A cloud platform construction method and system for big data alliance algorithm
Through the big data alliance algorithm, analyzing user behavior, filtering and weighting data resources, the problem of singleness and insufficient quality of cloud platform data sources is solved, and the service quality is improved.
Patent Information
- Application Number
- CN202411281764.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-13
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-09-13
AI Technical Summary
When collecting user data, cloud platforms have problems such as single data source and poor quality, which affects service quality.
By using the big data alliance algorithm, by obtaining user information, extracting user behavior characteristics, performing cluster analysis, identifying group preferences, screening high-quality data resources, performing weighted calculations, and building a collection of alliance weighted data resources for business service development and operation and maintenance .
It improves the quality and diversity of data resources, enhances the service capabilities of the cloud platform to the user group, and solves the problems of singleness and insufficient quality of data sources.
Smart Images

Figure CN118797176B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of cloud computing technology, and in particular to a cloud platform building method and system for a big data alliance algorithm. Background Art
[0002] A cloud platform is a cloud service-based platform for data storage, processing, and analysis. The cloud platform can dynamically expand or reduce resources based on demand, and enterprises can flexibly adjust computing and storage capabilities based on actual workloads without investing in expensive servers and hardware. In addition, it also has many advantages such as security and easy access.
[0003] However, cloud platforms often need a lot of data to support the services they provide to users. For example, if you want to understand user interests and push them to users, or provide the latest and most comprehensive agricultural information for users to search, you need to collect a lot of data. The quality of the collected data determines the quality of cloud platform services. For example, if more specialized platforms such as agricultural websites want to further improve the service quality of cloud platforms for user groups, they must solve the problem of single data source and poor quality when collecting user data. Summary of the invention
[0004] The present invention provides a cloud platform construction method and system for a big data alliance algorithm, the main purpose of which is to solve the problems of single data source and poor quality when collecting user data.
[0005] To achieve the above purpose, the present invention provides a cloud platform construction method for a big data alliance algorithm, comprising:
[0006] Obtain user information of a user cluster in a target cloud platform to obtain a user information set, perform a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set, perform a clustering operation based on preference categories on the user behavior information feature sequence set to obtain individual user clusters, identify the user proportion and preference type of the user clusters, and obtain a group preference set;
[0007] Extracting data resources based on the group preference set from a pre-built alliance database to obtain a candidate data resource set;
[0008] Determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various scoring types including accuracy, completeness, timeliness and ease of use;
[0009] Determine the candidate data resources with the internal score in the candidate data resource set as the in-platform data resource set;
[0010] Determine the candidate data resources that do not have the internal score as a set of off-platform data resources;
[0011] Identify the in-platform scoring set and the out-platform scoring set corresponding to the target scoring type of the target data resource in the in-platform data resource set, calculate the mean of the in-platform scoring set to obtain the in-platform target scoring type mean, and calculate the mean of the out-platform scoring set to obtain the out-platform target scoring type mean;
[0012] Performing weighted calculation on the mean of the target scoring type within the platform and the mean of the target scoring type outside the platform to obtain a first data resource confirmation weight corresponding to the data resource set within the platform;
[0013] Classifying the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set, and calculating the external similarity target rating type mean and the external irrelevant target rating type mean based on the external similarity rating data resource set and the external irrelevant rating data resource set;
[0014] Performing weighted calculation on the mean value of the external similar target scoring type and the mean value of the external irrelevant target scoring type to obtain a second data resource confirmation weight value corresponding to the external platform data resource set;
[0015] Concatenating the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix;
[0016] Performing weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and performing weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set;
[0017] According to preset business activity requirements, the alliance preference weighted data resource set is screened to obtain business-usable data, and the business services in the target cloud platform are constructed using the business-usable data.
[0018] Optionally, performing a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set includes:
[0019] Using a pre-built Jieba word segmentation tool to perform word segmentation processing on the user information set to obtain a word segmentation word set, and performing a stop word deletion operation on the word segmentation word set to obtain a concise word segmentation set;
[0020] Using a pre-built Word2Vec model, a word quantization operation based on the word segmentation order is performed on the concise word segmentation set to obtain a word segmentation vector set;
[0021] Performing a feature recognition operation on the word segmentation vector set to obtain a user feature sequence set;
[0022] A feature classification operation based on user behavior is performed on the user feature sequence set to obtain a user behavior information feature sequence set.
[0023] Optionally, performing a clustering operation based on preference categories on the user behavior information feature sequence set to obtain each user cluster, identifying the user ratio and preference type of the user cluster, and obtaining a group preference set includes:
[0024] Performing a spatial mapping operation of a preset feature engineering dimension on the user behavior information feature sequence set to obtain a user behavior feature point set;
[0025] Performing a K-means clustering operation with an adaptive K value on the user behavior feature point set to obtain K user clusters;
[0026] The user ratio of each user cluster is obtained, each user cluster is arranged according to the user ratio, and the preference types corresponding to the N user clusters with the largest user ratio are screened to obtain a group preference set.
[0027] Optionally, the determining whether each candidate data resource in the candidate data resource set has an internal score of the user cluster includes:
[0028] Obtain the scoring user ID corresponding to the score of each candidate data resource, and use each scoring user ID to build a user vest portrait based on the pre-built real-name association relationship;
[0029] Determine whether there is a rating user ID of a registered user of the target cloud platform in each user's vest portrait;
[0030] When there is a rating user ID belonging to the registered user in the user vest portrait, it is determined that each rating user ID in the user vest portrait belongs to a user within the platform;
[0031] When there is no rating user ID belonging to the registered user in the user vest portrait, it is determined that each rating user ID in the user vest portrait does not belong to a user on the platform;
[0032] Determine whether the candidate data resource has an internal rating of the user cluster based on whether the rating user IDs corresponding to the ratings all belong to users within the platform;
[0033] When the candidate data resource contains a scoring user ID that belongs to a user within the platform, determining that the candidate data resource has an internal score of the user cluster;
[0034] When the scoring user ID that does not have a rating in the candidate data resource belongs to a user within the platform, it is determined that the candidate data resource does not have an internal rating of the user cluster.
