CN112132660B - Commodity recommendation method, system, equipment and storage medium - Google Patents
Commodity recommendation method, system, equipment and storage medium Download PDFInfo
- Publication number
- CN112132660B CN112132660B CN202011027186.1A CN202011027186A CN112132660B CN 112132660 B CN112132660 B CN 112132660B CN 202011027186 A CN202011027186 A CN 202011027186A CN 112132660 B CN112132660 B CN 112132660B
- Authority
- CN
- China
- Prior art keywords
- commodity
- recommendation
- vector
- log data
- preference
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000003860 storage Methods 0.000 title claims abstract description 13
- 239000013598 vector Substances 0.000 claims abstract description 186
- 238000012545 processing Methods 0.000 claims abstract description 62
- 230000006399 behavior Effects 0.000 claims description 127
- 238000009826 distribution Methods 0.000 claims description 34
- 238000007477 logistic regression Methods 0.000 claims description 21
- 230000008569 process Effects 0.000 claims description 11
- 230000003542 behavioural effect Effects 0.000 claims description 8
- 238000013519 translation Methods 0.000 claims description 5
- 238000004422 calculation algorithm Methods 0.000 abstract description 9
- 238000010586 diagram Methods 0.000 description 10
- 238000004590 computer program Methods 0.000 description 7
- 244000062793 Sorghum vulgare Species 0.000 description 6
- 238000004891 communication Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 235000019713 millet Nutrition 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 5
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 241000282421 Canidae Species 0.000 description 2
- 241000282461 Canis lupus Species 0.000 description 2
- 230000004075 alteration Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- DFEYYRMXOJXZRJ-UHFFFAOYSA-N sevoflurane Chemical compound FCOC(C(F)(F)F)C(F)(F)F DFEYYRMXOJXZRJ-UHFFFAOYSA-N 0.000 description 1
- 229960002078 sevoflurane Drugs 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000026676 system process Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0623—Item investigation
- G06Q30/0625—Directed, with specific intent or strategy
Abstract
The invention discloses a commodity recommendation method, a commodity recommendation system, commodity recommendation equipment and a storage medium, wherein the commodity recommendation method comprises the following steps: acquiring latest behavior log data generated when a user browses or purchases goods; processing the latest behavior log data by adopting a preset preference model group to generate commodity preference vectors of users; processing the latest behavior log data by adopting a preset entity relation model to generate commodity vectors; and calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector, and recommending the commodity according to the recommendation probability. The problem that the recommendation accuracy of a recommendation algorithm adopted by the conventional shopping network for recommending commodities to a user is inaccurate and the actual demands of the user are difficult to meet is solved, so that the recommendation system of the shopping network has stronger timeliness for recommending commodities to the user, the accuracy of commodity recommendation is improved, behavior data of browsing and purchasing the commodities by the user can be synthesized, and the relationship between the commodities is evaluated.
Description
Technical Field
The present invention relates to the field of data analysis technologies, and in particular, to a method, a system, an apparatus, and a storage medium for recommending commodities.
Background
The online shopping is also called shopping, is a representation form of electronic commerce, and refers to the process that a user browses and searches related information in a virtual shopping environment for completing shopping or related tasks, thereby providing necessary information for purchasing decision making and practicing decision making and purchasing.
After the user browses or clicks the commodity, the background of the website can automatically save the operation trace of the user, and recommend the related commodity to the user according to the historical operation trace of the user. However, the recommendation algorithms employed by current shopping networks to recommend items to users include conventional recommendation algorithms and general recommendation algorithms.
The traditional recommendation algorithm system generally adopts a collaborative filtering algorithm, but has poor recommendation precision in the specific field with higher professionality, and simultaneously only meets objective requirements of users, and has a certain deviation from understanding of actual requirements of the users. The general recommendation algorithm only adopts a CBOW model and a Skip-gram model although combining related word vector related text processing technology, and the depth requirement of a user is not deeply understood although the sequential sequence of clicking articles by the user is considered, so that the recommendation precision is inaccurate.
Therefore, the recommendation algorithm adopted by the current shopping network for recommending commodities to users is inaccurate in recommendation accuracy, and actual demands of users are difficult to meet.
Disclosure of Invention
The embodiment of the application aims to solve the problems that the recommendation accuracy of a recommendation algorithm adopted by the current shopping network for recommending commodities to a user is inaccurate and the actual demands of the user are difficult to meet by providing the commodity recommendation method, the system, the equipment and the storage medium.
The embodiment of the application provides a commodity recommendation method, which comprises the following steps:
acquiring latest behavior log data generated when a user browses or purchases goods;
processing the latest behavior log data by adopting a preset preference model group to generate commodity preference vectors of users;
processing the latest behavior log data by adopting a preset entity relation model to generate commodity vectors;
and calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector, and recommending the commodity according to the recommendation probability.
In one embodiment, the determining the latest behavior log data generated when the user browses or purchases the commodity includes:
acquiring all historical behavior log data generated when a user browses or purchases goods;
Dividing the historical time of browsing or purchasing commodities by a user into a plurality of time windows with equal length;
and extracting historical behavior log data corresponding to the latest preset number of time windows as the latest behavior log data.
In an embodiment, the processing the latest behavior log data by using a preset preference model group to generate a commodity preference vector of the user includes:
extracting data statistical characteristics of the commodity according to the latest behavior log data;
and processing the data statistical characteristics by adopting a machine model in the preference model group to obtain the commodity preference vector.
In an embodiment, the processing the data statistics using the machine model in the preference model group to obtain the commodity preference vector includes:
forming a behavior vector of the user according to the data statistics characteristics;
processing the behavior vector by adopting machine models corresponding to different types of commodities respectively to obtain first probability distribution;
processing the behavior vector by adopting machine models corresponding to different brands of commodities in the same class of commodities respectively to obtain second probability distribution;
and splicing the first probability distribution and the second probability distribution to generate the commodity preference vector.
