CN110990698A - Recommendation model construction method and device - Google Patents
Recommendation model construction method and device Download PDFInfo
- Publication number
- CN110990698A CN110990698A CN201911204928.0A CN201911204928A CN110990698A CN 110990698 A CN110990698 A CN 110990698A CN 201911204928 A CN201911204928 A CN 201911204928A CN 110990698 A CN110990698 A CN 110990698A
- Authority
- CN
- China
- Prior art keywords
- sample data
- clustering
- recommendation
- recommendation model
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a recommendation model construction method and device, and belongs to the technical field of computers. The method comprises the following steps: the method comprises the steps of receiving a first sample data set respectively sent by a plurality of second servers, carrying out decryption processing on the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets, obtaining a plurality of sample data of at least one same user in the plurality of second sample data sets according to a preset rule, merging the plurality of sample data of each same user to obtain a target sample data set, and constructing a recommendation model according to the target sample data set. By arranging the third-party server, the sample data of each enterprise is encrypted and then sent to the third party for processing to obtain a target sample data set, the third party constructs a recommendation model through the target sample data set, and the data of each enterprise is shared through the third-party server, so that the data barrier between enterprises is broken, more accurate recommendation models can be obtained, and the recommendation accuracy is improved.
Description
Technical Field
The invention belongs to the technical field of computers, and particularly relates to a recommendation model construction method and device.
Background
With the development of computer technology, how to recommend products meeting the user requirements for users and improve recommendation efficiency become the focus of attention of various industries.
A large amount of user information is collected, a recommendation model is built according to the large amount of user information, and information is recommended for users through the recommendation model, so that a conventional recommendation method is provided. The construction of the recommendation model requires a large amount of user information, and the more information is collected, the more accurate the recommendation model is obtained. However, for the reasons of protecting user privacy and enterprise data security, data sharing cannot be performed between enterprises having different information of users. An accurate recommendation model cannot be constructed based on unilateral user information, and an accurate recommendation result cannot be obtained.
Disclosure of Invention
The invention provides a recommendation model construction method and device, which are used for solving the problems that data cannot be shared among enterprises to a certain extent and an accurate recommendation model cannot be constructed.
According to a first aspect of the present invention, there is provided a recommendation model building method applied to a first server, the method including:
receiving first sample data sets respectively sent by a plurality of second servers, wherein the first sample data sets are obtained after the second servers encrypt second sample data sets according to a preset encryption algorithm;
decrypting the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets;
obtaining a plurality of sample data of at least one same user in a plurality of second sample data sets according to a preset rule;
merging the plurality of sample data of each same user to obtain a target sample data set;
and constructing a recommendation model according to the target sample data set.
According to a second aspect of the present invention, there is provided a recommendation model building apparatus, disposed on a first server, the apparatus including:
the receiving module is used for receiving a first sample data set respectively sent by a plurality of second servers, wherein the first sample data set is obtained after the second servers encrypt a second sample data set according to a preset encryption algorithm;
the decryption module is used for decrypting the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets;
an obtaining module, configured to obtain multiple sample data of at least one same user in multiple second sample data sets;
the merging module is used for merging the plurality of sample data of each same user to obtain a target sample data set;
and the construction module is used for constructing a recommendation model according to the target sample data set.
The recommendation model construction method provided by the embodiment of the invention is applied to a first server and comprises the following steps: the method comprises the steps of receiving a first sample data set respectively sent by a plurality of second servers, carrying out decryption processing on the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets, obtaining a plurality of sample data of at least one same user in the plurality of second sample data sets according to a preset rule, merging the plurality of sample data of each same user to obtain a target sample data set, and constructing a recommendation model according to the target sample data set. By arranging the third-party server, the sample data of each enterprise is encrypted and then sent to the third party for processing to obtain a target sample data set, the third party constructs a recommendation model through the target sample data set, and the constructed recommendation model can be used for recommending relevant information for a user. Data sharing of the data of each enterprise is achieved through the third-party server, data barriers among the enterprises are broken, a more accurate recommendation model can be obtained, and recommendation accuracy is improved.
The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
FIG. 1 is a flowchart illustrating steps of a recommendation model construction method according to an embodiment of the present invention;
FIG. 2 is a system architecture diagram of a recommendation model construction provided by an embodiment of the present invention;
FIG. 3 is a flowchart illustrating steps of another method for constructing a recommendation model according to an embodiment of the present invention;
fig. 4 is a block diagram of a recommendation model building apparatus according to an embodiment of the present invention.
