CN110866181A - Resource recommendation method, device and storage medium - Google Patents
Resource recommendation method, device and storage medium Download PDFInfo
- Publication number
- CN110866181A CN110866181A CN201910970985.3A CN201910970985A CN110866181A CN 110866181 A CN110866181 A CN 110866181A CN 201910970985 A CN201910970985 A CN 201910970985A CN 110866181 A CN110866181 A CN 110866181A
- Authority
- CN
- China
- Prior art keywords
- user
- information
- resource
- resource information
- vector
- 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 relates to a data analysis technology, and discloses a resource recommendation method, a resource recommendation device and a storage medium, wherein the method comprises the following steps: acquiring first user information of a user, acquiring first resource information explicitly associated with the user, and generating a user explicit vector; acquiring second resource information, acquiring second user information of a user explicitly associated with the second resource information, and generating a resource explicit vector; acquiring implicit behavior characteristics of a user, acquiring third user information and third resource information associated with the implicit behavior characteristics, constructing a triple relation matrix, and performing decomposition calculation on the triple relation matrix to obtain a user implicit vector and a resource implicit vector of the user; calculating a first similarity between a user explicit vector of a user and a corresponding resource explicit vector, and calculating a second similarity between a user implicit vector and a corresponding resource implicit vector; and performing weighted summation, and selecting and recommending the resource information based on the result of the weighted summation. The resource recommendation method and the resource recommendation device can improve the accuracy of resource recommendation.
Description
Technical Field
The present invention relates to the field of data analysis technologies, and in particular, to a method, an apparatus, and a storage medium for resource recommendation.
Background
User portrayal and resource portrayal are important ways for improving accuracy of a recommendation system, user characteristics and resource characteristics can be fully embodied through comprehensive and accurate labels, and an individualized resource pool can be generated for a user according to characteristics formed by the portrayal, so that the effect of thousands of people is achieved, recommendation accuracy is improved, and user satisfaction is improved.
Currently, in the application of portraits in recommendation systems, two methods are used for screening resources and forming predictions: one is to adopt dominant features (e.g., similar features such as content and/or attributes) to perform resource screening, and this method generally needs to perform a large amount of feature engineering to find a suitable feature combination, and the effect of the feature combination determines the quality of the final screening and prediction effect to a certain extent, and the accuracy needs to be improved; the other method is to adopt a machine learning algorithm to calculate implicit characteristics (for example, content and/or attributes and the like are dissimilar, but certain associated characteristics exist) to perform resource screening, and under the condition of large data quantity, the method can relieve data sparsity to a certain extent, but has the characteristics of slow resource updating and low interpretability of results, and the accuracy also needs to be improved.
Disclosure of Invention
The invention aims to provide a resource recommendation method, a resource recommendation device and a storage medium, aiming at improving the accuracy of resource recommendation.
In order to achieve the above object, the present invention provides a method for resource recommendation, which comprises:
acquiring first user information of a user, acquiring first resource information explicitly associated with the user, and generating a user explicit vector based on the first user information and the first resource information;
acquiring second resource information, acquiring second user information of a user explicitly associated with the second resource information, and generating a resource explicit vector with the same dimension as the user explicit vector based on the second resource information and the second user information;
acquiring implicit behavior characteristics of a user, acquiring third user information and third resource information associated with the implicit behavior characteristics, constructing a triple relation matrix based on the third user information and the third resource information, and performing decomposition calculation on the triple relation matrix by using a predetermined algorithm to obtain a user implicit vector and a resource implicit vector of the user;
calculating a first similarity between a user explicit vector of the user and a corresponding resource explicit vector, and calculating a second similarity between the user implicit vector and a corresponding resource implicit vector;
and carrying out weighted summation on the first similarity and the second similarity, selecting resource information based on the result of the weighted summation, and recommending the resource information to the user.
Preferably, the first user information includes basic information and behavior information of a user, the first resource information includes resource information having the same or different service attributes, and the step of generating a user explicit vector based on the first user information and the first resource information specifically includes:
acquiring a predefined multidimensional vector, wherein the multidimensional vector comprises basic information and service attributes;
and assigning a value to the basic information in the multi-dimensional vector based on the basic information, assigning a value to the service attribute in the multi-dimensional vector based on the resource information, the basic information and the behavior information in the first resource information, and taking the assigned multi-dimensional vector as the user explicit vector.
Preferably, the behavior information includes an explicit behavior characteristic and an implicit behavior characteristic, and the step of assigning a value to the service attribute in the multidimensional vector based on each resource information, the basic information, and the behavior information in the first resource information specifically includes:
acquiring dominant behavior characteristics and time information generated when the user operates on each resource information in the first resource information, calculating the preference degree of the user on the corresponding resource information based on the dominant behavior characteristics and the time information, and taking the preference degree as the value of the corresponding service attribute; or
Grouping the users based on the basic information, predicting the preference degree of the users in each group to each resource information in the first resource information through an association rule in the group, and taking the preference degree as the value of the corresponding service attribute.
Preferably, the step of performing weighted summation on the first similarity and the second similarity, selecting resource information based on a result of the weighted summation, and recommending the resource information to the user specifically includes:
normalizing the first similarity and the second similarity respectively to obtain a preset weight, and carrying out weighted summation based on the normalized first similarity, the normalized second similarity and the weight to obtain a total similarity;
and acquiring the time and the heat of the resource information, selecting a plurality of resource information based on the total similarity and the time and the heat of the resource information, and recommending the resource information to the user.
Preferably, the step of selecting and recommending a plurality of resource information to the user based on the total similarity, the time to put on shelf and the degree of heat of each resource information specifically includes:
and calculating the priority of each resource information based on the total similarity, the time of putting each resource information on shelf and the heat, selecting a plurality of resource information according to the priority of each resource information, and recommending the resource information to the user.