In one embodiment, the processing the latest behavior log data using a predetermined entity relationship model to generate a commodity vector includes
Forming the latest behavior log data into triple information; wherein the triplet information includes: the first commodity, the second commodity and the behavioral relationship between the first commodity and the second commodity;
and calculating the triplet information by adopting the entity relation model to obtain the commodity vector.
In one embodiment, the entity-relationship model consists of a word vector model and a translation model.
In an embodiment, the calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector, and the recommending the commodity according to the recommendation probability, includes:
splicing the commodity preference vector and the commodity vector to generate a spliced vector;
processing the spliced vector by adopting a logistic regression model, and outputting the recommendation probability of the commodity;
and recommending the commodity if the recommendation probability is greater than a recommendation probability threshold.
In addition, in order to achieve the above object, the present invention also provides a commodity recommendation system, including:
the data acquisition module is used for acquiring the latest behavior log data generated when a user browses or purchases goods;
The first processing module is used for processing the latest behavior log data by adopting a preset preference model group and generating commodity preference vectors of users;
the second processing module is used for processing the latest behavior log data by adopting a preset entity relation model to generate commodity vectors;
and the commodity recommendation module is used for calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector and recommending the commodity according to the recommendation probability.
In addition, in order to achieve the above object, the present invention further provides a commodity recommendation method and apparatus, including: the commodity recommending system comprises a memory, a processor and a commodity recommending program which is stored in the memory and can run on the processor, wherein the commodity recommending program realizes the steps of the commodity recommending method when being executed by the processor.
In addition, in order to achieve the above object, the present invention also provides a storage medium having stored thereon a commodity recommendation program which, when executed by a processor, implements the steps of the commodity recommendation method described above.
The technical scheme of the commodity recommendation method, the system, the equipment and the storage medium provided by the embodiment of the application has at least the following technical effects or advantages:
The method comprises the steps of acquiring the latest behavior log data generated when a user browses or purchases commodities, processing the latest behavior log data by a preset preference model group, generating commodity preference vectors of the user, processing the latest behavior log data by a preset entity relation model, generating commodity vectors, calculating the recommendation probability of the commodities according to the commodity preference vectors and the commodity vectors, and carrying out commodity recommendation according to the recommendation probability.
Drawings
FIG. 1 is a schematic diagram of a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a first embodiment of the merchandise recommendation method of the present invention;
FIG. 3 is a flowchart of a second embodiment of the merchandise recommendation method of the present invention;
FIG. 4 is a flowchart of a third embodiment of a merchandise recommendation method according to the present invention;
FIG. 5 is a flowchart of a fourth embodiment of a merchandise recommendation method according to the present invention;
FIG. 6 is a flowchart of a fifth embodiment of a merchandise recommendation method according to the present invention;
FIG. 7 is a flowchart of a sixth embodiment of a merchandise recommendation method according to the present invention;
FIG. 8 is a functional block diagram of the merchandise recommendation system of the present invention.
Detailed Description
In order that the above-described aspects may be better understood, exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may 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 disclosure to those skilled in the art.
The invention provides commodity recommending equipment. As shown in fig. 1, fig. 1 is a schematic structural diagram of a hardware running environment according to an embodiment of the present invention.
It should be noted that fig. 1 is a schematic structural diagram of a hardware operation environment of the commodity recommendation device.
As shown in fig. 1, the commodity recommendation apparatus may include: a processor 1001, such as a CPU, memory 1005, user interface 1003, network interface 1004, communication bus 1002. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
Optionally, the merchandise recommendation device may also include RF (Radio Frequency) circuitry, sensors, audio circuitry, wiFi modules, and the like.
It will be appreciated by those skilled in the art that the merchandise recommendation apparatus structure shown in fig. 1 is not meant to be limiting and may include more or fewer components than shown, or may combine certain components, or a different arrangement of components.
As shown in fig. 1, an operating system, a network communication module, a user interface module, and a commodity recommendation program may be included in the memory 1005 as one type of storage medium. The operating system is a program for managing and controlling hardware and software resources of the commodity recommendation device, and a commodity recommendation program and other software or running programs.
In the commodity recommendation apparatus shown in fig. 1, the user interface 1003 is mainly used for connecting a terminal, and performs data communication with the terminal; the network interface 1004 is mainly used for a background server and is in data communication with the background server; the processor 1001 may be configured to invoke the merchandise recommendation program stored in the memory 1005.
In the present embodiment, the commodity recommendation apparatus includes: a memory 1005, a processor 1001, and a merchandise recommendation program stored on the memory and executable on the processor, wherein:
When the processor 1001 calls the commodity recommendation program stored in the memory 1005, the following operations are performed:
acquiring latest behavior log data generated when a user browses or purchases goods;
processing the latest behavior log data by adopting a preset preference model group to generate commodity preference vectors of users;
processing the latest behavior log data by adopting a preset entity relation model to generate commodity vectors;
and calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector, and recommending the commodity according to the recommendation probability.
When the processor 1001 calls the commodity recommendation program stored in the memory 1005, the following operations are performed:
acquiring all historical behavior log data generated when a user browses or purchases goods;
dividing the historical time of browsing or purchasing commodities by a user into a plurality of time windows with equal length;
and extracting historical behavior log data corresponding to the latest preset number of time windows as the latest behavior log data.
When the processor 1001 calls the commodity recommendation program stored in the memory 1005, the following operations are performed:
extracting data statistical characteristics of the commodity according to the latest behavior log data;
And processing the data statistical characteristics by adopting a machine model in the preference model group to obtain the commodity preference vector.