Detailed Description
Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
At present, in order to protect privacy of users and data security of enterprises, data sharing cannot be performed between enterprises in different industries, and for the same user, enterprises in different industries may collect different information of the user. For example, with a 4S store as a user, for the same 4S store (an automatic Sales service 4S store), a bank may collect information such as credit line, asset liability, Sales volume, and per-capita profit of the 4S store, and an insurance company may collect information such as an exhibition area, an exhibition insurance line, and a vehicle insurance line of the 4S store. Data sharing cannot be carried out between the bank and the insurance company, the bank cannot recommend loan products meeting the requirements for the 4S shop by using the data of the insurance company, and the insurance company cannot recommend insurance products meeting the requirements for the 4S shop by using the data of the bank. While 4S stores with the same amount of credit and asset load may have the same insurance requirements, 4S stores with the same amount of insurance may have the same loan requirements. Due to the fact that data cannot be shared, the bank and the insurance company cannot use data of the other side, products can be recommended to the 4S shop more accurately, more user requirements are mined, and benefits are improved.
Fig. 1 is a flowchart of steps of a recommendation model construction method according to an embodiment of the present invention, and the recommendation model construction method according to the embodiment is suitable for construction of a recommendation model and can improve accuracy of the recommendation model. The recommendation model construction method provided by this embodiment may be executed by a recommendation model construction device, and the recommendation model construction device may be disposed in the first server. The recommendation model building apparatus is typically implemented in software and/or hardware, and as shown in fig. 1, the method may include:
and 101, receiving first sample data sets respectively sent by a plurality of second servers.
And the first sample data set is obtained after the second server encrypts the second sample data set according to a preset encryption algorithm. Referring to fig. 2, fig. 2 is a system structure diagram of a recommendation model construction according to an embodiment of the present invention, and as shown in fig. 2, a first server may receive a first sample data set respectively sent by a plurality of second servers, where the second servers are data centers of sample data generators, such as banks, insurance companies, and securities companies. For convenience of description, in this example, a data center of an enterprise, such as a bank, an insurance company, a security company, and the like, which is easy to distinguish and collects different information of users is selected, and a person skilled in the art may select the second server according to a requirement in actual use, which is not limited in this embodiment.
Banks, insurance companies, security companies and the like can generate a plurality of sample data with unified standards according to the user information stored in the data center and the user information is required to be generated, and the second sample data set is obtained through combination. The standard of the sample data may be set according to a requirement, which is not limited in this embodiment, and the process of generating the sample data with the unified standard may refer to the prior art, which is not described herein again.
In this embodiment, each second server may encrypt the respective generated second sample data set by using a preset encryption algorithm to obtain the first sample data set. For example, the second sample data set X of the bank may include three sample data a1, B1 and C1, where the sample data a1 is sample data of 4S store a, the sample data B1 is sample data of 4S store B, the sample data C1 is sample data of 4S store C, and each sample data includes data such as a credit line, employee number, loan amount, and the like of each 4S store. Similarly, the second sample data set Y of the insurance company may include sample data a2, B2, C2 and D2 of the 4S store A, B, C and D, and each sample data includes an exhibition hall area and an exhibition hall insurance amount of each 4S store. Sample data A3, B3, C3, D3 and E3 of 4S stores A, B, C, D and E may be included in the second sample data set Z of the security company, and each sample number includes data of a security trade amount and a holdup total amount of each 4S store. Each second server (data centers of banks, insurance companies and securities companies) can encrypt the respective second sample data set by adopting an asymmetric encryption algorithm, for example, each second server can respectively generate a public key and a private key, encrypt the respective second sample data set by using the public key to obtain a first sample data set (namely the encrypted second sample data set), and send the first sample data set to the first server. At the same time, the private key is hosted to the first server. The process of encrypting using asymmetric encryption algorithm can refer to the prior art, and the embodiment will not be described in detail here.
And 102, decrypting the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets.
In this embodiment, the first server may receive the first sample data sets respectively sent by each second server to obtain a plurality of first sample data sets, and decrypt each first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets.
In connection with step 101, the first server may receive the second sample data sets X, Y and Z encrypted by the public key and sent by the bank, the insurance company, and the securities company, respectively, and the first server may decrypt the encrypted second sample data sets X, Y and Z by using the private key hosted by each second server to obtain the second sample data sets X, Y and Z. The process of decrypting by the first server using the private key can refer to the prior art, and the embodiment will not be described in detail here.