[0035] Optionally, the mean of the target scoring type within the platform and the mean of the target scoring type outside the platform are respectively expressed as:
[0036]
[0037]
[0038] In the formula, represents the mean value of the target rating type in the platform, represents the mean value of the target rating type outside the platform, Indicates the number of ratings in the rating set within the platform. Indicates the number of ratings in the off-platform rating set, Indicates the first Ratings, , Indicates the first Ratings, , represents one of the rating types, ,in, They stand for {accuracy, completeness, timeliness and usability} respectively.
[0039] Optionally, the first data resource confirmation weight is expressed as
[0040]
[0041] In the formula, Indicates the first data resource confirmation value, The adjustment factor representing the mean of the target rating type within the platform is set to , indicating that the influence of the mean of the target score type within the platform is greater than the influence of the mean of the target score type outside the platform by default.
[0042] Optionally, the off-platform data resource set is classified based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set, including:
[0043] The similarity between the off-platform users corresponding to each score in the off-platform data resource set and the on-platform users corresponding to each score in the on-platform data resource set is calculated based on the Pearson correlation coefficient:
[0044]
[0045] In the formula, Indicates similarity, represents a user in the platform, represents users outside the platform, n1 represents the data resource collection within the platform, and n2 represents the data resource collection outside the platform. represents the intersection data volume of the public data resources of the data resource set within the platform and the data resource set outside the platform, represents the amount of data of the union of the data resource set within the platform and the data resource set outside the platform, represents candidate data resources rated by both users on the platform and users outside the platform, Indicates users within the platform Candidate data resources Ratings, Indicates users outside the platform Candidate data resources Rating, Indicates users within the platform Candidate data resources The average of all ratings of Indicates users outside the platform Candidate data resources The mean of all the ratings;
[0046] Determine whether the similarity of users outside each platform is greater than a preset similarity threshold;
[0047] Determine the off-platform users whose similarity is greater than the similarity threshold as off-platform similar users, and obtain the external similarity rating data resource set corresponding to the off-platform similar users;
[0048] Off-platform users whose similarity is less than or equal to the similarity threshold are determined as off-platform irrelevant users, and an off-platform irrelevant rating data resource set corresponding to the off-platform irrelevant users is obtained.
[0049] Optionally, the second data resource confirmation weight is expressed as:
[0050]
[0051] In the formula, represents the second data resource confirmation value, represents the adjustment factor of the mean of the external similarity target score type, which is set to , indicating that the influence of the mean of the external similar target score type is greater than the influence of the mean of the external irrelevant target score type by default.
[0052] Optionally, the data resource weight matrix is expressed as:
[0053]
[0054] In the formula, represents the data resource confirmation weight matrix, Indicates The candidate data resource The weight value of each scoring type, , Represents the total amount of data in the candidate data resource set, .
[0055] To achieve the above objectives, the present invention also provides a cloud platform construction system for a big data alliance algorithm, comprising:
[0056] A group preference collection module is used to obtain user information of a user cluster in a target cloud platform, obtain a user information set, perform a feature extraction operation based on user behavior on the user information set, obtain a user behavior information feature sequence set, perform a clustering operation based on preference categories on the user behavior information feature sequence set, obtain individual user clusters, identify the user ratio and preference type of the user clusters, and obtain a group preference set;
[0057] The alliance big data acquisition module is used to extract data resources based on the group preference set from the pre-built alliance database to obtain a candidate data resource set, and determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various scoring types including accuracy, completeness, timeliness and ease of use, and determine the candidate data resources in the candidate data resource set that have the internal score as the platform-based data resource set, and determine the candidate data resources that do not have the internal score as the platform-based data resource set;
[0058] The data resource confirmation module within the platform is used to identify the on-platform scoring set and the off-platform scoring set corresponding to the target scoring type of the target data resource in the on-platform data resource set, and calculate the mean of the on-platform scoring set to obtain the on-platform target scoring type mean, and calculate the mean of the off-platform scoring set to obtain the off-platform target scoring type mean, and perform weighted calculation on the on-platform target scoring type mean and the off-platform target scoring type mean to obtain the first data resource confirmation value corresponding to the on-platform data resource set;
[0059] The off-platform data resource rights confirmation module is used to classify the off-platform data resource set based on the similarity between off-platform users and on-platform users, obtain an external similarity scoring data resource set and an external irrelevant scoring data resource set, and calculate the external similarity target scoring type mean and the external irrelevant target scoring type mean according to the external similarity scoring data resource set and the external irrelevant scoring data resource set, and perform weighted calculation on the external similarity target scoring type mean and the external irrelevant target scoring type mean to obtain a second data resource rights confirmation value corresponding to the off-platform data resource set;
[0060] An alliance preference weighting module, used to concatenate the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix, and to perform weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and to perform weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set;
[0061] The service building module is used to screen the alliance preference weighted data resource set according to preset business activity requirements, obtain business-usable data, and use the business-usable data to build business services in the target cloud platform.
[0062] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:
[0063] a memory storing at least one instruction; and
[0064] The processor executes the instructions stored in the memory to implement the cloud platform construction method of the big data alliance algorithm described above.
[0065] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is executed by a processor in an electronic device to implement the cloud platform building method of the big data alliance algorithm described above.
