.

[0050] External data sources: In addition to directly collecting user behavior data, you can also consider using external data sources, such as social media platforms, third-party data providers, etc., to obtain more comprehensive user portraits and behavior data.

[0051] In a further embodiment, data is collected from multiple channels, such as websites, social media, mobile applications, third-party data sources, etc., to ensure the comprehensiveness and diversity of the data. The collected data types include static attributes (such as age, gender, region, etc.), dynamic attributes (such as browsing history, purchase history, search history, etc.) and psychological attributes (such as interest preferences, consumption habits, etc.).

[0052] In a further embodiment, after data processing, data from different channels are integrated to form a unified user data view, and key user features are extracted from the original data, which will be used for subsequent portrait construction.

[0053] In a further embodiment, based on the user's static attributes, dynamic attributes, and psychological attributes, combined with the user's activity, historical purchase records, and multi-dimensional data of user feedback, the commonalities and differences of each user's tags are determined, and finally the user portrait tags are intelligently grouped.

[0054] In a further embodiment, before the lottery rules are formulated, it is necessary to define the grouping criteria of the user portrait tags. These criteria may be based on the user's basic information (such as age, gender), behavioral characteristics (such as activity, purchase frequency), interest preferences (such as product preferences, content interests), etc. Make sure that the grouping criteria are closely related to your business goals and the purpose of the lottery. Then design the lottery level, and design different levels of lottery activities according to the grouping criteria. Each level sets different prizes, winning probabilities and lottery conditions for users in a specific group. For example: High-value user group: For users with high purchase frequency and large consumption amounts, set higher-value prizes and higher winning probabilities to recognize their contributions and encourage continued consumption. Active user group: For users who frequently visit the website and participate in interactions, set interesting interactive links and moderate prizes to increase their participation and loyalty. Potential user group: For newly registered users or users who occasionally purchase, set entry-level prizes and simple lottery conditions to attract them to further understand and experience your products or services. Then formulate specific rules for each level of lucky draw, including participation conditions: clarify what conditions users need to meet to participate in the lucky draw, such as completing specific tasks, reaching a certain score or level, etc.; lucky draw method: determine the method of lucky draw, such as random draw, point exchange, answering questions and lucky draw, etc. Ensure that the lucky draw method is fair and transparent, and can effectively attract user participation.

[0055] The present invention allows different lottery rules to be set according to different tag combinations. This flexibility allows lottery activities to be personalized according to different user groups and scenarios, meeting diverse lottery needs.

[0056] In a further embodiment, different winning probabilities are set according to the grouping situation, and prizes matching the portrait tags are set according to the activity purpose and budget, including physical prizes, virtual goods, coupons, etc. The winning probability can also be adjusted according to the value and quantity of the prizes and the characteristics of the user group.

[0057] By assigning a corresponding lottery level and winning probability to each user according to their tag, the present invention ensures the fairness and transparency of the lottery. This helps to enhance the trust of users and encourages users to participate in the lottery more actively. The present invention also configures the corresponding prize type, quantity and winning probability for each lottery rule, ensuring the diversity and attractiveness of the prizes. At the same time, the automated prize distribution process improves the efficiency of the lottery and reduces errors and delays in manual operations.

[0058] In a further embodiment, when executing a lottery activity, data analysis tools are used to monitor activity data, such as the number of participants, the number of winners, prize collection status, etc., so as to adjust the lottery rules and strategies in a timely manner.

[0059] In a further embodiment, when a user triggers a lottery action, the user's winning probability is calculated by real-time analysis of the user's tags and the lottery rules to which they belong based on probability theory and statistics. The lottery algorithm design has good scalability and adaptability, and can be easily integrated into existing big data analysis platforms and lottery systems. This provides enterprises and organizations with more flexibility, enabling them to adjust and optimize according to different business needs and scenarios.