In practical use, the specific encryption and decryption algorithms may be set according to requirements, for example, a homomorphic encryption algorithm may also be used to encrypt and decrypt the sample data set, which is not limited in this example.
In this embodiment, after the first server decrypts the first sample data set to obtain a plurality of second sample data sets, the plurality of second sample data sets, and a plurality of sample data of at least one same user may be determined according to a preset rule. For example, multiple sample data for at least one same user may be determined from an organization code in the sample data. A. B and C may respectively represent organization codes of corresponding 4S stores, may determine each second sample data set, each sample data respectively corresponding to an organization code, determine organization codes commonly included in the plurality of second sample data sets, and determine sample data respectively corresponding to each organization code in the plurality of second sample data sets. As the second sample data set X, Y and Z collectively include the organisational codes A, B and C, sample data a1, a2 and A3 for the 4S store a in the second sample data set X, Y and Z respectively may be determined, as may a plurality of sample data B1, B2 and B3 for the 4S store B, and a plurality of sample data C1, C2 and C3 for the 4S store C.
The preset rule may be set according to a requirement, for example, a plurality of sample data of at least one same user may be determined according to a user name, a tax payment identification number, and the like in the sample data, which is not limited in this embodiment.
And step 104, merging a plurality of sample data of each same user to obtain a target sample data set.
In this embodiment, the first server determines a plurality of sample data of at least one same user, may combine the plurality of sample data of each user to obtain target sample data of the user, and combines the target sample data of the plurality of users to obtain a target sample data set. In combination with step 103, after determining the user a, the first server may merge sample data a1, a2, and A3 of the user a to obtain target sample data a0 of the user a, where the target sample data a0 includes a credit line, employee number, loan amount, area of the exhibition hall, insurance amount of the exhibition hall, stock exchange amount, and total holding share of the user a. Similarly, target sample data B0 of user B and target sample data C0 of user C may be obtained, and target sample data a0, B0 and C0 constitute a target sample data set.
And 105, constructing a recommendation model according to the target sample data set.
After the first server obtains the target sample data set, a recommendation model can be constructed according to the target sample data set. With reference to step 104, the first server may construct the recommendation model according to the target sample data a0, B0, and C0 in the target sample data set, and the process of constructing the recommendation model by the first server according to the target sample data may be set according to requirements, which is not limited in this embodiment.
It should be noted that, for convenience of understanding, fewer sample data are selected in this embodiment for description, and those skilled in the art can understand that in the construction process of the recommendation model, the number of sample data may be set according to requirements.
In summary, the recommendation model construction method provided in the embodiment of the present invention is applied to a first server, and includes: the method comprises the steps of receiving a first sample data set respectively sent by a plurality of second servers, carrying out decryption processing on the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets, obtaining a plurality of sample data of at least one same user in the plurality of second sample data sets according to a preset rule, merging the plurality of sample data of each same user to obtain a target sample data set, and constructing a recommendation model according to the target sample data set. By arranging the third-party server, the sample data of each enterprise is encrypted and then sent to the third-party server for processing to obtain a target sample data set, the third-party server constructs a recommendation model according to the target sample data set, and the constructed recommendation model can be used for recommending relevant information for a user. Data sharing of the data of each enterprise is achieved through a third party, data barriers among the enterprises are broken, a more accurate recommendation model can be obtained, and recommendation accuracy is improved.
Fig. 3 is a flowchart of steps of another recommendation model building method provided in an embodiment of the present invention, and as shown in fig. 3, the method may include:
And the first sample data set is obtained after the second server encrypts the second sample data set according to a preset encryption algorithm.
And 302, the first server decrypts the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets.
And step 304, the first server merges a plurality of sample data of each same user to obtain a target sample data set.
And 305, clustering the target sample data set by the first server to obtain a first clustering result.
In this embodiment, clustering the target sample data set to obtain the first clustering result may be implemented in the following manner:
clustering target sample data by adopting a first clustering algorithm to obtain a second clustering result;
and clustering the second clustering result by adopting a second clustering algorithm to obtain a first clustering result.
In this example, after the target sample data set is obtained, a first clustering algorithm may be first used to perform first clustering on sample data in the target sample data set, so as to divide the sample data in the target sample data set. For example, the first clustering algorithm may be a hierarchical clustering algorithm, and the hierarchical clustering algorithm may be used for performing the first clustering, in the hierarchical clustering process, first, each sample data in the target sample data set is used as one class, the distance between two sample data is calculated by using the euclidean distance, the two sample data with small distance values are merged into one class, the above steps are repeatedly performed, the second clustering result is obtained, and the sample data in the target sample data set is divided into multiple classes.