In order to achieve the above object, the present invention further provides a resource recommendation apparatus, where the resource recommendation apparatus includes a memory and a processor connected to the memory, the memory stores therein a processing system that is executable on the processor, and when executed by the processor, the processing system implements the following steps:
acquiring first user information of a user, acquiring first resource information explicitly associated with the user, and generating a user explicit vector based on the first user information and the first resource information;
acquiring second resource information, acquiring second user information of a user explicitly associated with the second resource information, and generating a resource explicit vector with the same dimension as the user explicit vector based on the second resource information and the second user information;
acquiring implicit behavior characteristics of a user, acquiring third user information and third resource information associated with the implicit behavior characteristics, constructing a triple relation matrix based on the third user information and the third resource information, and performing decomposition calculation on the triple relation matrix by using a predetermined algorithm to obtain a user implicit vector and a resource implicit vector of the user;
calculating a first similarity between a user explicit vector of the user and a corresponding resource explicit vector, and calculating a second similarity between the user implicit vector and a corresponding resource implicit vector;
and carrying out weighted summation on the first similarity and the second similarity, selecting resource information based on the result of the weighted summation, and recommending the resource information to the user.
Preferably, the first user information includes basic information and behavior information of a user, the first resource information includes resource information having the same or different service attributes, and the step of generating a user explicit vector based on the first user information and the first resource information specifically includes:
acquiring a predefined multidimensional vector, wherein the multidimensional vector comprises basic information and service attributes;
and assigning a value to the basic information in the multi-dimensional vector based on the basic information, assigning a value to the service attribute in the multi-dimensional vector based on the resource information, the basic information and the behavior information in the first resource information, and taking the assigned multi-dimensional vector as the user explicit vector.
Preferably, the behavior information includes an explicit behavior characteristic and an implicit behavior characteristic, and the step of assigning a value to the service attribute in the multidimensional vector based on each resource information, the basic information, and the behavior information in the first resource information specifically includes:
acquiring dominant behavior characteristics and time information generated when the user operates on each resource information in the first resource information, calculating the preference degree of the user on the corresponding resource information based on the dominant behavior characteristics and the time information, and taking the preference degree as the value of the corresponding service attribute; or
Grouping the users based on the basic information, predicting the preference degree of the users in each group to each resource information in the first resource information through an association rule in the group, and taking the preference degree as the value of the corresponding service attribute.
Preferably, the step of performing weighted summation on the first similarity and the second similarity, selecting resource information based on a result of the weighted summation, and recommending the resource information to the user specifically includes:
normalizing the first similarity and the second similarity respectively to obtain a preset weight, and carrying out weighted summation based on the normalized first similarity, the normalized second similarity and the weight to obtain a total similarity;
and acquiring the time and the heat of the resource information, selecting a plurality of resource information based on the total similarity and the time and the heat of the resource information, and recommending the resource information to the user.
The invention also provides a computer readable storage medium having stored thereon a processing system, which when executed by a processor implements the steps of the method for resource recommendation described above.
The invention has the beneficial effects that: according to the method, the user explicit vector and the resource explicit vector are generated firstly, then the triple relation matrix is constructed based on the implicit behavior characteristics of the user, the triple relation matrix is decomposed to obtain the user implicit vector and the resource implicit vector, finally, the first similarity between the user explicit vector of the user and the corresponding resource explicit vector is calculated, the second similarity between the user implicit vector and the corresponding resource implicit vector is calculated, the first similarity and the second similarity are subjected to weighted summation, and the resource information is selected and recommended to the user based on the result of the weighted summation.
Drawings
FIG. 1 is a diagram illustrating a hardware architecture of an apparatus for resource recommendation according to an embodiment of the present invention;
FIG. 2 is a block diagram of a process of one embodiment of the processing system of FIG. 1;
FIG. 3 is a flowchart illustrating a resource recommendation method according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the description relating to "first", "second", etc. in the present invention is for descriptive purposes only and is not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In addition, technical solutions between various embodiments may be combined with each other, but must be realized by a person skilled in the art, and when the technical solutions are contradictory or cannot be realized, such a combination should not be considered to exist, and is not within the protection scope of the present invention.
Fig. 1 is a schematic diagram of a hardware architecture of an embodiment of the resource recommendation apparatus of the present invention. The resource recommendation device 1 is a device capable of automatically performing numerical calculation and/or information processing in accordance with a command set or stored in advance. The resource recommendation device 1 may be a computer, or may be a single network server, a server group composed of a plurality of network servers, or a cloud composed of a large number of hosts or network servers based on cloud computing, where cloud computing is one of distributed computing and is a super virtual computer composed of a group of loosely coupled computers.
In the present embodiment, the resource recommendation apparatus 1 may include, but is not limited to, a memory 11, a processor 12, and a network interface 13, which are communicatively connected to each other through a system bus, wherein the memory 11 stores a processing system capable of running on the processor 12. It is noted that fig. 1 only shows the apparatus 1 with resource recommendation of components 11-13, but it is to be understood that not all shown components are required to be implemented, and that more or less components may be implemented instead.
The storage 11 includes a memory and at least one type of readable storage medium. The memory provides cache for the operation of the resource recommendation device 1; the readable storage medium may be a non-volatile storage medium such as flash memory, a hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a Read Only Memory (ROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), a Programmable Read Only Memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the readable storage medium may be an internal storage unit of the apparatus 1 for resource recommendation, such as a hard disk of the apparatus 1 for resource recommendation; in other embodiments, the non-volatile storage medium may also be an external storage device of the resource recommendation apparatus 1, such as a plug-in hard disk provided on the resource recommendation apparatus 1, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like. In this embodiment, the readable storage medium of the memory 11 is generally used for storing an operating system and various application software installed in the resource recommendation device 1, for example, program codes of the processing system 10 in an embodiment of the present invention are stored. Further, the memory 11 may also be used to temporarily store various types of data that have been output or are to be output.
The processor 12 may be, in some embodiments, a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other data Processing chip, and is used for executing program codes stored in the memory 11 or Processing data, such as the Processing system 10.
The network interface 13 may comprise a standard wireless network interface, a wired network interface, and the network interface 13 is generally used for establishing a communication connection between the resource recommendation apparatus 1 and other electronic devices.