[0060] The present invention creates multi-dimensional and refined portrait tags for each user through in-depth analysis of user behavior data. These tags can accurately reflect the user's personal characteristics, interest preferences and consumption habits, and provide data support for subsequent personalized lottery settings. Personalized lottery rules are formulated according to the user's portrait tags. This means that different user groups will enjoy different lottery treatments according to their characteristics and behaviors, which greatly improves the pertinence and fairness of the lottery. The prize type and quantity, as well as the corresponding winning probability, are flexibly configured according to the lottery rules. This flexibility makes the lottery activities more diverse and attractive, and can also better control the activity costs and risks. When a user participates in the lottery, the system can analyze the user's tags in real time, determine their lottery level and winning probability, and immediately trigger the reward distribution process after the user wins the prize. This real-time performance not only improves the user experience, but also enhances the credibility and attractiveness of the activity.

[0061] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge scope of technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A lottery method based on multi-dimensional portrait tags of platform users, characterized in that: It includes two parts: early data preparation and later lottery drawing; In the data preparation part, we establish multi-dimensional profile labels for platform users, formulate lottery rules based on label groups, and configure corresponding prizes and winning probabilities; In the lottery part, the winning rules are matched by the lottery user groups, and then the winning results are calculated.

2. The lottery method based on multi-dimensional portrait tags of platform users according to claim 1 is characterized in that: The method of establishing multi-dimensional portrait labels for platform users is as follows: Data collection: Identify the types of user online behavior data that need to be captured, and collect the corresponding types of user behavior data on the network platform; Data processing: The processing methods include cleaning, deduplication, conversion, formatting and standardization; Data analysis: After data processing, use machine learning algorithms and data mining techniques to conduct in-depth analysis of the data; Portrait construction: Based on business needs and data characteristics, design a reasonable portrait label system to cover multiple feature dimensions of users; select appropriate machine learning algorithms based on the label system and data characteristics; use machine learning algorithms to train preprocessed data to generate user portrait models; Label generation: Based on the trained model, generate multi-dimensional portrait labels for each user.

3. The lottery method based on multi-dimensional portrait tags of platform users according to claim 2 is characterized in that: Data collection methods include using web analysis tools, tracking technology, server logs, client data collection or external data sources.

4. The lottery method based on multi-dimensional portrait tags of platform users according to claim 2 is characterized in that: Data is collected from multiple channels, and the types of data collected include static attributes, dynamic attributes, and psychological attributes.

5. The lottery method based on multi-dimensional portrait tags of platform users according to claim 2 is characterized in that: After data processing, data from different channels will be integrated to form a unified user data view, and key user features will be extracted from the original data. These features will be used for subsequent portrait construction.

6. The lottery method based on multi-dimensional portrait tags of platform users according to claim 4 is characterized in that: Based on the user's static attributes, dynamic attributes, and psychological attributes, combined with the user's activity, historical purchase records, and multi-dimensional data of user feedback, we determine the commonalities and differences of each user's labels, and finally intelligently group the user portrait labels.

7. The lottery method based on multi-dimensional portrait tags of platform users according to claim 1 is characterized in that: Before formulating the lottery rules, it is necessary to define the grouping criteria of user portrait tags, then design the lottery levels, and design lottery activities at different levels based on the grouping criteria. Each level sets different prizes, winning probabilities, and lottery conditions for users in specific groups, and then formulate specific rules for the lottery activities at each level, including participation conditions and lottery methods.

8. The lottery method based on multi-dimensional portrait tags of platform users according to claim 1 is characterized in that: Set different winning probabilities based on grouping, and set prizes that match portrait tags based on the purpose and budget of the event.

9. The lottery method based on multi-dimensional portrait tags of platform users according to claim 1 is characterized in that: Monitor activity data through data analysis tools when executing sweepstakes activities.

10. The lottery method based on multi-dimensional portrait tags of platform users according to claim 1, characterized in that: When a user triggers a lottery draw, the user's probability of winning is calculated through real-time analysis of the user's tags and the lottery rules they belong to, based on probability theory and statistical methods.

Images