When the processor 1001 calls the commodity recommendation program stored in the memory 1005, the following operations are performed:
forming a behavior vector of the user according to the data statistics characteristics;
processing the behavior vector by adopting machine models corresponding to different types of commodities respectively to obtain first probability distribution;
processing the behavior vector by adopting machine models corresponding to different brands of commodities in the same class of commodities respectively to obtain second probability distribution;
and splicing the first probability distribution and the second probability distribution to generate the commodity preference vector.
When the processor 1001 calls the commodity recommendation program stored in the memory 1005, the following operations are performed:
forming the latest behavior log data into triple information; wherein the triplet information includes: the first commodity, the second commodity and the behavioral relationship between the first commodity and the second commodity;
and calculating the triplet information by adopting the entity relation model to obtain the commodity vector.
The entity relation model consists of a word vector model and a translation model.
When the processor 1001 calls the commodity recommendation program stored in the memory 1005, the following operations are performed:
splicing the commodity preference vector and the commodity vector to generate a spliced vector;
processing the spliced vector by adopting a logistic regression model, and outputting the recommendation probability of the commodity;
and recommending the commodity if the recommendation probability is greater than a recommendation probability threshold.
The embodiments of the present invention provide embodiments of commodity recommendation methods, it being noted that although a logical order is illustrated in the flowchart, in some cases the steps illustrated or described may be performed in a different order than that illustrated herein.
As shown in fig. 2, in a first embodiment of the present application, the commodity recommendation method of the present application includes the following steps:
step S210: and acquiring the latest behavior log data generated when the user browses or purchases the commodity.
In this embodiment, the recommendation system obtains the latest behavior log data generated when the user browses or purchases the merchandise. The object of the application of the recommendation system is a transaction platform, such as Taobao, beijing east, unique meeting, etc. The behavior log data refers to behavior data recorded by the recommendation system when a user purchases or browses goods on the internet, and comprises goods clicking data, goods collecting data, goods order data, goods purchasing data and the like. For example, a user purchases a millet mobile phone displayed in store 1 through a treasured washing net, clicks a Hua mobile phone displayed in store 2 before purchase, and stores a VIVO mobile phone displayed in store 3, which belong to the behavior log data, at 14 points, namely, 30 months in 2020. The latest behavior log data refers to behavior data recorded by the recommendation system when a user purchases or browses goods on the internet at a time closest to the current time or for a period of time. For example, when 10 minutes and 25 seconds are at the current month 7 and 8 in 2020, and 10 minutes and 10 seconds are at the current month 7 and 8 in 2020, a user clicks on the sports bracelet displayed in store 4. The "7/8/2020/10 minutes/10 seconds" is considered as the latest behavior log data when a user clicks the sports bracelet displayed in store 4.
Step S220: and processing the latest behavior log data by adopting a preset preference model group to generate commodity preference vectors of users.
In this embodiment, the preference model group is pre-constructed and is used for processing the obtained behavior log data to obtain the commodity preference vector of the user. Specifically, after acquiring the latest behavior log data generated when a user browses or purchases a commodity, the recommendation system takes the latest behavior log data as input data and inputs the input data into a preference model group, and the preference model group acquires commodity preference of the user and the preference degree of the user on the commodity according to the latest behavior log data, so that a commodity preference vector is obtained.
Step S230: and processing the latest behavior log data by adopting a preset entity relation model to generate commodity vectors.
In this embodiment, the entity relationship model is pre-constructed and is used to process the obtained behavior log data to obtain the commodity vector in the commodity library of the transaction platform. The commodity library of the transaction platform stores various commodities, each commodity is provided with a commodity vector, and the commodity vector can be understood as a commodity label for facilitating understanding. Every new commodity is stored in commodity library, and the recommendation system automatically generates a commodity vector, and the commodity can obtain its own commodity vector. Specifically, after the recommendation system obtains the latest behavior log data generated when the user browses or purchases the commodity, the latest behavior log data is used as input data and is input into the entity relation model, the entity relation model obtains commodity information browsed or purchased by the user from the latest behavior log data, and the obtained commodity information is compared with commodity information in a commodity library to obtain two same commodity information, so that a commodity vector can be obtained.
Step S240: and calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector, and recommending the commodity according to the recommendation probability.
In this embodiment, the recommendation system may determine, according to the commodity preference vector and the commodity vector, a commodity vector of each commodity and a commodity preference vector of the user when the user browses or purchases the commodity. The recommendation probability of each commodity can be predicted through the commodity preference vector, the recommended commodity can be predicted through the commodity vector, and the corresponding commodity can be recommended according to the recommendation probability of each commodity.
According to the technical scheme, the latest behavior log data generated when the user browses or purchases the commodity is acquired, the latest behavior log data is processed by the preset preference model group to generate the commodity preference vector of the user, the latest behavior log data is processed by the preset entity relation model to generate the commodity vector, the commodity recommendation probability is calculated according to the commodity preference vector and the commodity vector, and the technical means of commodity recommendation is carried out according to the recommendation probability, so that the timeliness of recommending the commodity to the user by the recommendation system of the shopping network is higher, the accuracy of commodity recommendation is improved, the behavior data of the user browsing and purchasing the commodity can be synthesized, and the relationship between the commodity and the commodity is evaluated.
As shown in fig. 3, in the second embodiment of the present application, the method for recommending goods of the present application, step S210 specifically includes the following steps:
step S211: all historical behavior log data generated when a user browses or purchases a commodity is obtained.
In this embodiment, before the recommendation system obtains the latest behavior log data, it is necessary to obtain all the historical behavior log data generated when the user browses or purchases the merchandise. When a user browses or purchases goods on the transaction platform, the recommendation system adopts a buried point mode to collect behavior log data generated when the user browses or purchases goods in real time, and the collected behavior log data is stored to be historical behavior log data.