The processing system 10 is stored in the memory 11 and includes at least one computer-readable instruction stored in the memory 11, which is executable by the processor 12 to implement the methods of the embodiments of the present application; and the at least one computer readable instruction may be divided into different logic blocks depending on the functions implemented by the respective portions.
In one embodiment, the processing system 10 as described above, when executed by the processor 12, implements the steps of:
acquiring first user information of a user, acquiring first resource information explicitly associated with the user, and generating a user explicit vector based on the first user information and the first resource information;
acquiring second resource information, acquiring second user information of a user explicitly associated with the second resource information, and generating a resource explicit vector with the same dimension as the user explicit vector based on the second resource information and the second user information;
acquiring implicit behavior characteristics of a user, acquiring third user information and third resource information associated with the implicit behavior characteristics, constructing a triple relation matrix based on the third user information and the third resource information, and performing decomposition calculation on the triple relation matrix by using a predetermined algorithm to obtain a user implicit vector and a resource implicit vector of the user;
calculating a first similarity between a user explicit vector of the user and a corresponding resource explicit vector, and calculating a second similarity between the user implicit vector and a corresponding resource implicit vector;
and carrying out weighted summation on the first similarity and the second similarity, selecting resource information based on the result of the weighted summation, and recommending the resource information to the user.
Preferably, the first user information includes basic information and behavior information of a user, the first resource information includes resource information having the same or different service attributes, and the step of generating a user explicit vector based on the first user information and the first resource information specifically includes:
acquiring a predefined multidimensional vector, wherein the multidimensional vector comprises basic information and service attributes;
and assigning a value to the basic information in the multi-dimensional vector based on the basic information, assigning a value to the service attribute in the multi-dimensional vector based on the resource information, the basic information and the behavior information in the first resource information, and taking the assigned multi-dimensional vector as the user explicit vector.
Preferably, the behavior information includes an explicit behavior characteristic and an implicit behavior characteristic, and the step of assigning a value to the service attribute in the multidimensional vector based on each resource information, the basic information, and the behavior information in the first resource information specifically includes:
acquiring dominant behavior characteristics and time information generated when the user operates on each resource information in the first resource information, calculating the preference degree of the user on the corresponding resource information based on the dominant behavior characteristics and the time information, and taking the preference degree as the value of the corresponding service attribute; or
Grouping the users based on the basic information, predicting the preference degree of the users in each group to each resource information in the first resource information through an association rule in the group, and taking the preference degree as the value of the corresponding service attribute.
Preferably, the step of performing weighted summation on the first similarity and the second similarity, selecting resource information based on a result of the weighted summation, and recommending the resource information to the user specifically includes:
normalizing the first similarity and the second similarity respectively to obtain a preset weight, and carrying out weighted summation based on the normalized first similarity, the normalized second similarity and the weight to obtain a total similarity;
and acquiring the time and the heat of the resource information, selecting a plurality of resource information based on the total similarity and the time and the heat of the resource information, and recommending the resource information to the user.
Referring now to FIG. 2, a block diagram of a process of the processing system 10 of FIG. 1 is shown. The processing system 10 is partitioned into a plurality of modules that are stored in the memory 12 and executed by the processor 13 to accomplish the present invention. The modules referred to herein are referred to as a series of computer program instruction segments capable of performing specified functions.
The processing system 10 may be divided into: a first generation module 101, a second generation module 102, a decomposition module 103, a calculation module 104 and a recommendation module 105.
The first generating module 101 is configured to obtain first user information of a user, obtain first resource information explicitly associated with the user, and generate a user explicit vector based on the first user information and the first resource information;
the second generating module 102 is configured to obtain second resource information, obtain second user information of a user explicitly associated with the second resource information, and generate a resource explicit vector having the same dimension as the user explicit vector based on the second resource information and the second user information;
the decomposition module 103 is configured to obtain a implicit behavior feature of a user, obtain third user information and third resource information associated with the implicit behavior feature, construct a triple relationship matrix based on the third user information and the third resource information, and perform decomposition calculation on the triple relationship matrix by using a predetermined algorithm to obtain a user implicit vector and a resource implicit vector of the user;
the calculating module 104 is configured to calculate a first similarity between the user explicit vector of the user and the corresponding resource explicit vector, and calculate a second similarity between the user implicit vector and the corresponding resource implicit vector;
the recommending module 105 is configured to perform weighted summation on the first similarity and the second similarity, select resource information based on a result of the weighted summation, and recommend the resource information to the user.
For the detailed principle, refer to the following description of fig. 3 regarding the flow chart of the method.
As shown in fig. 3, fig. 3 is a flowchart illustrating an embodiment of a method for resource recommendation according to the present invention, when a processor 13 of a device 1 for resource recommendation executes a processing system 10 stored in a memory 12, the method is implemented by the following steps:
step S1, acquiring first user information of a user, acquiring first resource information explicitly associated with the user, and generating a user explicit vector based on the first user information and the first resource information;
the first user information comprises basic information and behavior information of the user, the basic information comprises gender, age, consumption capability, working information and the like, the behavior information is behavior operation information of the user when browsing or operating resources, and the behavior information can be acquired from a log and comprises explicit behavior characteristics and implicit behavior characteristics. The explicit behavior characteristics can directly reflect the preference of a user to resources, such as collection, praise, sharing and the like, the implicit behavior characteristics cannot directly reflect the preference of the user to resources, and the implicit behavior characteristics can be resource page browsing time, search keywords, comments, clicks, mouse sliding and the like.
The first resource information is resource information on the network, which includes resource information with the same or different business attributes, and is distinguished according to the business attributes, for example, the resource information may be product information, sales information, training information, artificial intelligence information, and the like.
From the perspective of the user, if the behavior information of the user when browsing or operating the resource information is an explicit behavior characteristic, the user is explicitly associated with the resource information.