After the first clustering, a second clustering may be performed on the second clustered results using a second clustering algorithm. For example, the second clustering algorithm may be a k-means clustering algorithm, and after the first clustering, the k-means clustering algorithm is used to cluster the second clustering result again, and the sample data in the target sample data set is refined and classified again. For example, in combination with the example in the first embodiment, three indexes, namely a brand of a 4S store, an area of an exhibition hall, and an operating profit, may be respectively selected for each category of the first clustering result, and a k-means clustering algorithm is used to perform clustering, so as to obtain the first clustering result. The specific calculation processes of the hierarchical clustering algorithm and the k-means clustering algorithm can be set according to requirements, and the embodiment does not limit the calculation processes.
The specific algorithms of the first clustering algorithm and the second clustering algorithm can be selected according to requirements, the target sample data is clustered twice, the sample data in the target sample data set can be finely divided, and the recommendation precision is improved.
And step 306, the first server performs association analysis on each class in the first clustering result to obtain an association rule, and a recommendation model is constructed according to the first clustering result and the association rule.
In this example, after the first server performs clustering to obtain the first clustering result, the first server may perform association analysis on each class in the first clustering result to obtain an association rule, and construct the recommendation model according to the first clustering result and the association rule. For example, a correlation analysis algorithm may be used to perform a correlation analysis on sample data in the target sample data set. By way of example in the first embodiment, for a certain class in the first clustering result, the feature factor tag of each 4S store may be set according to the target sample data of each 4S store. If the feature label "collateral" exists in the sample data of 4S store A, B or C in this class, the feature factor label that "collateral exists" is 4S store may be set. The support (S) of the feature factor tag "with mortgage" in the 4S-like stores is calculated, for example, S ═ 4S store number with mortgage feature }/{ such user number }, and features with S greater than or equal to 0.5 are extracted as a frequent item set. And (4) taking combined characteristic factors in the frequent item set, for example, characteristic factors of 'collateral' and 'credit line' as combined characteristic factors, calculating the support degree (S), and screening out a combined item set with S being greater than or equal to 0.5. Creating an association rule, such as an association rule of { mortgage } - > { credit line }, using a Bayesian formula:
and calculating the confidence (T), and screening out the association rules with the confidence greater than or equal to 0.8. And constructing a recommendation model according to the first clustering result and the association rule with the confidence coefficient being greater than or equal to 0.8, and recommending related information for the 4S store user. The specific values of the support degree and the confidence degree may be set according to requirements, and the process of specifically using the association analysis algorithm may be set according to requirements, which is not limited in this embodiment.
Optionally, the first server may further obtain a plurality of first recommendation results by using a collaborative filtering algorithm according to each class in the first clustering results, and construct a recommendation model according to the first clustering results, the association rules, and the plurality of first recommendation results.
In this embodiment, after the first server obtains the first clustering result, each class in the first clustering result may also be analyzed by using a collaborative filtering algorithm, so as to obtain a first recommendation result. For example, the similarity between 4S stores may be calculated using cosine similarity:
for example, in a certain category of the first clustering result, the brand of the 4S store a is bmw, the area of the exhibition hall is 1000 square meters, the number of staff is 100, and there is a business: the loan amount is 1000 thousands and the insurance limit of the exhibition hall is 500 thousands. At this time, the brand of the 4S shop B is bmw, the area of the exhibition hall is 1000 square meters, and the number of employees is 100, and it can be analyzed that the 4S shop B has the same demand as the 4S shop a in terms of loan and the insurance amount of the exhibition hall. The first recommendation result regarding the 4S shop B is obtained, i.e., the loan amount of 1000 ten thousand and the extension room insurance line of 500 ten thousand. After the plurality of first recommendation results are obtained through calculation, a recommendation model can be constructed according to the first clustering results, the association rules and the first recommendation results. The prior art may be referred to in the specific process of obtaining the first recommendation result by using the collaborative filtering algorithm, which is not limited in this embodiment.
The process of constructing the recommendation model according to the first clustering result and the association rule, and according to the first clustering result, the association rule, and the first recommendation result by the first server may refer to the prior art, which is not described in detail in this embodiment.