[0066] In order to solve the problems described in the background technology, the present invention first collects the group preferences of users in the cloud platform, so as to provide directions for the expansion of cloud platform business, the organization of activities or the push of interfaces; then the present invention realizes the collection of massive data through the alliance algorithm, which is conducive to improving the level of data resource push; the present invention screens the alliance database once according to the group preferences to obtain a candidate data resource set, and the candidate data resource set can be used as the underlying data for the development of cloud platform business; wherein, the candidate data resource set contains data rated by cloud platform users and data not rated by cloud platform users; the present invention can play a decisive role in the selection of data resources through the scoring of candidate data resources containing the ratings of users in the platform; and for candidate data resources with only external users of the platform, the candidate data resource set can be used as the underlying data for the development of cloud platform business; wherein, the candidate data resource set contains data rated by cloud platform users and data not rated by cloud platform users; the present invention can play a decisive role in the selection of data resources through the scoring of candidate data resources containing the ratings of users in the platform; and for The candidate data resources of the user's rating are considered. Considering the similar users outside the platform who are similar to the users inside the platform, the ratings of the similar users outside the platform will have a higher degree of influence on the recommendation results than the irrelevant users outside the platform. Therefore, the external similarity rating data resource set of the data resource set outside the platform is used as an important reference for the operation and maintenance of the cloud platform; the present invention calculates the first data resource confirmation value corresponding to the data resource set inside the platform, and the second data resource confirmation value corresponding to the data resource set outside the platform, and then performs comprehensive weighting according to the group preference set including the user proportion and preference type, to obtain the alliance preference weighted data resource set that takes into account the quality of the alliance data resources and the preferences of the users of this platform; the cloud platform can realize business development and activity operation and maintenance according to the alliance preference weighted data resource set. Therefore, the present invention can solve the problems of single data source and poor quality when collecting user data. BRIEF DESCRIPTION OF THE DRAWINGS
[0067] Figure 1 A schematic diagram of a process for building a cloud platform for a big data alliance algorithm provided by an embodiment of the present invention;
[0068] Figure 2 A functional module diagram of a cloud platform building system for a big data alliance algorithm provided in one embodiment of the present invention;
[0069] Figure 3 A schematic diagram of the structure of an electronic device for implementing a cloud platform construction method for the big data alliance algorithm provided in one embodiment of the present invention.
[0070] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0071] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.
[0072] The embodiment of the present application provides a method for building a cloud platform for a big data alliance algorithm. The execution subject of the cloud platform construction method for the big data alliance algorithm includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the cloud platform construction method for the big data alliance algorithm can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.
[0073] Reference Figure 1 FIG. 1 is a flow chart of a method for building a cloud platform for a big data alliance algorithm according to an embodiment of the present invention. In this embodiment, the method for building a cloud platform for a big data alliance algorithm includes:
[0074] S1. Obtain user information of a user cluster in a target cloud platform to obtain a user information set, perform a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set, and perform a clustering operation based on preference categories on the user behavior information feature sequence set to obtain individual user clusters, identify the user proportion and preference type of the user clusters, and obtain a group preference set.
[0075] In the embodiment of the present invention, the target cloud platform is a platform for sorting out agricultural related credit based on cloud technology, so that users can query agricultural information and purchase agricultural products.
[0076] Furthermore, the user cluster refers to the user group registered in the target cloud platform. The present invention can view the behavior habits of the user group based on the user information database inside the cloud platform, for example, output the user information set based on the click volume of each data resource in the cloud platform, the seasonal fluctuation of the consumption of physical resources, etc. Among them, the user information set can obtain the changes in user preferences, which is beneficial for the target cloud platform to push information, hold activities and build new businesses.
[0077] In detail, in the embodiment of the present invention, the user information set is subjected to a feature extraction operation based on user behavior to obtain a user behavior information feature sequence set, including:
[0078] Using a pre-built Jieba word segmentation tool to perform word segmentation processing on the user information set to obtain a word segmentation word set, and performing a stop word deletion operation on the word segmentation word set to obtain a concise word segmentation set;
[0079] Using a pre-built Word2Vec model, a word quantization operation based on the word segmentation order is performed on the concise word segmentation set to obtain a word segmentation vector set;
[0080] Performing a feature recognition operation on the word segmentation vector set to obtain a user feature sequence set;
[0081] A feature classification operation based on user behavior is performed on the user feature sequence set to obtain a user behavior information feature sequence set.
[0082] The Jieba word segmentation tool is a Chinese word segmentation tool used to segment continuous Chinese text into individual words to facilitate subsequent text analysis, information retrieval, machine learning, etc. The present invention first converts the messy text data into a word segmentation word set through the Jieba word segmentation tool, and then eliminates stop words such as this and which in the word segmentation word set to reduce the text volume and obtain a concise word segmentation set.
[0083] Furthermore, the Word2Vec model is a word embedding model for natural language processing, which is used to map words into vectors in a high-dimensional space so that similar words have similar representations in the vector space. The embodiment of the present invention converts a concise word segmentation set into a quantized code through the Word2Vec model to obtain a word segmentation vector set, and then performs feature recognition on the word segmentation vector set through the convolution layer, pooling layer, and fully connected layer commonly used in the neural network model to obtain a user feature sequence set.
[0084] In order to obtain group preferences, the embodiment of the present invention does not consider the user's own identity information and other information, and only identifies the user's user behavior, such as clicks, views, browsing, and browsing objects. Therefore, the present invention screens user behaviors by classifying user behaviors based on their features, and obtains a user behavior information feature sequence set.
[0085] The user behavior information feature sequence set includes data resource objects, click rates, click volumes, change trends, etc. By analyzing the user behavior information feature sequence set, group preferences can be seen. For example, a large number of users are interested in modules such as global food price changes and agricultural by-products.
[0086] Furthermore, in an embodiment of the present invention, the clustering operation based on preference categories is performed on the user behavior information feature sequence set to obtain each user cluster, and the user ratio and preference type of the user cluster are identified to obtain a group preference set, including:
[0087] Performing a spatial mapping operation of a preset feature engineering dimension on the user behavior information feature sequence set to obtain a user behavior feature point set;
[0088] Performing a K-means clustering operation with an adaptive K value on the user behavior feature point set to obtain K user clusters;
[0089] The user ratio of each user cluster is obtained, each user cluster is arranged according to the user ratio, and the preference types corresponding to the N user clusters with the largest user ratio are screened to obtain a group preference set.
[0090] Specifically, in an embodiment of the present invention, the preset feature engineering dimensions may include visit frequency, dwell time, interactive behavior, purchasing behavior, browsing history, product category preference, brand preference, price sensitivity and seasonal behavior, etc., and a set of user behavior feature points is obtained through the multi-dimensional feature space of feature engineering.
[0091] Then, the present invention uses the K-means clustering algorithm to implement the grouping operation for user behavior. The present invention does not limit the value of K, so that the K value changes adaptively, for example, using the elbow rule, by plotting the clustering error (SSE) under different K values, the "elbow" point is automatically found to help select a suitable K.
[0092] Then, the present invention obtains the user ratio of each user cluster, and selects the preference types corresponding to the top-N user clusters by sorting, as a group preference set.
[0093] S2. Extracting data resources based on the group preference set from a pre-built alliance database to obtain a candidate data resource set.
[0094] In the embodiment of the present invention, the alliance database is a large database built based on an alliance algorithm, and may include user data of other platforms that cooperate with the target cloud platform.