The step of generating the user explicit vector based on the first user information and the first resource information specifically includes:
predefining a multidimensional vector (a)1,a2,…,aj,b1,b2,…,bk) Wherein the multidimensional vector supports extensible, configurable operations in the form of a configuration file. a is1,a2,…,ajThe basic information (including sex, age, consumption ability, work information, etc.) of the user is 0 orAnd 1, assigning values for basic information in the multidimensional vector based on the basic information, directly obtaining corresponding values for discrete variables, and discretizing values of continuous variables by adopting a minimum entropy binning method to obtain corresponding values. For example, for gender, a value of 0 for gender male and a value of 1 for gender female; for age, 20 years old or older, including 20 years old, corresponds to a value of 0, and 20 years old or younger corresponds to a value of 1; for the job information, the value corresponding to the writer is 0, and the value corresponding to the writer is not 1.
b1,b2,…,bkThe service attribute of each resource information of the first resource information which is explicitly associated with the first user information. For the service attribute of each resource information of the first resource information, the service attribute may be determined by: the method comprises the steps of constructing a service attribute label structure which accords with a service development target in advance, then extracting text information of each resource information of first resource information, and performing subsequent processing on the text information by adopting the prior art, namely word segmentation, data cleaning, LDA theme extraction, vectorization and vector-based service attribute similarity calculation.
The value of each service attribute may be obtained in any of the following predetermined manners:
the first way is to obtain the dominant behavior characteristics and the time information generated when the user operates each resource information in the first resource information, calculate the preference degree of the user for the corresponding resource information based on the dominant behavior characteristics and the time information, and use the preference degree as the value of the corresponding service attribute, that is, calculate the preference degree of the user for the resource information of each service attribute through the dominant behavior characteristics and the time factor of the user, as the value of each service attribute: for the resource information of a certain service attribute, the corresponding behavior executed by the user is closely related to the time. Acquiring the explicit behavior characteristics (such as praise, collection and the like) of the user from the behavior information, and calculating the preference degree b of the user by adopting the following formula:
wherein t is the number of days from the current time of the user performing the dominant behavior feature on the resource information, α, β, c, tγAre constant parameters, α > 0, β > 0, c > 0, α, β, c, tγThe default values are 1, 0.42, 0.025 and 0.0025, but it is needless to say that corresponding α, β, c and t can be generated according to the data change of the service attributeγThe value of (c). Because the user may browse or operate the same resource information at different times, the preference degrees can be summarized according to the user and the service attributes, for example, the user generates an interactive behavior with the same resource information in a certain time period, the maximum preference degree of the time period can be taken as the value of b, and finally, the preference degrees corresponding to the service attributes are respectively taken as the values of the service attributes b1,b2,…,bkThe value of (c).
The preference degree of the user is calculated by combining the time factors, and the accuracy of resource recommendation is improved by the association of the time factors.
The second way is to group users based on the basic information, predict the preference degree of the users to the resource information of each service attribute through the association rule in the group, as the value of each service attribute: the users are grouped, and the definition of the group can be judged according to the data size of the users, for example, if the data size of the users is small, all the users are the whole group, and if the data size of the users is large, the groups can be grouped based on the basic information of the users, such as according to regions or industries, and each user has a corresponding group. In the spark platform, resource information preferred by a user is predicted by utilizing an association rule (FP-Growth) algorithm in each group. Specifically, each packet is constructed as G ═ G1,g2,g3,…,gnWhere n is the number of groups. Acquiring the dominant behavior characteristics of a user, user uiDominant behavioral characteristic v ofiIs denoted as { ui,viU, each user uiWith a corresponding packet g1Generating a corresponding relation R ═ R1,r2,…,rm},ri＝{gi,viAnd m is the number of users, so that resource frequent items corresponding to each group are constructed. Generating each group preference resource list and recommendation score based on the resource frequent items, further obtaining (recommendation scores of service attributes corresponding to the user and the preference resources) according to the relation R, each group preference resource list and the recommendation scores, and taking the recommendation scores as corresponding service attributes b1,b2,…,bkThe value of (c).
The intra-group association rule provided in the embodiment is beneficial to solving the problem of excessive consumption of computing resources of an association rule algorithm, enhancing the group effect of users and improving the accuracy of resource recommendation.
The third way is to merge the first way and the second way, that is, the value b of the service attribute in the first way1,b2,…,bkWith the value b of the service attribute in the second mode1,b2,…,bkAfter the correspondence is performed, weighted summation is performed, and each weight value can be predetermined. Wherein, the weights corresponding to the values of the service attributes in the first mode are all the same, for example, all 0.55, the weights corresponding to the values of the service attributes in the second mode are all the same, for example, all 0.45, and the final service attribute b is obtained after weighted summation1,b2,…,bkThe value of (c).
The above-mentioned multidimensional vector (a)1,a2,…,aj,b1,b2,…,bk) After the values of (2) are determined, a user explicit vector is generated.
Step S2, acquiring second resource information, acquiring second user information of a user explicitly associated with the second resource information, and generating a resource explicit vector with the same dimension as the user explicit vector based on the second resource information and the second user information;
the second resource information is also resource information on the network, and includes resource information with the same or different service attributes. The second user information also includes basic information and behavior information of the user.
From the resource perspective, if the behavior information of the user when browsing or operating the resource information is an explicit behavior characteristic, the resource information is explicitly associated with the user to obtain the user information of the user, and the user information of all explicitly associated users forms second user information.
Generating a resource explicit vector with the same dimension as the user explicit vector based on the second resource information and the second user information, specifically including:
predefining a multidimensional vector (A)1,A2,…,Aj,B1,B2,…,Bk) And the multidimensional vector supports extensible and configurable operation in the form of a configuration file, and the dimension of the multidimensional vector is the same as that of the user explicit vector. B is1,B2,…,BkDetermining the service attribute of each resource information in the second resource information by the following method: the method comprises the steps of constructing a service attribute label structure which accords with a service development target in advance, then extracting text information of each resource information in second resource information, and performing subsequent processing on the text information by adopting the prior art, namely word segmentation, data cleaning, LDA theme extraction, vectorization and vector-based service attribute similarity calculation.