In this embodiment, the first server may send the recommendation model to the second server, so that the second server runs the recommendation model, and the second server may obtain, according to the second user information, a third recommendation result matching the second user information through the recommendation model. For example, as shown in fig. 2, if the second server is a server of an insurance company, and the second user information is 4S shop H, the area of the exhibition hall is 1000 square meters, and the number of employees is 100, the second server inputs the second user information into the recommendation model, and can output a third recommendation result of 500 ten thousand of insurance lines of the exhibition hall, and recommend a business of 500 ten thousand of insurance lines of the exhibition hall to the 4S shop H. The process of the second server obtaining the third recommendation result matching the second user information through the recommendation model may refer to the prior art.
In summary, the recommendation model construction method provided in the embodiment of the present invention is applied to a first server, and includes: the method comprises the steps of receiving a first sample data set respectively sent by a plurality of second servers, carrying out decryption processing on the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets, obtaining a plurality of sample data of at least one same user in the plurality of second sample data sets according to a preset rule, merging the plurality of sample data of each same user to obtain a target sample data set, and constructing a recommendation model according to the target sample data set. By arranging the third-party server, the sample data of each enterprise is encrypted and then sent to the third party for processing to obtain a target sample data set, the third party constructs a recommendation model through the target sample data set, and the constructed recommendation model can be used for recommending relevant information for a user. Data sharing of the data of each enterprise is achieved through a third party, data barriers among the enterprises are broken, a more accurate recommendation model can be obtained, and recommendation accuracy is improved.
Optionally, the first server may receive the first user information sent by the second server after constructing the recommendation model according to the first clustering result;
obtaining a second recommendation result matched with the first user information through a recommendation model according to the first user information;
and sending the second recommendation result to the second server.
In this embodiment, after the recommendation model is constructed, the first server may run the recommendation model, receive the first user information sent by any one of the second servers, obtain a second recommendation result matching the first user information through the recommendation model, and send the second recommendation result to the second server. For example, referring to fig. 2, when a bank sends first user information 4S store L to a first server, the area of an exhibition hall is 1000 square meters, the first server receives the first user information, and obtains a second recommendation result, namely the loan amount is 1000 ten thousand through a recommendation model, so as to excavate loan requirements of the 4S store L for the bank. The second server sends the first user information to the first server, and the process of sending the second user information to the first server after the second server obtains the second recommendation result may refer to the prior art, which is not limited in this embodiment.
Fig. 4 is a block diagram of a recommendation model building apparatus according to an embodiment of the present invention, and as shown in fig. 4, the apparatus 400 may include: a receiving module 401, a decryption module 402, an obtaining module 403, a merging module 404 and a building module 405.
The receiving module 401 is configured to receive first sample data sets respectively sent by multiple second servers, where the first sample data sets are obtained by the second servers after encrypting second sample data sets according to a preset encryption algorithm.
The decryption module 402 is configured to decrypt the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets.
The obtaining module 403 is configured to obtain a plurality of sample data of at least one same user in a plurality of second sample data sets.
The merging module 404 is configured to merge multiple sample data of each same user to obtain a target sample data set.
The building module 405 is configured to build a recommendation model according to the target sample data set.
In summary, the recommendation model construction device provided in the embodiment of the present invention is disposed in the first server, and by disposing the third-party server, the sample data of each enterprise is encrypted and then sent to the third party for processing, so as to obtain the target sample data set, and the third party constructs the recommendation model through the target sample data set, and can recommend relevant information to the user by using the constructed recommendation model. Data sharing of the data of each enterprise is achieved through the third-party server, data barriers among the enterprises are broken, a more accurate recommendation model can be obtained, and recommendation accuracy is improved.
Optionally, the building module 405 may include: the device comprises a clustering unit, a first analysis unit and a construction unit.
The clustering unit is used for clustering the target sample data set to obtain a first clustering result.
The first analysis unit is used for performing association analysis on each class in the first clustering result to obtain an association rule.
The construction unit is used for constructing a recommendation model according to the first clustering result and the association rule.
Optionally, the building module 405 may further include: a second analysis unit.
The second analysis unit is used for obtaining a plurality of first recommendation results by adopting a collaborative filtering algorithm according to each class in the first clustering results.
The construction unit is specifically configured to construct the recommendation model according to the first clustering result, the association rule, and the plurality of first recommendation results.
Optionally, the clustering unit is specifically configured to cluster the target sample data by using a first clustering algorithm to obtain a second clustering result; and clustering the second clustering result by adopting a second clustering algorithm to obtain a first clustering result.