Step S212: the historical time for a user to browse or purchase goods is divided into a plurality of time windows of equal length.
In this embodiment, when a user browses or purchases a commodity on the transaction platform, each time corresponds to a time in which behavior log data of the user is stored, and a time corresponding to the behavior log data stored by the recommendation system is a history time. Specifically, before the recommendation system obtains the latest behavior log data, the historical time for browsing or purchasing the commodity by the user needs to be divided into a plurality of time windows, and each time window is equal in length so as to select the corresponding historical behavior log data by selecting each time window.
Step S213: and extracting historical behavior log data corresponding to the latest preset number of time windows as the latest behavior log data.
In this embodiment, after dividing the historical time of browsing or purchasing goods by the user into a plurality of time windows with equal length, the recommendation system sorts all the time windows according to the time sequence of browsing or purchasing goods by the user, and then extracts the historical behavior log data corresponding to a preset number of time windows close to the current time as the latest behavior log data. Assuming that the current time is 6 months and 10 days in 2020, a user browses or purchases goods on a transaction platform in 5-10 days in 2020, and the recommendation system divides the five-day time into 20 time windows according to hours, wherein the length of each time window is 6 hours, namely the time windows in each day of 5-10 days in 2020 are respectively as follows: 1 time-6 time, 7 time-12 time, 13 time-18 time, 17 time-24 time. Wherein, 6 month 9 day is nearest to 6 month 10 day, the historical behavior log data corresponding to four time windows in 6 month 9 day can be used as the latest behavior log data.
According to the technical scheme, due to the adoption of the technical means that all the historical behavior log data generated when the user browses or purchases the commodity are acquired, the historical time when the user browses or purchases the commodity is divided into a plurality of equal-length time windows, the historical behavior log data corresponding to the latest preset number of time windows are extracted to serve as the latest behavior log data, the commodity favored by the latest time of the user can be acquired, and the acquired latest behavior log data is more accurate.
As shown in fig. 4, in a third embodiment of the present application, step S220 specifically includes the following steps:
step S221: and extracting the data statistical characteristics of the commodity according to the latest behavior log data.
In this embodiment, after the recommendation system obtains the latest behavior log data, the latest behavior log data is processed to obtain the data statistics feature of the commodity. The data statistics feature refers to the number of clicks that a user clicks at different times when browsing or purchasing a commodity.
Step S222: and processing the data statistical characteristics by adopting a machine model in the preference model group to obtain the commodity preference vector.
In this embodiment, the preference model group includes a plurality of machine learning models, such as a logistic regression model (LR model), a support vector machine network (SVM model), and the like, where each category of commodity corresponds to one machine learning model, and commodities of different brands in the same category of commodity correspond to one machine learning model. The data statistics features corresponding to different types of commodities are input into corresponding machine learning models, commodity preference vectors corresponding to the types of commodities can be output, the machine learning models corresponding to different brands of commodities in the same type of commodities can be output, and commodity preference vectors corresponding to the different brands of commodities in the type of commodities can be output.
According to the technical scheme, the data statistics characteristics of the commodities are extracted according to the latest behavior log data, and the machine model in the preference model group is adopted to process the data statistics characteristics, so that the technical means of commodity preference vectors is obtained, corresponding commodity preference vectors of commodities of different categories can be obtained in a reasonable and standard mode, and confusion of the data is avoided.
As shown in fig. 5, in a fourth embodiment of the present application, step S222 specifically includes the following steps:
step S2221: and forming a behavior vector of the user according to the data statistics characteristics.
In this embodiment, different data statistics features exist for each time window, that is, the number of clicks that the product is clicked by the user. And the recommendation system processes the click times of the commodities in the same category corresponding to the preset number of time windows to obtain the behavior vector of the user. For example, the user clicks the class a commodity and the class B commodity in four time windows of 6 months and 9 days, the clicking times of the class a commodity clicked by the user are respectively 4,2,3 and 3, and the clicking times of the class B commodity clicked by the user are respectively 1,2,1 and 3 according to the time sequence. And after the recommendation system performs splicing processing on the clicking times of the class A commodity clicked by the user and the clicking times of the class B commodity clicked by the user, obtaining the behavior vector of the user on the day of 6 months and 9 days, namely [4,2,3,3,1,2,1,3].
Step S2222: and processing the behavior vector by adopting machine models corresponding to different types of commodities respectively to obtain first probability distribution.
In this embodiment, the recommendation system inputs the obtained behavior vector into the machine model corresponding to the different types of commodities, and outputs the preference probability of the user for each type of commodity through processing, so that the first probability distribution can be obtained according to the preference probability of each type of commodity. For example, assuming that the class a commodity is clothes and the class B commodity is electronic equipment, the behavior vector [4,2,3,3,1,2,1,3] of the user on day 6 and 9 is respectively input into a logistic regression model of the clothes and a logistic regression model of the electronic equipment, the clothes preference probability output by the logistic regression model of the clothes is 20%, the electronic equipment preference probability output by the logistic regression model of the electronic equipment is 70%, the user can be determined to be interested in the electronic equipment, the obtained preference vector of the commodity class is [0.2,0.7], and the preference vector of the commodity class is used as the first probability distribution.
Step S2223: and processing the behavior vector by adopting machine models corresponding to different brands of commodities in the same class of commodities respectively to obtain second probability distribution.