[0095] The present invention screens the data resources corresponding to the group preference set in the alliance database by means of keyword extraction to obtain a candidate data resource set.
[0096] S3. Determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various score types including accuracy, completeness, timeliness and ease of use.
[0097] In detail, in the embodiment of the present invention, the determining whether each candidate data resource in the candidate data resource set has an internal score of the user cluster includes:
[0098] Obtain the scoring user ID corresponding to the score of each candidate data resource, and use each scoring user ID to build a user vest portrait based on the pre-built real-name association relationship;
[0099] Determine whether there is a rating user ID of a registered user of the target cloud platform in each user's vest portrait;
[0100] When there is a rating user ID belonging to the registered user in the user vest portrait, it is determined that each rating user ID in the user vest portrait belongs to a user within the platform;
[0101] When there is no rating user ID belonging to the registered user in the user vest portrait, it is determined that each rating user ID in the user vest portrait does not belong to a user on the platform;
[0102] Determine whether the candidate data resource has an internal rating of the user cluster based on whether the rating user IDs corresponding to the ratings all belong to users within the platform;
[0103] When the candidate data resource contains a scoring user ID that belongs to a user within the platform, determining that the candidate data resource has an internal score of the user cluster;
[0104] When the scoring user ID that does not have a rating in the candidate data resource belongs to a user within the platform, it is determined that the candidate data resource does not have an internal rating of the user cluster.
[0105] It should be known that the ratings of each candidate data resource may come from different rating user IDs on different platforms, but since users use a platform in a data alliance, they are likely to be exposed to other platforms in the alliance. For example, if you use *du, you are likely to use *hub. Although the rating user IDs in each platform are different, different rating user IDs may be the same user. The candidate data resources that Zhang San rated on the target cloud platform can be called in-platform data resources, and the data resources that Zhang San rated on another platform can also be regarded as in-platform data resources that can be collected by the target cloud platform, because these data resources are all data evaluated by users within the target cloud platform.
[0106] Therefore, in the embodiment of the present invention, each scoring user ID is associated with the user ID directly or indirectly through the real-name system, mobile phone number, etc. to obtain the user vest portrait. Then, the user vest portrait is used for scoring identification to determine whether each candidate data resource in the candidate data resource set has the internal score of the user cluster.
[0107] Furthermore, in the embodiment of the present invention, the internal rating is a score set of various rating types including accuracy, completeness, timeliness and ease of use.
[0108] Among them, accuracy refers to whether the data correctly reflects the facts or realities it represents; completeness refers to whether the data set contains all the required information and fields and whether there are missing values; timeliness refers to whether the data is available within the required time, reflecting the frequency and delay of data updates; and ease of use refers to whether the use and access of data resources are easy and can meet the needs of users. The four standards of accuracy, completeness, timeliness and ease of use are the cornerstones of candidate data resources. High-quality candidate data resources can effectively support the analytical decision-making of the target cloud platform, optimize business processes, improve user satisfaction, and promote the overall efficiency and effectiveness of the target cloud platform.
[0109] S4. Determine the candidate data resources with the internal scores in the candidate data resource set as the in-platform data resource set.
[0110] In the embodiment of the present invention, for the candidate data resources containing the ratings of the users in the platform, the distribution of the ratings can play a decisive role in the selection of the data resources, and therefore, they are preferentially screened to obtain the set n1 of data resources in the platform.
[0111] S5. Determine the candidate data resources that do not have the internal score as a set of off-platform data resources.
[0112] In the embodiment of the present invention, the candidate data resources without the ratings may be regarded as data resources that have not been evaluated by users in the platform.
[0113] The data quality and content of the data resources that the user has not evaluated may not be recognized by the users in the platform, but there may be data that will be recognized. For example, when the preferences and behaviors of a user group C on a platform are similar to the preferences and behaviors of a user cluster B in the target cloud platform, the score of a data resource by the user group C may be accepted by the user cluster B. Therefore, the present invention also needs to determine the candidate data resources that do not have the internal score as the off-platform data resource set n2 to further collect data resources.
[0114] S6. Identify the in-platform scoring set and the out-platform scoring set corresponding to the target scoring type of the target data resource in the in-platform data resource set, and calculate the mean of the in-platform scoring set to obtain the in-platform target scoring type mean, and calculate the mean of the out-platform scoring set to obtain the out-platform target scoring type mean.
[0115] In the embodiment of the present invention, there are n1 data resources in the data resource set in the platform, and a target data resource p (p=1, 2, 3, ..., n1) is randomly selected from the data resource set in the platform. The target data resource p has a target scoring type j, ,in, They stand for {accuracy, completeness, timeliness and usability} respectively.
[0116] The target rating type j may be m1 user ratings within the platform, that is, the number of data resources in the rating set within the platform is m1, and the remaining m2 data resources are summarized as the rating set outside the platform, where .
[0117] The embodiment of the present invention can calculate the mean value s1 of the target rating type within the platform and the mean value s2 of the target rating type outside the platform respectively according to the rating set within the platform and the rating set outside the platform.
[0118] In detail, in the embodiment of the present invention, the mean of the target scoring type within the platform and the mean of the target scoring type outside the platform are respectively expressed as:
[0119]
[0120]
[0121] In the formula, represents the mean value of the target rating type in the platform, represents the mean value of the target rating type outside the platform, Indicates the number of ratings in the rating set within the platform. Indicates the number of ratings in the off-platform rating set, Indicates the first Ratings, , Indicates the first Ratings, , represents one of the rating types, ,in, They stand for {accuracy, completeness, timeliness and usability} respectively.
[0122] In the embodiment of the present invention, the calculation and This allows us to directly compare the quality of data resources within and outside the platform. By comparing the means, we can determine which platform's data resources have a higher score, thus providing a basis for data selection or decision-making.
[0123] S7. Perform weighted calculation on the average of the target scoring types within the platform and the average of the target scoring types outside the platform to obtain a first data resource confirmation weight corresponding to the data resource set within the platform.
[0124] In detail, in the embodiment of the present invention, the first data resource confirmation weight is expressed as
[0125]
[0126] In the formula, Indicates the first data resource confirmation value, The adjustment factor representing the mean of the target rating type within the platform is set to , indicating that the influence of the mean of the target score type within the platform is greater than the influence of the mean of the target score type outside the platform by default.