A1,A2,…,AjIs the respective basis information of the user explicitly associated with the second resource information. The user explicitly associated with the second resource information may have multiple attribute tags (e.g., male, high consumer group, research and development engineer, etc.), which makes the base information of the user more distributed. The resource information of a service attribute may be browsed and operated by different users, but the resource information of the service attribute may not be actually suitable for all the users, so this embodiment selects the users corresponding to the attribute tags in a clustering manner, and bases on these usersThe basic information of the user of the attribute tag is obtained A1,A2,…,AjThe value of (c).
The clustering of this embodiment can adopt a kmeans algorithm: grouping users (based on the basic information of users, such as grouping according to regions or industries), analyzing user groups to obtain preset central point number, setting the central point number as k, clustering the basic information related to users with historical behavior information records, obtaining the clustering center of the basic information, and obtaining the relationship between the basic information of users and the clustering center [ basic information list, clustering center)]. After clustering, the relation [ resource information, basic information list ] can be obtained through the historical behavior information of the user]The two relations are fused to obtain the relation (resource information, clustering center)]Because one resource information corresponds to a plurality of clustering centers, the values of N clustering centers are measured according to the attribute category users, the values of the first 3 clustering centers can be taken as default, finally, weighted summation is carried out according to the user quantity ratio as the weight, and the final A is obtained through the weighted summation result1,A2,…,AjThe value of (c).
The embodiment vectorizes the user explicit characteristics and the resource explicit characteristics to obtain the user explicit vectors and the resource explicit vectors, thereby avoiding the processes of characteristic combination and related massive characteristic preprocessing and reducing the complexity of calculation.
Step S3, obtaining the implicit behavior characteristics of the user, obtaining third user information and third resource information associated with the implicit behavior characteristics, constructing a triple relation matrix based on the third user information and the third resource information, and performing decomposition calculation on the triple relation matrix by using a predetermined algorithm to obtain a user implicit vector and a resource implicit vector of the user;
in this embodiment, the implicit behavior feature of the user may be obtained from the log, and third user information and third resource information associated with the implicit behavior feature may be obtained, where the third user information also includes basic information of the user, and the third resource information is also resource information on the network, and includes resource information with the same or different service attributes. A triple relationship matrix R [ user, product, rating ] is constructed based on the third user information and the third resource information, where the triple relationship matrix includes m users and n products, the users represent the users, the products represent the resources, and the rating represents a score (i.e., a preference degree), and the rating score corresponding to the implicit behavior feature is defined as 1 in a unified manner in this embodiment.
In practical applications, the size of the triplet relation matrix R is very large, since the number of n and m is very large. At this time, the conventional matrix decomposition method is difficult to handle for such a large amount of data; furthermore, a user cannot score all resource products, so the triple relation matrix R is a sparse matrix with many missing entries.
In this embodiment, based on the spark platform, the user implicit vector and the resource implicit vector are calculated by using a predetermined algorithm (alternating least squares (ALS)), so as to obtain the user implicit vector and the resource implicit vector. Since the triplet relation matrix R is a matrix of m × n, it can be regarded as a matrix obtained by multiplying two matrices of m × k and k × n, where k < < m, n, and k typically have values of 20-200, the following formula is obtained:
Rm*n＝um*k×pk*n；
in the above formula, um*kRepresenting the user's preference for implicit behavioral characteristics, pk*nRepresenting the degree of the resource containing the implicit behavior characteristics, and calculating u according to the formulam*k、pk*nU to be calculatedm*kAs a user implicit vector, pk*nAs a resource implicit vector.
Step S4, calculating a first similarity between the user explicit vector of the user and the corresponding resource explicit vector, and calculating a second similarity between the user implicit vector and the corresponding resource implicit vector;
because the problem of large calculation amount and high calculation complexity exists in directly calculating the similarity between the user explicit vector and the corresponding resource explicit vector, the similarity is calculated by adopting a Locality-Sensitive Hashing (LSH) algorithm. Firstly, performing hash mapping processing on a user explicit vector and a corresponding resource explicit vector, designating a feature column and a unique identification column, using the features as the input of an algorithm, designating an algorithm output column, and then calculating the Euclidean distance between vectors as a similarity value by adopting an approximate similarity connection method for the mapped vectors:
calculating Euclidean distance between user explicit vectors and resource explicit vectors, and acquiring first similarity sim by using the local sensitive Hash algorithm LSH in the implementation processexplict_init；
Calculating Euclidean distance between the implicit vector of the user and the implicit vector of the resource, and acquiring a second similarity sim by adopting the description local sensitive hashing algorithm (LSH) in the implementation processimplict_init。
And step S5, carrying out weighted summation on the first similarity and the second similarity, selecting resources based on the result of the weighted summation, and recommending the resources to the user.
Respectively combining the first similarity
Second degree of similarity
Normalizing to the interval of (0, 1) to obtain simexplict、simimplictAnd weighting and summing the two groups of similarity values to obtain the total similarity:
Sim＝α*simexplict+β*simexplict，
the weights α and β are mainly set in two ways, one way is expert scoring, and fixed values of α and β are set, the other way is determination through linear regression, a sampling user is used as an experience officer in a random sampling way of the user, the sampling user is made to score similarity of provided resources, and a scoring result is used as training data to generate values of α and β.
Finally, selecting topN resource information to calculate the priority of each resource information according to the total similarity Sim obtained by weighted summation, the shelf-loading time and the heat of the resource information, and adopting priority sorting
view represents the heat (default is the heat of the previous day), age represents the number of days at which the time of the resource information is on the shelf is current, and constant parameters i and j both default to 1. And finally, pushing the topN resource information to the user according to the sorted topN resource information.
According to the description, the implicit characteristic of the user and the implicit characteristic of the resource are fused on the basis of the explicit characteristic of the user and the explicit characteristic of the resource, the existing recommendation algorithm is modified, the resource recommendation accuracy can be improved, and the interpretability of the system is enhanced.