In this embodiment, the recommendation system inputs the obtained behavior vector into the machine model corresponding to the different brands of the same category of goods, outputs the preference probability of the user for the different brands of the same category of goods after processing, and can obtain the second probability distribution according to the preference probability of the different brands of the same category of goods. Continuing with the example in step S2222, the brands of the class a commodity include the pool of yours and seven wolves, and the brands of the class B commodity include the apple phone and the millet phone. The recommendation system inputs the behavior vector [4,2,3,3,1,2,1,3] into a logistic regression model of a top-quality library, a logistic regression model of seven wolves, a logistic regression model of a apple mobile phone and a logistic regression model of a millet mobile phone respectively, and then the output result of each logistic regression model is as follows: the preference probability of the user to the optimal clothes library is 50%, the preference probability of the user to the seven wolf brands is 40%, the preference probability of the user to the apple mobile phone brands is 80%, the preference probability of the user to the millet mobile phone brands is 20%, and in the class A commodity, the user interested in the optimal clothes library can be determined, so that the preference vector of the clothes brands is [0.5,0.4]; in the class B commodity, the user is interested in the brand of the apple mobile phone, the preference vector of the brand of the electronic equipment is obtained as [0.8,0.2], and the preference vector of the brand of the clothes and the preference vector of the brand of the electronic equipment are respectively used as second distribution probability.
Step S2224: and splicing the first probability distribution and the second probability distribution to generate the commodity preference vector.
In this embodiment, the recommendation system splices the obtained first probability distribution and the obtained second probability distribution, so as to obtain the commodity preference vector. Continuing with the example in step S2223, the recommender system concatenates the preference vector for the category of the good, the preference vector for the brand of the clothing, and the preference vector for the brand of the electronic device to obtain the preference vector for the good, [0.2,0.7,0.5,0.4,0.8,0.2].
According to the technical scheme, the behavior vectors of the user are formed according to the data statistics characteristics, the machine models corresponding to different types of commodities are respectively adopted to process the behavior vectors, so that first probability distribution is obtained, the machine models corresponding to different brands of commodities in the same type of commodities are respectively adopted to process the behavior vectors, second probability distribution is obtained, the first probability distribution and the second probability distribution are spliced, and the technical means of commodity preference vectors are generated, so that the preference degree of the user for the different types of commodities can be predicted, and the preference degree of the user for the different brands of commodities in the same type of commodities can be predicted.
As shown in fig. 6, in a fifth embodiment of the present application, step S230 specifically includes the following steps:
step S231: and constructing the latest behavior log data into triple information.
In this embodiment, the triplet information includes: the first article, the second article, and a behavioral relationship between the first article and the second article. The triplet information is expressed as (h, r, t), h is the first commodity, r is the behavioral relationship between the first commodity and the second commodity, and t is the second commodity.
Specifically, the behavioral relationship between the first commodity and the second commodity includes: post-browse, post-browse purchase, post-purchase browse, post-purchase. After the recommendation system collates the acquired latest behavior log data, the formation of the triplet information comprises the following steps:
(first commodity, purchased after browsing, second commodity), i.e. purchased after browsing the first commodity;
(first commodity, purchased after purchase, second commodity), i.e. purchased after purchase of the first commodity;
(first commodity, browse after browsing, second commodity), i.e. browse the second commodity after browsing the first commodity;
(first commodity, browse after purchase, second commodity), i.e. browse second commodity after purchase of first commodity;
The first commodity and the second commodity can be the same commodity or different commodities.
Step S232: and calculating the triplet information by adopting the entity relation model to obtain the commodity vector.
In this embodiment, the entity relationship model is used to calculate the triplet information, so as to obtain the commodity vector. The entity relation model consists of a word vector model (word 2vec model) and a translation model (trans-E model), namely the entity relation model can be obtained through the word2vec model and the trans-E model. The objective function can be obtained by a word2vec model and a transient model, and is specifically as follows:
（1）
wherein the left side of equation (1) is the Word2vec Model based on the CBOW Model (Continuous Bag-of-Word Model), the right side is the relational Word vector Model, γ is the parameter balancing the contribution ratio of the two models, |C| is the size of the whole corpus,is to browse w by user i Commodity w is pushed out from the front and back commodity i Is a function of the probability of (1),is to contact commodity w by user i And triggering (e.g. browsing after browsing) the behavior relation r between the first commodity and the second commodity, and recommending the probability of the second commodity t.
The commodity w of the relational word vector model is obtained by the formula (1) i And the behavior relation r gradient between the first commodity and the second commodity is updated as follows:
（2）
（3）
assuming that the expression (2) and the expression (3) obtain the commodity vector of the first commodity, the commodity vector of the second commodity, and the behavior relation vector between the first commodity and the second commodity as follows:
the commodity vector representation result of the first commodity is [0.1,0.2,0.1,0.1,0.15, … ], and the total dimension is 300;
the commodity vector representation result of the second commodity is [0.2,0.1,0.1,0.01,0.5, … ], which is 300 dimensions in total;
the behavior relation between the first commodity and the second commodity is that the relation vector is [0.1, -0.1,0, -0.09, -0.1, … ] which is 300 dimensionality in total.
Here, the relationship between the commodity vector of the first commodity and the browsing relationship vector=the commodity vector of the second commodity, and the vector relationship of the relationship can predict that the user will browse the second commodity after browsing the first commodity. For example, after the user 1 browses the mobile phone, he can browse the mobile phone shell back; after the user 2 browses the garment, he browses the pants, etc.
Therefore, the commodity vector representation result of the first commodity and the commodity vector representation result of the second commodity are commodity vectors obtained by calculating the triplet information through the entity relation model.
According to the technical scheme, the technical means that the latest behavior log data form the triplet information and the entity relation model is adopted to calculate the triplet information to obtain the commodity vector can further improve the expression capability of the commodity vector.
As shown in fig. 7, in a sixth embodiment of the present application, step S240 specifically includes the following steps:
step S241: and splicing the commodity preference vector and the commodity vector to generate a spliced vector.