[0127] In the embodiment of the present invention, by It can represent the weight of the data resources in the data resource set n1 within the platform.
[0128] S8. Classify the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set, and calculate the external similarity target rating type mean and the external irrelevant target rating type mean based on the external similarity rating data resource set and the external irrelevant rating data resource set.
[0129] In detail, in an embodiment of the present invention, the classification of the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set includes:
[0130] The similarity between the off-platform users corresponding to each score in the off-platform data resource set and the on-platform users corresponding to each score in the on-platform data resource set is calculated based on the Pearson correlation coefficient:
[0131]
[0132] In the formula, Indicates similarity, represents a user in the platform, represents users outside the platform, n1 represents the data resource collection within the platform, and n2 represents the data resource collection outside the platform. represents the intersection data volume of the public data resources of the data resource set within the platform and the data resource set outside the platform, represents the amount of data of the union of the data resource set within the platform and the data resource set outside the platform, represents candidate data resources rated by both users on the platform and users outside the platform, Indicates users within the platform Candidate data resources Rating, Indicates users outside the platform Candidate data resources Rating, Indicates users within the platform Candidate data resources The average of all ratings of Indicates users outside the platform Candidate data resources The mean of all the ratings;
[0133] Determine whether the similarity of users outside each platform is greater than a preset similarity threshold;
[0134] Determine the off-platform users whose similarity is greater than the similarity threshold as off-platform similar users, and obtain the external similarity rating data resource set corresponding to the off-platform similar users;
[0135] Off-platform users whose similarity is less than or equal to the similarity threshold are determined as off-platform irrelevant users, and an off-platform irrelevant rating data resource set corresponding to the off-platform irrelevant users is obtained.
[0136] The Pearson Correlation Coefficient is a statistical indicator used to measure the strength and direction of the linear relationship between two variables, and its value range is between -1 and 1.
[0137] In the embodiment of the present invention, through the above Formula, combined with the Pearson correlation coefficient, calculates the similarity between users outside the platform and users inside the platform, and then classifies the data resource set outside the platform through a preset similarity threshold to obtain an external similarity score data resource set and an external irrelevant score data resource set. Among them, the similarity threshold can be adjusted according to the data volume requirements and user satisfaction requirements of the actual business needs of the target cloud platform. The larger the data volume requirement, the smaller the similarity threshold, and the greater the user satisfaction requirement, the larger the similarity threshold.
[0138] S9. Perform weighted calculation on the mean value of the external similar target score type and the mean value of the external irrelevant target score type to obtain a second data resource confirmation weight value corresponding to the external data resource set of the platform.
[0139] The same weighting algorithm is configured as in the above-mentioned operation steps in S7.
[0140] In detail, in the embodiment of the present invention, the second data resource confirmation value is expressed as:
[0141]
[0142] In the formula, represents the second data resource confirmation value, represents the adjustment factor of the mean of the external similarity target score type, which is set to , indicating that the influence of the mean of the external similar target score type is greater than the influence of the mean of the external irrelevant target score type by default.
[0143] In the embodiment of the present invention, the The weight coefficient of each candidate data resource in the off-platform data resource set may be represented.
[0144] S10. Concatenate the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix.
[0145] In detail, in the embodiment of the present invention, the data resource confirmation weight matrix is expressed as:
[0146]
[0147] In the formula, represents the data resource confirmation weight matrix, Indicates The candidate data resource The weight value of each scoring type, , Represents the total amount of data in the candidate data resource set, .
[0148] Specifically, in the embodiment of the present invention, when When (1, m1) is taken, Represents the first data resource confirmation value, and when When (m1, n) is taken, Represents the second data resource confirmation weight value, and through splicing, the data resource confirmation weight value matrix corresponding to the candidate data resource set is obtained .
[0149] S11. Perform weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and perform weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set.
[0150] In the embodiment of the present invention, it is expected that the cloud platform construction can be realized through the candidate data resource set, and the collected data resources are required to meet both the data quality requirements and the preferences of users in the platform.
[0151] Therefore, in the embodiment of the present invention, the weight matrix is first determined by the data resource A weighted calculation is performed on the candidate data resource set, and then a weighted calculation is performed on the candidate data resource set using a group preference set including user proportions and preference types, and finally an alliance preference weighted data resource set is obtained.
[0152] Among them, the data resource weight matrix The weighted calculation process is relatively simple and can be calculated using a matrix formula. The process of weighting using the group preference set requires identifying the preference type corresponding to each candidate data resource in the candidate data resource set, and then performing a separate weighted calculation on each candidate data resource using the normalized user ratio and preference type.
[0153] S12. According to preset business activity requirements, the alliance preference weighted data resource set is screened to obtain business-usable data, and the business services in the target cloud platform are constructed using the business-usable data.
[0154] In the embodiment of the present invention, the business activity requirements include price reduction activities for agricultural and sideline products, display of popular agricultural information homepage, or establishment of other personalized push services.
[0155] The present invention can select appropriate data from the alliance preference weighted data resource set as the basis for business operation and maintenance according to different business needs. No matter what kind of data is extracted, it can meet the data quality and the preferences of users in the platform, so as to better complete the construction and update of the target cloud platform.
[0156] In order to solve the problems described in the background technology, the present invention first collects the group preferences of users in the cloud platform, so as to provide directions for the expansion of cloud platform business, the organization of activities or the push of interfaces; then the present invention realizes the collection of massive data through the alliance algorithm, which is conducive to improving the level of data resource push; the present invention screens the alliance database once according to the group preferences to obtain a candidate data resource set, and the candidate data resource set can be used as the underlying data for the development of cloud platform business; wherein, the candidate data resource set contains data rated by cloud platform users and data not rated by cloud platform users; the present invention can play a decisive role in the selection of data resources through the scoring of candidate data resources containing the ratings of users in the platform; and for candidate data resources with only external users of the platform, the candidate data resource set can be used as the underlying data for the development of cloud platform business; wherein, the candidate data resource set contains data rated by cloud platform users and data not rated by cloud platform users; the present invention can play a decisive role in the selection of data resources through the scoring of candidate data resources containing the ratings of users in the platform; and for The candidate data resources of the user's rating are considered. Considering the similar users outside the platform who are similar to the users inside the platform, the ratings of the similar users outside the platform will have a higher degree of influence on the recommendation results than the irrelevant users outside the platform. Therefore, the external similarity rating data resource set of the data resource set outside the platform is used as an important reference for the operation and maintenance of the cloud platform; the present invention calculates the first data resource confirmation value corresponding to the data resource set inside the platform, and the second data resource confirmation value corresponding to the data resource set outside the platform, and then performs comprehensive weighting according to the group preference set including the user proportion and preference type, to obtain the alliance preference weighted data resource set that takes into account the quality of the alliance data resources and the preferences of the users of this platform; the cloud platform can realize business development and activity operation and maintenance according to the alliance preference weighted data resource set. Therefore, the present invention can solve the problems of single data source and poor quality when collecting user data.