Furthermore, the embodiment of the present invention also provides a computer-readable storage medium, which may be any one or any combination of a hard disk, a multimedia card, an SD card, a flash memory card, an SMC, a Read Only Memory (ROM), an Erasable Programmable Read Only Memory (EPROM), a portable compact disc read only memory (CD-ROM), a USB memory, and the like. The computer readable storage medium includes a processing system, and the processing system implements the functions when executed by the processor, please refer to the description related to fig. 3, which is not repeated herein.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A method for resource recommendation, the method for resource recommendation comprising:
acquiring first user information of a user, acquiring first resource information explicitly associated with the user, and generating a user explicit vector based on the first user information and the first resource information;
acquiring second resource information, acquiring second user information of a user explicitly associated with the second resource information, and generating a resource explicit vector with the same dimension as the user explicit vector based on the second resource information and the second user information;
acquiring implicit behavior characteristics of a user, acquiring third user information and third resource information associated with the implicit behavior characteristics, constructing a triple relation matrix based on the third user information and the third resource information, and performing decomposition calculation on the triple relation matrix by using a predetermined algorithm to obtain a user implicit vector and a resource implicit vector of the user;
calculating a first similarity between a user explicit vector of the user and a corresponding resource explicit vector, and calculating a second similarity between the user implicit vector and a corresponding resource implicit vector;
and carrying out weighted summation on the first similarity and the second similarity, selecting resource information based on the result of the weighted summation, and recommending the resource information to the user.
2. The method according to claim 1, wherein the first user information includes basic information and behavior information of a user, the first resource information includes resource information having the same or different service attributes, and the step of generating a user explicit vector based on the first user information and the first resource information specifically includes:
acquiring a predefined multidimensional vector, wherein the multidimensional vector comprises basic information and service attributes;
and assigning a value to the basic information in the multi-dimensional vector based on the basic information, assigning a value to the service attribute in the multi-dimensional vector based on the resource information, the basic information and the behavior information in the first resource information, and taking the assigned multi-dimensional vector as the user explicit vector.
3. The method according to claim 2, wherein the behavior information includes explicit behavior characteristics and implicit behavior characteristics, and the step of assigning the service attribute in the multidimensional vector based on each resource information, the basic information, and the behavior information in the first resource information specifically includes:
acquiring dominant behavior characteristics and time information generated when the user operates on each resource information in the first resource information, calculating the preference degree of the user on the corresponding resource information based on the dominant behavior characteristics and the time information, and taking the preference degree as the value of the corresponding service attribute; or
Grouping the users based on the basic information, predicting the preference degree of the users in each group to each resource information in the first resource information through an association rule in the group, and taking the preference degree as the value of the corresponding service attribute.
4. The method according to any one of claims 1 to 3, wherein the step of performing weighted summation on the first similarity and the second similarity, selecting resource information based on a result of the weighted summation, and recommending the resource information to the user specifically includes:
normalizing the first similarity and the second similarity respectively to obtain a preset weight, and carrying out weighted summation based on the normalized first similarity, the normalized second similarity and the weight to obtain a total similarity;
and acquiring the time and the heat of the resource information, selecting a plurality of resource information based on the total similarity and the time and the heat of the resource information, and recommending the resource information to the user.
5. The method according to claim 4, wherein the step of selecting and recommending to the user a plurality of resource information based on the total similarity, the time spent on shelves and the heat of each resource information specifically comprises:
and calculating the priority of each resource information based on the total similarity, the time of putting each resource information on shelf and the heat, selecting a plurality of resource information according to the priority of each resource information, and recommending the resource information to the user.
6. An apparatus for resource recommendation, comprising a memory and a processor connected to the memory, wherein the memory stores a processing system operable on the processor, and the processing system when executed by the processor implements the following steps:
acquiring first user information of a user, acquiring first resource information explicitly associated with the user, and generating a user explicit vector based on the first user information and the first resource information;
acquiring second resource information, acquiring second user information of a user explicitly associated with the second resource information, and generating a resource explicit vector with the same dimension as the user explicit vector based on the second resource information and the second user information;
acquiring implicit behavior characteristics of a user, acquiring third user information and third resource information associated with the implicit behavior characteristics, constructing a triple relation matrix based on the third user information and the third resource information, and performing decomposition calculation on the triple relation matrix by using a predetermined algorithm to obtain a user implicit vector and a resource implicit vector of the user;
calculating a first similarity between a user explicit vector of the user and a corresponding resource explicit vector, and calculating a second similarity between the user implicit vector and a corresponding resource implicit vector;
and carrying out weighted summation on the first similarity and the second similarity, selecting resource information based on the result of the weighted summation, and recommending the resource information to the user.
7. The apparatus for resource recommendation according to claim 6, wherein the first user information includes basic information and behavior information of a user, the first resource information includes resource information having the same or different service attributes, and the step of generating a user explicit vector based on the first user information and the first resource information specifically includes:
acquiring a predefined multidimensional vector, wherein the multidimensional vector comprises basic information and service attributes;
and assigning a value to the basic information in the multi-dimensional vector based on the basic information, assigning a value to the service attribute in the multi-dimensional vector based on the resource information, the basic information and the behavior information in the first resource information, and taking the assigned multi-dimensional vector as the user explicit vector.
8. The apparatus for resource recommendation according to claim 7, wherein the behavior information includes explicit behavior features and implicit behavior features, and the step of assigning the service attribute in the multidimensional vector based on each resource information, the basic information, and the behavior information in the first resource information specifically includes:
acquiring dominant behavior characteristics and time information generated when the user operates on each resource information in the first resource information, calculating the preference degree of the user on the corresponding resource information based on the dominant behavior characteristics and the time information, and taking the preference degree as the value of the corresponding service attribute; or
Grouping the users based on the basic information, predicting the preference degree of the users in each group to each resource information in the first resource information through an association rule in the group, and taking the preference degree as the value of the corresponding service attribute.