In this embodiment, the recommendation system splices the commodity preference vector of the obtained user and the commodity vector to obtain a spliced vector. For example, the commodity preference vector is [0.2,0.7,0.5,0.4,0.8,0.2], the commodity vector is [0.1,0.2,0.1,0.1,0.15, … ], and the splice vector obtained by splicing the commodity preference vector and the commodity vector is [0.2,0.7,0.5,0.4,0.8,0.2,0.1,0.2,0.1,0.1,0.15, … ].
Step S242: and processing the spliced vector by adopting a logistic regression model, and outputting the recommendation probability of the commodity.
In this embodiment, the recommendation system inputs the obtained spliced vector into a logistic regression model, and outputs the recommendation probability of the commodity through the logistic regression model processing. Specifically, it is assumed that the commodity preference vector in step S241 includes preference probabilities of the user for the clothes class commodity and the electronic device class commodity, preference probabilities for the excellent clothes library commodity and the sevoflurane commodity in the clothes class commodity, and preference probabilities for the apple mobile phone and the millet mobile phone commodity in the electronic device class commodity, respectively. The obtained commodity corresponding to the commodity vector belongs to one of the clothes type commodity and the electronic equipment type commodity, the logistic regression model obtains data of a commodity vector part from the spliced vector, and determines a specific commodity corresponding to the commodity vector from a commodity library according to the data of the commodity vector part, for example, determines that the commodity library is an apple mobile phone, and can determine the preference probability of a user on the apple mobile phone in the electronic equipment type commodity through the apple mobile phone. The preference probability of a user for electronic equipment type commodities is 0.7, the preference probability of the user for clothes type commodities is 0.2, the preference probability of the user for excellent clothes warehouse commodities is 0.5, the preference probability of the user for seven wolf commodities is 0.8, the preference probability of the user for apple mobile phones is 0.8, the preference probability of the user for millet mobile phones is 0.2, the user can see that the user prefers apple mobile phones in the electronic equipment type commodities, the recommendation probability of the output commodities is 0.8, and the commodities are apple mobile phones.
Step S243: and recommending the commodity if the recommendation probability is greater than a recommendation probability threshold.
In this embodiment, the recommendation system compares the recommendation probability of the commodity output by the logistic regression model with a preset recommendation probability threshold, if the recommendation probability of the commodity output is greater than the recommendation probability threshold, the corresponding commodity is recommended to the user, otherwise, the recommendation is not performed.
According to the technical scheme, the commodity preference vector and the commodity vector are spliced to generate the spliced vector, the logistic regression model is adopted to process the spliced vector, the recommendation probability of the commodity is output, if the recommendation probability is larger than the recommendation probability threshold, the technical means of recommending the commodity is adopted, the commodity recommendation accuracy is improved, the actual shopping requirement of a user is favorably met, and therefore sales of the commodity is promoted.
As shown in fig. 8, the present application further provides a commodity recommendation system, including:
the data acquisition module 310 is configured to acquire latest behavior log data generated when a user browses or purchases a commodity;
the first processing module 320 is configured to process the latest behavior log data by using a preset preference model group, and generate a commodity preference vector of the user;
The second processing module 330 is configured to process the latest behavior log data by using a preset entity relationship model, and generate a commodity vector;
the commodity recommendation module 340 is configured to calculate a recommendation probability of a commodity according to the commodity preference vector and the commodity vector, and perform commodity recommendation according to the recommendation probability.
Further, the data acquisition module 310 includes:
the first acquisition unit is used for acquiring all historical behavior log data generated when a user browses or purchases goods;
the time dividing unit is used for dividing the historical time of browsing or purchasing commodities by a user into a plurality of time windows with equal length;
and the second acquisition unit is used for extracting historical behavior log data corresponding to the latest preset number of time windows as the latest behavior log data.
Further, the first processing module 320 includes:
the feature extraction unit is used for extracting the data statistics features of the commodity according to the latest behavior log data;
and the first model processing unit is used for processing the data statistical characteristics by adopting a machine model in the preference model group to obtain the commodity preference vector.
Further, the model processing unit includes:
A vector forming subunit, configured to form a behavior vector of a user according to the data statistics feature;
the first model subunit is used for processing the behavior vectors by adopting machine models corresponding to different types of commodities respectively to obtain first probability distribution;
the second model subunit is used for processing the behavior vectors by adopting machine models corresponding to different brands of commodities in the same class of commodities respectively to obtain second probability distribution;
and the data integration subunit is used for splicing the first probability distribution and the second probability distribution to generate the commodity preference vector.
Further, the second processing module 330 includes:
the data composing unit is used for composing the latest behavior log data into triple information; wherein the triplet information includes: the first commodity, the second commodity and the behavioral relationship between the first commodity and the second commodity;
and the data calculation unit is used for calculating the triplet information by adopting the entity relation model to obtain the commodity vector.
Further, the entity relation model consists of a word vector model and a translation model.
Further, the commodity recommendation module 340 includes:
The data splicing unit is used for splicing the commodity preference vector and the commodity vector to generate a spliced vector;
the second model processing unit is used for processing the spliced vectors by adopting a logistic regression model and outputting the recommendation probability of the commodity;
and the data comparison unit is used for recommending the commodity if the recommendation probability is greater than a recommendation probability threshold value.