[0157] like Figure 2 The figure shows a functional module diagram of a cloud platform building system for a big data alliance algorithm provided by an embodiment of the present invention.
[0158] The cloud platform construction system 100 of the big data alliance algorithm of the present invention can be installed in an electronic device. According to the functions implemented, the cloud platform construction system 100 of the big data alliance algorithm can include a group preference collection module 101, an alliance big data collection module 102, a platform data resource right confirmation module 103, an off-platform data resource right confirmation module 104, an alliance preference weighting module 105 and a service construction module 106. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.
[0159] The group preference acquisition module 101 is used to obtain user information of a user cluster in a target cloud platform to obtain a user information set, perform a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set, perform a clustering operation based on preference categories on the user behavior information feature sequence set to obtain individual user clusters, identify the user ratio and preference type of the user clusters, and obtain a group preference set;
[0160] The alliance big data collection module 102 is used to extract data resources based on the group preference set from the pre-built alliance database to obtain a candidate data resource set, and determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various scoring types including accuracy, completeness, timeliness and ease of use, and determine the candidate data resources in the candidate data resource set that have the internal score as the in-platform data resource set, and determine the candidate data resources that do not have the internal score as the out-platform data resource set;
[0161] The platform data resource right confirmation module 103 is used to identify the platform scoring set and the platform scoring set corresponding to the target scoring type of the target data resource in the platform data resource set, and calculate the mean of the platform scoring set to obtain the platform target scoring type mean, and calculate the mean of the platform scoring set to obtain the platform target scoring type mean, and perform weighted calculation on the platform target scoring type mean and the platform target scoring type mean to obtain the first data resource right confirmation value corresponding to the platform data resource set;
[0162] The off-platform data resource right confirmation module 104 is used to classify the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity scoring data resource set and an external irrelevant scoring data resource set, and calculate the external similarity target scoring type mean and the external irrelevant target scoring type mean according to the external similarity scoring data resource set and the external irrelevant scoring data resource set, and perform weighted calculation on the external similarity target scoring type mean and the external irrelevant target scoring type mean to obtain a second data resource right confirmation value corresponding to the off-platform data resource set;
[0163] The alliance preference weighting module 105 is used to concatenate the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix, and to perform weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and to perform weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set;
[0164] The service building module 106 is used to screen the alliance preference weighted data resource set according to preset business activity requirements, obtain business available data, and use the business available data to build business services in the target cloud platform.
[0165] In detail, each module in the cloud platform building system 100 of the big data alliance algorithm in the embodiment of the present invention is used in the same manner as above. Figure 1 The cloud platform construction method of the big data alliance algorithm described in the text is the same technical means and can produce the same technical effects, so I will not go into details here.
[0166] like Figure 3 , is a schematic diagram of the structure of an electronic device for implementing a cloud platform construction method for a big data alliance algorithm provided by an embodiment of the present invention.
[0167] The electronic device 1 may include a processor 10, a memory 11 and a bus 12, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as a cloud platform building method program for a big data alliance algorithm.
[0168] Among them, the memory 11 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (for example: SD or DX memory, etc.), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 11 can be an internal storage unit of the electronic device 1, such as a mobile hard disk of the electronic device 1. In other embodiments, the memory 11 can also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 1. Further, the memory 11 also includes an internal storage unit of the electronic device 1 and an external storage device. The memory 11 can not only be used to store application software and various types of data installed in the electronic device 1, such as the code of the cloud platform construction method program of the big data alliance algorithm, but also can be used to temporarily store data that has been output or is to be output.
[0169] The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips, etc. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect various components of the entire electronic device, and executes or executes programs or modules stored in the memory 11 (such as a cloud platform building method program for a big data alliance algorithm, etc.), and calls data stored in the memory 11 to execute various functions of the electronic device 1 and process data.
[0170] The bus 12 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus 12 may be divided into an address bus, a data bus, a control bus, etc. The bus 12 is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.
[0171] Figure 3 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.
[0172] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to various components. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.
[0173] Furthermore, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices.
[0174] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.
[0175] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.
[0176] The cloud platform building method program of the big data alliance algorithm stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve:
[0177] Obtain user information of a user cluster in a target cloud platform to obtain a user information set, perform a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set, perform a clustering operation based on preference categories on the user behavior information feature sequence set to obtain individual user clusters, identify the user proportion and preference type of the user clusters, and obtain a group preference set;
[0178] Extracting data resources based on the group preference set from a pre-built alliance database to obtain a candidate data resource set;
[0179] Determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various scoring types including accuracy, completeness, timeliness and ease of use;
[0180] Determine the candidate data resources with the internal score in the candidate data resource set as the in-platform data resource set;
[0181] Determine the candidate data resources that do not have the internal score as a set of off-platform data resources;
[0182] Identify the in-platform scoring set and the out-platform scoring set corresponding to the target scoring type of the target data resource in the in-platform data resource set, calculate the mean of the in-platform scoring set to obtain the in-platform target scoring type mean, and calculate the mean of the out-platform scoring set to obtain the out-platform target scoring type mean;
[0183] Performing weighted calculation on the mean of the target scoring type within the platform and the mean of the target scoring type outside the platform to obtain a first data resource confirmation weight corresponding to the data resource set within the platform;
[0184] Classifying the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set, and calculating the external similarity target rating type mean and the external irrelevant target rating type mean based on the external similarity rating data resource set and the external irrelevant rating data resource set;
[0185] Performing weighted calculation on the mean value of the external similar target scoring type and the mean value of the external irrelevant target scoring type to obtain a second data resource confirmation weight value corresponding to the external platform data resource set;
[0186] Concatenating the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix;
[0187] Performing weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and performing weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set;
[0188] According to preset business activity requirements, the alliance preference weighted data resource set is screened to obtain business-usable data, and the business services in the target cloud platform are constructed using the business-usable data.