9. The apparatus for resource recommendation according to any one of claims 6 to 8, wherein the step of performing weighted summation on the first similarity and the second similarity, selecting resource information based on a result of the weighted summation, and recommending the resource information to the user specifically comprises:
normalizing the first similarity and the second similarity respectively to obtain a preset weight, and carrying out weighted summation based on the normalized first similarity, the normalized second similarity and the weight to obtain a total similarity;
and acquiring the time and the heat of the resource information, selecting a plurality of resource information based on the total similarity and the time and the heat of the resource information, and recommending the resource information to the user.
10. A computer-readable storage medium, having stored thereon a processing system which, when executed by a processor, performs the steps of the method of resource recommendation of any of claims 1-5.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910970985.3A CN110866181B (en) | 2019-10-12 | 2019-10-12 | Resource recommendation method, device and storage medium |
| PCT/CN2020/105925 WO2021068610A1 (en) | 2019-10-12 | 2020-07-30 | Resource recommendation method and apparatus, electronic device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910970985.3A CN110866181B (en) | 2019-10-12 | 2019-10-12 | Resource recommendation method, device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110866181A true CN110866181A (en) | 2020-03-06 |
| CN110866181B CN110866181B (en) | 2022-04-22 |
Family
ID=69652645
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910970985.3A Active CN110866181B (en) | 2019-10-12 | 2019-10-12 | Resource recommendation method, device and storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN110866181B (en) |
| WO (1) | WO2021068610A1 (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111488526A (en) * | 2020-04-14 | 2020-08-04 | 北京声智科技有限公司 | Recommendation method and device |
| CN111581505A (en) * | 2020-04-28 | 2020-08-25 | 海南太美航空股份有限公司 | Flight recommendation method and system based on combined recommendation |
| CN112559901A (en) * | 2020-12-11 | 2021-03-26 | 百度在线网络技术（北京）有限公司 | Resource recommendation method and device, electronic equipment, storage medium and computer program product |
| WO2021068610A1 (en) * | 2019-10-12 | 2021-04-15 | 平安国际智慧城市科技股份有限公司 | Resource recommendation method and apparatus, electronic device and storage medium |
| CN113095583A (en) * | 2021-04-23 | 2021-07-09 | 何桂霞 | Data analysis method applied to business management and business management server |
| CN113538053A (en) * | 2021-07-20 | 2021-10-22 | 深圳市炆石数据有限公司 | OTT resource bit classification method, system and storage medium for brand construction |
| CN114003826A (en) * | 2021-12-31 | 2022-02-01 | 思创数码科技股份有限公司 | Resource directory recommendation method and device, readable storage medium and electronic equipment |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113220999A (en) * | 2021-05-14 | 2021-08-06 | 北京百度网讯科技有限公司 | User feature generation method and device, electronic equipment and storage medium |
| CN113343086B (en) * | 2021-06-01 | 2022-09-16 | 合肥工业大学 | Data-driven green design knowledge pushing method |
| CN113590935B (en) * | 2021-06-30 | 2022-04-19 | 深圳市东信时代信息技术有限公司 | Information recommendation method and device, computer equipment and storage medium |
| CN113538108A (en) * | 2021-07-27 | 2021-10-22 | 北京沃东天骏信息技术有限公司 | Resource information determination method and device, electronic equipment and storage medium |
| CN114039744B (en) * | 2021-09-29 | 2024-02-27 | 中孚信息股份有限公司 | Abnormal behavior prediction method and system based on user feature labels |
| CN115422438B (en) * | 2022-07-21 | 2023-07-28 | 中国铁道科学研究院集团有限公司电子计算技术研究所 | Method, system and storage medium for recommending railway material supply resources |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104462385A (en) * | 2014-12-10 | 2015-03-25 | 山东科技大学 | Personalized movie similarity calculation method based on user interest model |
| CN105589916A (en) * | 2016-01-11 | 2016-05-18 | 西华大学 | Extraction method for explicit and implicit interest knowledge |
| CN106815297A (en) * | 2016-12-09 | 2017-06-09 | 宁波大学 | A kind of academic resources recommendation service system and method |
| CN107330461A (en) * | 2017-06-27 | 2017-11-07 | 安徽师范大学 | Collaborative filtering recommending method based on emotion with trust |
| CN108628999A (en) * | 2018-05-02 | 2018-10-09 | 南京大学 | It is a kind of based on explicit and implicit information video recommendation method |
| CN109241405A (en) * | 2018-08-13 | 2019-01-18 | 华中师范大学 | A kind of associated education resource collaborative filtering recommending method of knowledge based and system |
| US20190114687A1 (en) * | 2017-10-17 | 2019-04-18 | Adobe Systems Incorporated | Global Vector Recommendations Based on Implicit Interaction and Profile Data |
| CN109947983A (en) * | 2017-09-19 | 2019-06-28 | Tcl集团股份有限公司 | Video recommendation method, system, terminal and computer readable storage medium |
| CN110188208A (en) * | 2019-06-04 | 2019-08-30 | 河海大学 | A kind of the information resources inquiry recommended method and system of knowledge based map |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106354867A (en) * | 2016-09-12 | 2017-01-25 | 传线网络科技（上海）有限公司 | Multimedia resource recommendation method and device |
| CN106802915B (en) * | 2016-12-09 | 2020-07-28 | 宁波大学 | Academic resource recommendation method based on user behaviors |
| CN107492008B (en) * | 2017-08-09 | 2020-06-30 | 创新先进技术有限公司 | Information recommendation method and device, server and computer storage medium |
| CN110309405B (en) * | 2018-03-08 | 2022-09-30 | 腾讯科技（深圳）有限公司 | Project recommendation method and device and storage medium |
| CN110866181B (en) * | 2019-10-12 | 2022-04-22 | 平安国际智慧城市科技股份有限公司 | Resource recommendation method, device and storage medium |
-
2019
- 2019-10-12 CN CN201910970985.3A patent/CN110866181B/en active Active
-
2020
- 2020-07-30 WO PCT/CN2020/105925 patent/WO2021068610A1/en active Application Filing
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104462385A (en) * | 2014-12-10 | 2015-03-25 | 山东科技大学 | Personalized movie similarity calculation method based on user interest model |