The specific implementation of the commodity recommendation system is basically the same as the above-mentioned examples of the commodity recommendation method, and will not be repeated here.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It should be noted that 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 use of the words first, second, third, etc. do not denote any order. These words may be interpreted as names.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (6)
1. A commodity recommendation method, characterized in that the commodity recommendation method comprises:
acquiring latest behavior log data generated when a user browses or purchases goods;
processing the latest behavior log data by adopting a preset preference model group to generate commodity preference vectors of users, wherein the method comprises the following steps: extracting data statistical characteristics of the commodity according to the latest behavior log data; forming a behavior vector of the user according to the data statistics characteristics; processing the behavior vector by adopting machine models corresponding to different types of commodities respectively to obtain first probability distribution; processing the behavior vector by adopting machine models corresponding to different brands of commodities in the same class of commodities respectively to obtain second probability distribution; splicing the first probability distribution and the second probability distribution to generate commodity preference vectors;
Processing the latest behavior log data by adopting a preset entity relation model to generate commodity vectors, wherein the method comprises the following steps: forming the latest behavior log data into triple information; wherein the triplet information includes: the first commodity, the second commodity and the behavioral relationship between the first commodity and the second commodity; calculating the triplet information by adopting the entity relation model to obtain commodity vectors;
calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector, and recommending the commodity according to the recommendation probability, wherein the commodity recommendation method comprises the following steps: splicing the commodity preference vector and the commodity vector to generate a spliced vector; processing the spliced vector by adopting a logistic regression model, and outputting the recommendation probability of the commodity; and recommending the commodity if the recommendation probability is greater than a recommendation probability threshold.
2. The method of claim 1, wherein determining the latest behavior log data generated when the user browses or purchases the commodity comprises:
acquiring all historical behavior log data generated when a user browses or purchases goods;
dividing the historical time of browsing or purchasing commodities by a user into a plurality of time windows with equal length;
And extracting historical behavior log data corresponding to the latest preset number of time windows as the latest behavior log data.
3. The method of claim 1, wherein the entity-relationship model consists of a word vector model and a translation model.
4. A merchandise recommendation system, comprising:
the data acquisition module is used for acquiring the latest behavior log data generated when a user browses or purchases goods;
the first processing module is configured to process the latest behavior log data by using a preset preference model group, and generate a commodity preference vector of a user, where the first processing module includes: extracting data statistical characteristics of the commodity according to the latest behavior log data; forming a behavior vector of the user according to the data statistics characteristics; processing the behavior vector by adopting machine models corresponding to different types of commodities respectively to obtain first probability distribution; processing the behavior vector by adopting machine models corresponding to different brands of commodities in the same class of commodities respectively to obtain second probability distribution; splicing the first probability distribution and the second probability distribution to generate commodity preference vectors;
the second processing module is configured to process the latest behavior log data by using a preset entity relationship model, and generate a commodity vector, where the second processing module includes: forming the latest behavior log data into triple information; wherein the triplet information includes: the first commodity, the second commodity and the behavioral relationship between the first commodity and the second commodity; calculating the triplet information by adopting the entity relation model to obtain commodity vectors;
The commodity recommendation module is used for calculating the recommendation probability of the commodity according to the commodity preference vector and the commodity vector and recommending the commodity according to the recommendation probability, and comprises the following steps: splicing the commodity preference vector and the commodity vector to generate a spliced vector; processing the spliced vector by adopting a logistic regression model, and outputting the recommendation probability of the commodity; and recommending the commodity if the recommendation probability is greater than a recommendation probability threshold.
5. A commodity recommendation device, characterized by comprising: a memory, a processor and a commodity recommendation program stored on the memory and executable on the processor, the commodity recommendation program when executed by the processor implementing the steps of the commodity recommendation method according to any one of claims 1 to 3.
6. A storage medium having stored thereon a commodity recommendation program which when executed by a processor implements the steps of the commodity recommendation method according to any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011027186.1A CN112132660B (en) | 2020-09-25 | 2020-09-25 | Commodity recommendation method, system, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011027186.1A CN112132660B (en) | 2020-09-25 | 2020-09-25 | Commodity recommendation method, system, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112132660A CN112132660A (en) | 2020-12-25 |
| CN112132660B true CN112132660B (en) | 2023-12-26 |
Family
ID=73840435
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011027186.1A Active CN112132660B (en) | 2020-09-25 | 2020-09-25 | Commodity recommendation method, system, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112132660B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112749996B (en) * | 2021-01-29 | 2024-02-20 | 吉林大学 | Automatic commodity recommendation method and system for unmanned lipstick vending machine |
| CN113222692B (en) * | 2021-04-27 | 2022-08-16 | 南京邮电大学 | Commodity online recommendation method and system |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010134733A (en) * | 2008-12-05 | 2010-06-17 | Dainippon Printing Co Ltd | Information recommendation device, information recommendation method, and program |