[0189] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figures 1 to 3 The description of the relevant steps in the corresponding embodiments will not be repeated here.
[0190] Furthermore, if the module / unit integrated in the electronic device 1 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).
[0191] The present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement:
[0192] Obtain user information of a user cluster in a target cloud platform to obtain a user information set, perform a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set, perform a clustering operation based on preference categories on the user behavior information feature sequence set to obtain individual user clusters, identify the user proportion and preference type of the user clusters, and obtain a group preference set;
[0193] Extracting data resources based on the group preference set from a pre-built alliance database to obtain a candidate data resource set;
[0194] Determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various scoring types including accuracy, completeness, timeliness and ease of use;
[0195] Determine the candidate data resources with the internal score in the candidate data resource set as the in-platform data resource set;
[0196] Determine the candidate data resources that do not have the internal score as a set of off-platform data resources;
[0197] Identify the in-platform scoring set and the out-platform scoring set corresponding to the target scoring type of the target data resource in the in-platform data resource set, calculate the mean of the in-platform scoring set to obtain the in-platform target scoring type mean, and calculate the mean of the out-platform scoring set to obtain the out-platform target scoring type mean;
[0198] Performing weighted calculation on the mean of the target scoring type within the platform and the mean of the target scoring type outside the platform to obtain a first data resource confirmation weight corresponding to the data resource set within the platform;
[0199] Classifying the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set, and calculating the external similarity target rating type mean and the external irrelevant target rating type mean based on the external similarity rating data resource set and the external irrelevant rating data resource set;
[0200] Performing weighted calculation on the mean value of the external similar target scoring type and the mean value of the external irrelevant target scoring type to obtain a second data resource confirmation weight value corresponding to the external platform data resource set;
[0201] Concatenating the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix;
[0202] Performing weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and performing weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set;
[0203] According to preset business activity requirements, the alliance preference weighted data resource set is screened to obtain business-usable data, and the business services in the target cloud platform are constructed using the business-usable data.
[0204] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative, and actual implementation may have other division methods.
[0205] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0206] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0207] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0208] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The second and other words are used to indicate names, but not to indicate any particular order.
[0209] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.
Claims
1. A cloud platform construction method for a big data alliance algorithm, characterized in that: The method comprises: Obtain user information of a user cluster in a target cloud platform to obtain a user information set, perform a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set, perform a clustering operation based on preference categories on the user behavior information feature sequence set to obtain individual user clusters, identify the user proportion and preference type of the user clusters, and obtain a group preference set; Extracting data resources based on the group preference set from a pre-built alliance database to obtain a candidate data resource set; Determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various score types including accuracy, completeness, timeliness and ease of use; Determine the candidate data resources with the internal score in the candidate data resource set as the in-platform data resource set; Determine the candidate data resources that do not have the internal score as a set of off-platform data resources; Identify the in-platform scoring set and the out-platform scoring set corresponding to the target scoring type of the target data resource in the in-platform data resource set, calculate the mean of the in-platform scoring set to obtain the in-platform target scoring type mean, and calculate the mean of the out-platform scoring set to obtain the out-platform target scoring type mean; Performing weighted calculation on the mean of the target scoring type within the platform and the mean of the target scoring type outside the platform to obtain a first data resource confirmation weight corresponding to the data resource set within the platform; Classifying the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set, and calculating the external similarity target rating type mean and the external irrelevant target rating type mean based on the external similarity rating data resource set and the external irrelevant rating data resource set; Performing weighted calculation on the mean value of the external similar target scoring type and the mean value of the external irrelevant target scoring type to obtain a second data resource confirmation weight value corresponding to the external platform data resource set; Concatenating the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix; Performing weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and performing weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set; According to preset business activity requirements, the alliance preference weighted data resource set is screened to obtain business-usable data, and the business services in the target cloud platform are constructed using the business-usable data.
2. The cloud platform construction method of the big data alliance algorithm according to claim 1, characterized in that: The performing of a feature extraction operation based on user behavior on the user information set to obtain a user behavior information feature sequence set includes: Using a pre-built Jieba word segmentation tool to perform word segmentation processing on the user information set to obtain a word segmentation word set, and performing a stop word deletion operation on the word segmentation word set to obtain a concise word segmentation set; Using a pre-built Word2Vec model, a word quantization operation based on the word segmentation order is performed on the concise word segmentation set to obtain a word segmentation vector set; Performing a feature recognition operation on the word segmentation vector set to obtain a user feature sequence set; A feature classification operation based on user behavior is performed on the user feature sequence set to obtain a user behavior information feature sequence set.
3. The cloud platform construction method of the big data alliance algorithm according to claim 2, characterized in that: The clustering operation based on preference categories is performed on the user behavior information feature sequence set to obtain each user clustering cluster, and the user proportion and preference type of the user clustering cluster are identified to obtain a group preference set, including: Performing a spatial mapping operation of a preset feature engineering dimension on the user behavior information feature sequence set to obtain a user behavior feature point set; Performing a K-means clustering operation with an adaptive K value on the user behavior feature point set to obtain K user clusters; The user ratio of each user cluster is obtained, each user cluster is arranged according to the user ratio, and the preference types corresponding to the N user clusters with the largest user ratio are screened to obtain a group preference set.