| CN105589916A (en) * | 2016-01-11 | 2016-05-18 | 西华大学 | Extraction method for explicit and implicit interest knowledge |
| CN106815297A (en) * | 2016-12-09 | 2017-06-09 | 宁波大学 | A kind of academic resources recommendation service system and method |
| CN107330461A (en) * | 2017-06-27 | 2017-11-07 | 安徽师范大学 | Collaborative filtering recommending method based on emotion with trust |
| CN109947983A (en) * | 2017-09-19 | 2019-06-28 | Tcl集团股份有限公司 | Video recommendation method, system, terminal and computer readable storage medium |
| US20190114687A1 (en) * | 2017-10-17 | 2019-04-18 | Adobe Systems Incorporated | Global Vector Recommendations Based on Implicit Interaction and Profile Data |
| CN108628999A (en) * | 2018-05-02 | 2018-10-09 | 南京大学 | It is a kind of based on explicit and implicit information video recommendation method |
| CN109241405A (en) * | 2018-08-13 | 2019-01-18 | 华中师范大学 | A kind of associated education resource collaborative filtering recommending method of knowledge based and system |
| CN110188208A (en) * | 2019-06-04 | 2019-08-30 | 河海大学 | A kind of the information resources inquiry recommended method and system of knowledge based map |
Non-Patent Citations (2)
| Title |
|---|
| JICHUN ZHAO等: ""A recommendation algorithm based on collaborative filtering technology in distance learning"", 《2017 IEEE 2ND ADVANCED INFORMATION TECHNOLOGY,ELECTRONIC AND AUTOMATION CONTROL CONFERENCE(IAEAC)》 * |
| 陈登科: ""基于潜在向量模型与项目的协同过滤混合推荐"", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021068610A1 (en) * | 2019-10-12 | 2021-04-15 | 平安国际智慧城市科技股份有限公司 | Resource recommendation method and apparatus, electronic device and storage medium |
| CN111488526A (en) * | 2020-04-14 | 2020-08-04 | 北京声智科技有限公司 | Recommendation method and device |
| CN111488526B (en) * | 2020-04-14 | 2024-04-05 | 北京声智科技有限公司 | Recommendation method and device |
| CN111581505A (en) * | 2020-04-28 | 2020-08-25 | 海南太美航空股份有限公司 | Flight recommendation method and system based on combined recommendation |
| CN111581505B (en) * | 2020-04-28 | 2023-07-07 | 海南太美航空股份有限公司 | Flight recommendation method and system based on combined recommendation |
| CN112559901A (en) * | 2020-12-11 | 2021-03-26 | 百度在线网络技术（北京）有限公司 | Resource recommendation method and device, electronic equipment, storage medium and computer program product |
| CN112559901B (en) * | 2020-12-11 | 2022-02-08 | 百度在线网络技术（北京）有限公司 | Resource recommendation method and device, electronic equipment, storage medium and computer program product |
| CN113095583A (en) * | 2021-04-23 | 2021-07-09 | 何桂霞 | Data analysis method applied to business management and business management server |
| CN113095583B (en) * | 2021-04-23 | 2024-05-03 | 浙江皓亮智享信息技术咨询有限公司 | Data analysis method applied to service management and service management server |
| CN113538053A (en) * | 2021-07-20 | 2021-10-22 | 深圳市炆石数据有限公司 | OTT resource bit classification method, system and storage medium for brand construction |
| CN113538053B (en) * | 2021-07-20 | 2023-09-01 | 深圳市爱易讯数据有限公司 | OTT resource bit classification method, system and storage medium for brand construction |
| CN114003826A (en) * | 2021-12-31 | 2022-02-01 | 思创数码科技股份有限公司 | Resource directory recommendation method and device, readable storage medium and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2021068610A1 (en) | 2021-04-15 |
| CN110866181B (en) | 2022-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110866181B (en) | Resource recommendation method, device and storage medium | |
| WO2019169756A1 (en) | Product recommendation method and apparatus, and storage medium | |
| CN111798273A (en) | Training method of purchase probability prediction model of product and purchase probability prediction method | |
| CN106844407B (en) | Tag network generation method and system based on data set correlation | |
| CN110503459B (en) | User credibility assessment method and device based on big data and storage medium | |
| CN110827112B (en) | Deep learning commodity recommendation method and device, computer equipment and storage medium | |
| CN111178949B (en) | Service resource matching reference data determining method, device, equipment and storage medium | |
| CN107818491A (en) | Electronic installation, Products Show method and storage medium based on user's Internet data | |
| US10937070B2 (en) | Collaborative filtering to generate recommendations | |
| CN111966886A (en) | Object recommendation method, object recommendation device, electronic equipment and storage medium | |
| CN110688565A (en) | Next item recommendation method based on multidimensional Hox process and attention mechanism | |
| CN114880566A (en) | User behavior analysis method, device, equipment and medium based on graph neural network | |
| CN113327132A (en) | Multimedia recommendation method, device, equipment and storage medium | |
| CN114329004A (en) | Digital fingerprint generation method, digital fingerprint generation device, data push method, data push device and storage medium | |
| CN111475720A (en) | Recommendation method, recommendation device, server and storage medium | |
| CN114282119B (en) | Scientific and technological information resource retrieval method and system based on heterogeneous information network | |
| US11709798B2 (en) | Hash suppression | |
| CN115186188A (en) | Product recommendation method, device and equipment based on behavior analysis and storage medium | |
| US20140324523A1 (en) | Missing String Compensation In Capped Customer Linkage Model | |
| CN110837596B (en) | Intelligent recommendation method and device, computer equipment and storage medium | |
| CN113468421A (en) | Product recommendation method, device, equipment and medium based on vector matching technology | |
| CA2485814A1 (en) | Method and apparatus for range processing in an n-dimensional space | |
| CN111291829A (en) | Automatic determination method and system for selected pictures | |
| CN112989020A (en) | Information processing method, apparatus and computer readable storage medium | |
| CN113837846B (en) | Commodity recommendation method, commodity recommendation device, computer equipment and storage medium |
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 |