Optionally, the apparatus 400 may further include: and a recommendation module.
The recommendation module is used for receiving the first user information sent by the second server, obtaining a second recommendation result matched with the first user information through the recommendation model according to the first user information, and sending the second recommendation result to the second server.
Optionally, the apparatus 400 may further include: and a sending module.
The sending module is used for sending the recommendation model to the second server so that the second server can obtain a third recommendation result matched with the second user information through the recommendation model according to the second user information.
For the above device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and for the relevant points, refer to the partial description of the method embodiment.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
As is readily imaginable to the person skilled in the art: any combination of the above embodiments is possible, and thus any combination between the above embodiments is an embodiment of the present invention, but the present disclosure is not necessarily detailed herein for reasons of space.
The operations performed herein are not inherently related to any particular computer, virtual machine system, or other apparatus. Various general purpose systems may also be used with the teachings herein. The structure required to construct a system incorporating aspects of the present invention will be apparent from the description above. Moreover, the present invention is not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose the best mode of the invention.
In the description provided herein, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the invention and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be interpreted as reflecting an intention that: that the invention as claimed requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
Those skilled in the art will appreciate that the modules in the device in an embodiment may be adaptively changed and disposed in one or more devices different from the embodiment. The modules or units or components of the embodiments may be combined into one module or unit or component, and furthermore they may be divided into a plurality of sub-modules or sub-units or sub-components. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or elements are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the claims, any of the claimed embodiments may be used in any combination.
The various component embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or Digital Signal Processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components of an operation execution method according to an embodiment of the present invention. The present invention may also be embodied as apparatus or device programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present invention may be stored on computer-readable media or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The usage of the words first, second and third, etcetera do not indicate any ordering. These words may be interpreted as names.
Claims (12)
1. A recommendation model building method applied to a first server, the method comprising:
receiving first sample data sets respectively sent by a plurality of second servers, wherein the first sample data sets are obtained after the second servers encrypt second sample data sets according to a preset encryption algorithm;
decrypting the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets;
obtaining a plurality of sample data of at least one same user in a plurality of second sample data sets according to a preset rule;
merging the plurality of sample data of each same user to obtain a target sample data set;
and constructing a recommendation model according to the target sample data set.
2. The method of claim 1, wherein said building a recommendation model from said target sample data set comprises:
clustering the target sample data set to obtain a first clustering result;
performing association analysis on each class in the first clustering result to obtain an association rule;
and constructing the recommendation model according to the first clustering result and the association rule.
3. The method of claim 2, further comprising:
obtaining a plurality of first recommendation results by adopting a collaborative filtering algorithm according to each class in the first clustering results;
the constructing the recommendation model according to the first clustering result and the association rule includes:
and constructing the recommendation model according to the first clustering result, the association rule and the plurality of first recommendation results.
4. The method according to claim 2, wherein said clustering said target sample data set to obtain a first clustering result comprises:
clustering the target sample data by adopting a first clustering algorithm to obtain a second clustering result;
and clustering the second clustering result by adopting a second clustering algorithm to obtain the first clustering result.
5. The method of claim 1, further comprising, after the building a recommendation model according to the first clustering result:
receiving first user information sent by the second server;
according to the first user information, obtaining a second recommendation result matched with the first user information through the recommendation model;
and sending the second recommendation result to the second server.
6. The method of claim 1, further comprising, after the building a recommendation model according to the first clustering result:
and sending the recommendation model to the second server so that the second server obtains a third recommendation result matched with the second user information through the recommendation model according to the second user information.
7. A recommendation model building apparatus, provided on a first server, the apparatus comprising:
the receiving module is used for receiving a first sample data set respectively sent by a plurality of second servers, wherein the first sample data set is obtained after the second servers encrypt a second sample data set according to a preset encryption algorithm;
the decryption module is used for decrypting the first sample data set according to a preset decryption algorithm to obtain a plurality of second sample data sets;
an obtaining module, configured to obtain multiple sample data of at least one same user in multiple second sample data sets;
the merging module is used for merging the plurality of sample data of each same user to obtain a target sample data set;
and the construction module is used for constructing a recommendation model according to the target sample data set.
8. The apparatus of claim 7, wherein the building module comprises:
the clustering unit is used for clustering the target sample data set to obtain a first clustering result;
the first analysis unit is used for carrying out association analysis on each class in the first clustering result to obtain an association rule;
and the construction unit is used for constructing the recommendation model according to the first clustering result and the association rule.