4. The cloud platform construction method of the big data alliance algorithm according to claim 3 is characterized in that: The determining whether each candidate data resource in the candidate data resource set has an internal score of the user cluster includes: Obtain the scoring user ID corresponding to the score of each candidate data resource, and use each scoring user ID to build a user vest portrait based on the pre-built real-name association relationship; Determine whether there is a rating user ID of a registered user of the target cloud platform in each user's vest portrait; When there is a rating user ID belonging to the registered user in the user vest portrait, it is determined that each rating user ID in the user vest portrait belongs to a user within the platform; When there is no rating user ID belonging to the registered user in the user vest portrait, it is determined that each rating user ID in the user vest portrait does not belong to a user on the platform; Determine whether the candidate data resource has an internal rating of the user cluster based on whether the rating user IDs corresponding to the ratings all belong to users within the platform; When the candidate data resource contains a scoring user ID that belongs to a user within the platform, determining that the candidate data resource has an internal score of the user cluster; When the scoring user ID that does not have a rating in the candidate data resource belongs to a user within the platform, it is determined that the candidate data resource does not have an internal rating of the user cluster.
5. The cloud platform construction method of the big data alliance algorithm according to claim 4, characterized in that: The mean of the target scoring type within the platform and the mean of the target scoring type outside the platform are respectively expressed as: In the formula, represents the mean value of the target rating type in the platform, represents the mean value of the target rating type outside the platform, Indicates the number of ratings in the rating set within the platform. Indicates the number of ratings in the off-platform rating set, Indicates the first Ratings, , Indicates the first Ratings, , represents one of the rating types, ,in, They stand for {accuracy, completeness, timeliness and usability} respectively.
6. The cloud platform construction method of the big data alliance algorithm according to claim 5, characterized in that: The first data resource confirmation value is expressed as In the formula, Indicates the first data resource confirmation value, The adjustment factor representing the mean of the target rating type within the platform is set to , indicating that the influence of the mean of the target score type within the platform is greater than the influence of the mean of the target score type outside the platform by default.
7. The cloud platform construction method of the big data alliance algorithm according to claim 6, characterized in that: The classification of the off-platform data resource set based on the similarity between off-platform users and on-platform users to obtain an external similarity rating data resource set and an external irrelevant rating data resource set includes: The similarity between the off-platform users corresponding to each score in the off-platform data resource set and the on-platform users corresponding to each score in the on-platform data resource set is calculated based on the Pearson correlation coefficient: In the formula, Indicates similarity, represents a user in the platform, represents users outside the platform, n1 represents the data resource collection within the platform, and n2 represents the data resource collection outside the platform. represents the intersection data volume of the public data resources of the data resource set within the platform and the data resource set outside the platform, represents the combined data volume of the data resource set within the platform and the data resource set outside the platform, represents candidate data resources rated by both users on the platform and users outside the platform, Indicates users within the platform Candidate data resources Ratings, Indicates users outside the platform Candidate data resources Ratings, Indicates users within the platform Candidate data resources The average of all ratings of Indicates users outside the platform Candidate data resources The mean of all the ratings; Determine whether the similarity of users outside each platform is greater than a preset similarity threshold; Determine the off-platform users whose similarity is greater than the similarity threshold as off-platform similar users, and obtain the external similarity rating data resource set corresponding to the off-platform similar users; Off-platform users whose similarity is less than or equal to the similarity threshold are determined as off-platform irrelevant users, and an off-platform irrelevant rating data resource set corresponding to the off-platform irrelevant users is obtained.
8. The cloud platform construction method of the big data alliance algorithm according to claim 7, characterized in that: The second data resource confirmation value is expressed as: In the formula, represents the second data resource confirmation value, represents the adjustment factor of the mean of the external similarity target score type, which is set to , indicating that the influence of the mean of the external similar target score type is greater than the influence of the mean of the external irrelevant target score type by default.
9. The cloud platform construction method of the big data alliance algorithm according to claim 8, characterized in that: The data resource weight matrix is expressed as: In the formula, represents the data resource confirmation weight matrix, Indicates The candidate data resource The weight value of each scoring type, , Represents the total amount of data in the candidate data resource set, .
10. A cloud platform construction system for big data alliance algorithm, characterized in that: The system comprises: A group preference collection module is used to obtain user information of a user cluster in a target cloud platform, obtain a user information set, perform a feature extraction operation based on user behavior on the user information set, obtain a user behavior information feature sequence set, perform a clustering operation based on preference categories on the user behavior information feature sequence set, obtain individual user clusters, identify the user proportion and preference type of the user clusters, and obtain a group preference set; The alliance big data acquisition module is used to extract data resources based on the group preference set from the pre-built alliance database to obtain a candidate data resource set, and determine whether each candidate data resource in the candidate data resource set has an internal score of the user cluster, wherein the internal score is a score set of various scoring types including accuracy, completeness, timeliness and ease of use, and determine the candidate data resources in the candidate data resource set that have the internal score as the platform-based data resource set, and determine the candidate data resources that do not have the internal score as the platform-based data resource set; The platform data resource right confirmation module is used to identify the platform scoring set and the platform scoring set corresponding to the target scoring type of the target data resource in the platform data resource set, and calculate the mean of the platform scoring set to obtain the platform target scoring type mean, and calculate the mean of the platform scoring set to obtain the platform target scoring type mean, and perform weighted calculation on the platform target scoring type mean and the platform target scoring type mean to obtain the first data resource right confirmation value corresponding to the platform data resource set; The off-platform data resource rights confirmation module is used to classify the off-platform data resource set based on the similarity between off-platform users and on-platform users, obtain an external similarity scoring data resource set and an external irrelevant scoring data resource set, and calculate the external similarity target scoring type mean and the external irrelevant target scoring type mean according to the external similarity scoring data resource set and the external irrelevant scoring data resource set, and perform weighted calculation on the external similarity target scoring type mean and the external irrelevant target scoring type mean to obtain a second data resource rights confirmation value corresponding to the off-platform data resource set; An alliance preference weighting module, used to concatenate the first data resource confirmation weight value and the second data resource confirmation weight value to obtain a data resource confirmation weight value matrix, and to perform weighted calculation on the candidate data resource set and the data resource confirmation weight value matrix to obtain an alliance weighted data resource set, and to perform weighted calculation on the corresponding preference labels of each data resource in the alliance weighted data resource set according to the group preference set to obtain an alliance preference weighted data resource set; The service building module is used to screen the alliance preference weighted data resource set according to preset business activity requirements, obtain business-usable data, and use the business-usable data to build business services in the target cloud platform.
Citation Information
Patent Citations
Deep recommendation model fusing comment timeliness and credible neighbor score influence
CN116502610A
Resource recommendation method and device, equipment and medium
CN117216358A