| CN104867021A (en) * | 2015-05-19 | 2015-08-26 | 腾讯科技（深圳）有限公司 | Goods homepage processing method and related apparatus |
| CN105469263A (en) * | 2014-09-24 | 2016-04-06 | 阿里巴巴集团控股有限公司 | Commodity recommendation method and device |
| CN106296242A (en) * | 2015-05-22 | 2017-01-04 | 苏宁云商集团股份有限公司 | A kind of generation method of commercial product recommending list in ecommerce and the system of generation |
| CN106485562A (en) * | 2015-09-01 | 2017-03-08 | 苏宁云商集团股份有限公司 | A kind of commodity information recommendation method based on user's history behavior and system |
| CN108109056A (en) * | 2018-01-10 | 2018-06-01 | 北京思特奇信息技术股份有限公司 | A kind of recommendation method and system of commodity |
| CN108261766A (en) * | 2018-01-18 | 2018-07-10 | 珠海金山网络游戏科技有限公司 | Game Method of Commodity Recommendation and device based on preference |
| JP2018181326A (en) * | 2017-04-06 | 2018-11-15 | ネイバー コーポレーションＮＡＶＥＲ Ｃｏｒｐｏｒａｔｉｏｎ | Personalized products recommendation using deep learning |
| CN110097433A (en) * | 2019-05-14 | 2019-08-06 | 苏州大学 | Recommended method, device, equipment and storage medium based on attention mechanism |
| CN110490685A (en) * | 2019-03-27 | 2019-11-22 | 南京国科双创信息技术研究院有限公司 | A kind of Products Show method based on big data analysis |
| CN110532544A (en) * | 2019-07-18 | 2019-12-03 | 中央民族大学 | Low-resource text tour field construction of knowledge base method and system |
| CN111178986A (en) * | 2020-02-18 | 2020-05-19 | 电子科技大学 | User-commodity preference prediction method and system |
| CN111209351A (en) * | 2020-01-02 | 2020-05-29 | 北京沃东天骏信息技术有限公司 | Object relation prediction method and device, object recommendation method and device, electronic equipment and medium |
| CN111507796A (en) * | 2020-04-07 | 2020-08-07 | 桂林电子科技大学 | Online shopping mall commodity recommendation method based on knowledge graph |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5902325B1 (en) * | 2015-01-07 | 2016-04-13 | 株式会社日立製作所 | Preference analysis system, preference analysis method |
-
2020
- 2020-09-25 CN CN202011027186.1A patent/CN112132660B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010134733A (en) * | 2008-12-05 | 2010-06-17 | Dainippon Printing Co Ltd | Information recommendation device, information recommendation method, and program |
| CN105469263A (en) * | 2014-09-24 | 2016-04-06 | 阿里巴巴集团控股有限公司 | Commodity recommendation method and device |
| CN104867021A (en) * | 2015-05-19 | 2015-08-26 | 腾讯科技（深圳）有限公司 | Goods homepage processing method and related apparatus |
| CN106296242A (en) * | 2015-05-22 | 2017-01-04 | 苏宁云商集团股份有限公司 | A kind of generation method of commercial product recommending list in ecommerce and the system of generation |
| CN106485562A (en) * | 2015-09-01 | 2017-03-08 | 苏宁云商集团股份有限公司 | A kind of commodity information recommendation method based on user's history behavior and system |
| JP2018181326A (en) * | 2017-04-06 | 2018-11-15 | ネイバー コーポレーションＮＡＶＥＲ Ｃｏｒｐｏｒａｔｉｏｎ | Personalized products recommendation using deep learning |
| CN108109056A (en) * | 2018-01-10 | 2018-06-01 | 北京思特奇信息技术股份有限公司 | A kind of recommendation method and system of commodity |
| CN108261766A (en) * | 2018-01-18 | 2018-07-10 | 珠海金山网络游戏科技有限公司 | Game Method of Commodity Recommendation and device based on preference |
| CN110490685A (en) * | 2019-03-27 | 2019-11-22 | 南京国科双创信息技术研究院有限公司 | A kind of Products Show method based on big data analysis |
| CN110097433A (en) * | 2019-05-14 | 2019-08-06 | 苏州大学 | Recommended method, device, equipment and storage medium based on attention mechanism |
| CN110532544A (en) * | 2019-07-18 | 2019-12-03 | 中央民族大学 | Low-resource text tour field construction of knowledge base method and system |
| CN111209351A (en) * | 2020-01-02 | 2020-05-29 | 北京沃东天骏信息技术有限公司 | Object relation prediction method and device, object recommendation method and device, electronic equipment and medium |
| CN111178986A (en) * | 2020-02-18 | 2020-05-19 | 电子科技大学 | User-commodity preference prediction method and system |
| CN111507796A (en) * | 2020-04-07 | 2020-08-07 | 桂林电子科技大学 | Online shopping mall commodity recommendation method based on knowledge graph |
Non-Patent Citations (1)
| Title |
|---|
| 基于用户浏览轨迹的商品推荐;郭俊霞;许文生;卢罡;;计算机科学;43(12);224-228 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112132660A (en) | 2020-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108665333B (en) | Commodity recommendation method and device, electronic equipment and storage medium | |
| CN103914492B (en) | Query word fusion method, merchandise news dissemination method and searching method and system | |
| US20180165745A1 (en) | Intelligent Recommendation Method and System | |
| CN105718184A (en) | Data processing method and apparatus | |
| CN108664513B (en) | Method, device and equipment for pushing keywords | |
| US9727906B1 (en) | Generating item clusters based on aggregated search history data | |
| CN110363604B (en) | Page generation method and device | |
| US10783561B2 (en) | Systems and methods for optimizing targeted advertising through social media interactions | |
| CN109840796B (en) | Decision factor analysis device and decision factor analysis method | |
| TW201342289A (en) | Method, web server and web browser of providing information | |
| CN110851699A (en) | Deep reinforcement learning-based information flow recommendation method, device, equipment and medium | |
| CN107423356A (en) | The processing method and processing device of evaluation information, computer-readable medium, electronic equipment | |
| CN112132660B (en) | Commodity recommendation method, system, equipment and storage medium | |
| JP6976207B2 (en) | Information processing equipment, information processing methods, and programs | |
| CN110933472B (en) | Method and device for realizing video recommendation | |
| CN108305181B (en) | Social influence determination method and device, information delivery method and device, equipment and storage medium | |
| US20220351239A1 (en) | Machine learning based methods and apparatus for automatically generating item rankings | |
| CN113689259A (en) | Commodity personalized recommendation method and system based on user behaviors | |
| CN113837842A (en) | Commodity recommendation method and equipment based on user behavior data | |
| CN104615721B (en) | For the method and system based on return of goods related information Recommendations | |
| CN111680213B (en) | Information recommendation method, data processing method and device | |
| WO2014201333A1 (en) | Systems, apparatuses and methods for providing a price point to a consumer for products in an electronic shopping cart of the consumer | |
| CN116308683B (en) | Knowledge-graph-based clothing brand positioning recommendation method, equipment and storage medium | |
| CN111787042B (en) | Method and device for pushing information | |
| CN110781399A (en) | Cross-platform information pushing method and device |
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 |