9. The apparatus of claim 8, wherein the building module further comprises:
the second analysis unit is used for obtaining a plurality of first recommendation results by adopting a collaborative filtering algorithm according to each class in the first clustering results;
the building unit is specifically configured to build the recommendation model according to the first clustering result, the association rule, and the plurality of first recommendation results.
10. The apparatus according to claim 8, wherein the clustering unit is specifically configured to cluster the target sample data by using a first clustering algorithm to obtain a second clustering result; and clustering the second clustering result by adopting a second clustering algorithm to obtain the first clustering result.
11. The apparatus of claim 7, further comprising: the recommendation module is used for receiving the first user information sent by the second server; according to the first user information, obtaining a second recommendation result matched with the first user information through the recommendation model; and sending the second recommendation result to the second server.
12. The apparatus of claim 7, further comprising: and the sending module is used for sending the recommendation model to the second server so that the second server can obtain a third recommendation result matched with the second user information through the recommendation model according to the second user information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911204928.0A CN110990698B (en) | 2019-11-29 | 2019-11-29 | Recommendation model construction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911204928.0A CN110990698B (en) | 2019-11-29 | 2019-11-29 | Recommendation model construction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110990698A true CN110990698A (en) | 2020-04-10 |
| CN110990698B CN110990698B (en) | 2021-01-08 |
Family
ID=70088652
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911204928.0A Active CN110990698B (en) | 2019-11-29 | 2019-11-29 | Recommendation model construction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110990698B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111371544A (en) * | 2020-05-27 | 2020-07-03 | 支付宝(杭州)信息技术有限公司 | Prediction method and device based on homomorphic encryption, electronic equipment and storage medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103164474A (en) * | 2011-12-15 | 2013-06-19 | 中国移动通信集团贵州有限公司 | Method for analyzing data service |
| US8484099B1 (en) * | 2012-02-06 | 2013-07-09 | Amazon Technologies, Inc. | Method, medium, and system for behavior-based recommendations of product upgrades |
| CN104750856A (en) * | 2015-04-16 | 2015-07-01 | 天天艾米（北京）网络科技有限公司 | System and method for multi-dimensional synergic recommendation |
| CN105224623A (en) * | 2015-09-22 | 2016-01-06 | 北京百度网讯科技有限公司 | The training method of data model and device |
| CN105608219A (en) * | 2016-01-07 | 2016-05-25 | 上海通创信息技术有限公司 | Stream-oriented recommended engine, recommendation system and recommendation method based on clustering |
| CN106407349A (en) * | 2016-09-06 | 2017-02-15 | 北京三快在线科技有限公司 | Product recommendation method and device |
| CN108256119A (en) * | 2018-02-14 | 2018-07-06 | 北京方正阿帕比技术有限公司 | A kind of construction method of resource recommendation model and the resource recommendation method based on the model |
| CN108830731A (en) * | 2018-07-03 | 2018-11-16 | 广州唯品会信息科技有限公司 | Financial product user account correlating method, device, system and readable storage medium storing program for executing |
| CN109165249A (en) * | 2018-08-07 | 2019-01-08 | 阿里巴巴集团控股有限公司 | Data processing model construction method, device, server and user terminal |
| CN110163647A (en) * | 2019-03-14 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN110263250A (en) * | 2019-06-12 | 2019-09-20 | 深圳前海微众银行股份有限公司 | A kind of generation method and device of recommended models |
| US20190303768A1 (en) * | 2016-12-30 | 2019-10-03 | Huawei Technologies Co., Ltd. | Community Question Answering-Based Article Recommendation Method, System, and User Device |
-
2019
- 2019-11-29 CN CN201911204928.0A patent/CN110990698B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103164474A (en) * | 2011-12-15 | 2013-06-19 | 中国移动通信集团贵州有限公司 | Method for analyzing data service |
| US8484099B1 (en) * | 2012-02-06 | 2013-07-09 | Amazon Technologies, Inc. | Method, medium, and system for behavior-based recommendations of product upgrades |
| CN104750856A (en) * | 2015-04-16 | 2015-07-01 | 天天艾米（北京）网络科技有限公司 | System and method for multi-dimensional synergic recommendation |
| CN105224623A (en) * | 2015-09-22 | 2016-01-06 | 北京百度网讯科技有限公司 | The training method of data model and device |
| CN105608219A (en) * | 2016-01-07 | 2016-05-25 | 上海通创信息技术有限公司 | Stream-oriented recommended engine, recommendation system and recommendation method based on clustering |
| CN106407349A (en) * | 2016-09-06 | 2017-02-15 | 北京三快在线科技有限公司 | Product recommendation method and device |
| US20190303768A1 (en) * | 2016-12-30 | 2019-10-03 | Huawei Technologies Co., Ltd. | Community Question Answering-Based Article Recommendation Method, System, and User Device |
| CN108256119A (en) * | 2018-02-14 | 2018-07-06 | 北京方正阿帕比技术有限公司 | A kind of construction method of resource recommendation model and the resource recommendation method based on the model |
| CN108830731A (en) * | 2018-07-03 | 2018-11-16 | 广州唯品会信息科技有限公司 | Financial product user account correlating method, device, system and readable storage medium storing program for executing |
| CN109165249A (en) * | 2018-08-07 | 2019-01-08 | 阿里巴巴集团控股有限公司 | Data processing model construction method, device, server and user terminal |
| CN110163647A (en) * | 2019-03-14 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN110263250A (en) * | 2019-06-12 | 2019-09-20 | 深圳前海微众银行股份有限公司 | A kind of generation method and device of recommended models |
Non-Patent Citations (2)
| Title |
|---|
| LU XING-HUA ETC: "A Hybrid Recommendation Model for Community Attributes of Social Networks Based on Association Rule Mining", 《2018 3RD INTERNATIONAL CONFERENCE ON MECHANICAL, CONTROL AND COMPUTER ENGINEERING》 * |
| 梁丽君: "基于用户属性聚类的协同过滤推荐算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111371544A (en) * | 2020-05-27 | 2020-07-03 | 支付宝(杭州)信息技术有限公司 | Prediction method and device based on homomorphic encryption, electronic equipment and storage medium |
| CN111371544B (en) * | 2020-05-27 | 2020-09-08 | 支付宝(杭州)信息技术有限公司 | Prediction method and device based on homomorphic encryption, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110990698B (en) | 2021-01-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Perera et al. | Blockchain technology: Is it hype or real in the construction industry? | |
| Sun Yin et al. | Regulating cryptocurrencies: a supervised machine learning approach to de-anonymizing the bitcoin blockchain | |
| Rejeb et al. | Blockchain technology in tourism: applications and possibilities | |
| Badzar | Blockchain for securing sustainable transport contracts and supply chain transparency-An explorative study of blockchain technology in logistics | |
| Bagloee et al. | Blockchain: The operating system of smart cities | |
| Badzar | Blockchain for securing sustainable transport contracts and supply chain transparency | |
| US11803823B2 (en) | Systems and methods for blockchain-based payment transactions, alerts, and dispute settlement, using a blockchain interface server | |
| US20060085370A1 (en) | System for identifying data relationships | |
| US11861619B1 (en) | Systems and methods for payment transactions, alerts, dispute settlement, and settlement payments, using multiple blockchains | |
| Bloomrosen et al. | Advancing the framework: use of health data—a report of a working conference of the American Medical Informatics Association | |
| Shah et al. | Government big data ecosystem: definitions, types of data, actors, and roles and the impact in public administrations | |
| CN110729028A (en) | Health data processing method and device, electronic equipment and storage medium | |
| CN111178005B (en) | Data processing system, method and storage medium | |
| Sarker et al. | Promoting information resource management for e-government through big data approach | |
| Elbashbishy et al. | Blockchain technology in the construction industry: mapping current research trends using social network analysis and clustering | |
| Yang et al. | How does the perceived physical risk of COVID-19 affect sharing economy services? | |
| US20240095734A1 (en) | Embedded data transaction exchange platform | |
| Treiblmaier | Blockchain and tourism: Paradoxes, misconceptions, and a research roadmap | |
| CN106156130A (en) | A kind of data processing method and device | |
| Dindarian et al. | Traceability of electronic waste using blockchain technology | |
| Saeed et al. | A systematic review of digital technology and innovation and its potential to address anti-corruption, transparency, and accountability in the pharmaceutical supply chain | |
| CN110990698B (en) | Recommendation model construction method and device | |
| Kim et al. | Developmental trajectories in blockchain technology using patent-based knowledge network analysis | |
| Nguyen et al. | Readiness of SMEs for adopt big data: an empirical study in Vietnam | |
| Ramazhamba et al. | Using distributed ledger technology for digital forensic investigation purposes on tendering projects |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |