CN110413870B - Commodity recommendation method and device and server - Google Patents
Commodity recommendation method and device and server Download PDFInfo
- Publication number
- CN110413870B CN110413870B CN201811549358.4A CN201811549358A CN110413870B CN 110413870 B CN110413870 B CN 110413870B CN 201811549358 A CN201811549358 A CN 201811549358A CN 110413870 B CN110413870 B CN 110413870B
- Authority
- CN
- China
- Prior art keywords
- commodity
- browsing
- commodities
- combination
- user
- 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
Images
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
Abstract
The embodiment of the invention provides a commodity recommendation method, a commodity recommendation device and a server, which are used for acquiring historical conversation data, wherein the historical conversation data comprises the following steps: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities; according to historical conversation data, at least one commodity combination is obtained, and each commodity combination comprises: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of the commodities spaced between the second commodity and the first commodity is smaller than a preset value; sorting second commodities corresponding to each first commodity according to at least one commodity combination to obtain a recommended commodity list corresponding to the first commodities; when the recommended commodity list is determined, the user browsing behavior in the historical conversation data is fully utilized, and the accuracy of the commodity recommendation result is improved; when the commodity combination is constructed, the quantity of commodities spaced between the second commodity and the first commodity is considered, and the accuracy of the commodity recommendation result is further improved.
Description
Technical Field
The embodiment of the invention relates to the technical field of internet, in particular to a commodity recommendation method, a commodity recommendation device and a server.
Background
With the rapid development of the internet, more and more people select online shopping, and users browse e-commerce websites through terminals, select desired goods and purchase the goods. The commodity recommendation method is widely applied to E-commerce, and has the value of mining the potential purchase demand of a user, avoiding the situation that the user spends a large amount of time to browse a large amount of irrelevant commodity information, helping the user to quickly find really needed commodities in a large amount of commodities and improving the shopping experience of the user.
Currently, the commodity recommendation methods generally used can be mainly classified into content-based recommendation and user group-based recommendation. In the recommendation method based on the content, according to the commodity which the user is interested in, the similarity between other commodities and the commodity is calculated, and according to the similarity, the similar commodity which the user is likely to be interested in is recommended to the user. In the recommendation method based on the user group, the evaluation of all users on commodities is scored, the users with similar preference are grouped through a clustering algorithm to obtain the user group, and the commodities purchased and browsed by the users in the group are recommended to other users in the group.
However, in a commodity recommendation scene of an e-commerce platform, the number of commodities reaches hundreds of millions of scales, and the accuracy of commodities recommended to a user is low in the existing commodity recommendation method, that is, commodities recommended to the user may not be interesting to the user, so that the user experience is low.
Disclosure of Invention
The embodiment of the invention provides a commodity recommendation method, a commodity recommendation device and a server, which are used for improving the accuracy of a commodity recommendation result and improving user experience.
In a first aspect, an embodiment of the present invention provides a commodity recommendation method, including:
obtaining historical session data, the historical session data comprising: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities;
according to the historical conversation data, at least one commodity combination is obtained, and each commodity combination comprises: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of commodities spaced between the second commodity and the first commodity is smaller than a preset value;
and sequencing the second commodities corresponding to the first commodities according to the at least one commodity combination to obtain a recommended commodity list corresponding to the first commodities.
Optionally, the sorting, according to the at least one product combination, the second products corresponding to each first product to obtain a recommended product list corresponding to the first product includes:
for each commodity combination, acquiring the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the commodity combination, the quantity of commodities spaced between the second commodity and the first commodity and the browsing stay time of each commodity between the first commodity and the second commodity;
and aiming at each first commodity, obtaining each second commodity corresponding to the first commodity according to each commodity combination, and sequencing each second commodity according to the skip relevance of each second commodity relative to the first commodity to obtain a recommended commodity list corresponding to the first commodity.
Optionally, the obtaining of the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the commodity combination, the number of the commodities spaced between the second commodity and the first commodity, and the browsing stay duration of each commodity between the first commodity and the second commodity includes:
acquiring the skip probability of the second commodity relative to the first commodity according to the browsing frequency of the first commodity and the browsing frequency of the commodity combination;
acquiring a skip coefficient of the second commodity relative to the first commodity according to the quantity of the commodities spaced between the second commodity and the first commodity;
acquiring the skip time of the second commodity relative to the first commodity according to the browsing stay time of each commodity between the first commodity and the second commodity;
and acquiring the jump relevance of the second commodity relative to the first commodity according to the jump probability, the jump coefficient and the jump duration.
Optionally, before the obtaining of the jump duration of the second commodity relative to the first commodity according to the browsing stay duration from the first commodity to each of the second commodities, the method further includes:
and acquiring the browsing stay time of each commodity in the historical conversation data according to the timestamp in the historical conversation data.
Optionally, the obtaining at least one commodity combination according to the historical conversation data includes:
cutting the historical conversation data according to the timestamp to obtain at least one conversation, wherein each conversation comprises a sequence of commodities browsed by a user in the conversation;
and for each conversation, combining each commodity and N subsequent adjacent commodities pairwise to obtain the commodity combination, wherein N is the preset value.
Optionally, after combining each commodity with the subsequent adjacent N commodities pairwise to obtain the commodity combination, the method further includes:
according to the historical conversation data, acquiring the browsing frequency of each commodity, and acquiring the browsing frequency of each commodity combination;
according to the browsing frequency of each commodity, deleting the commodity combination corresponding to the commodity which does not meet the first preset condition;
and deleting the commodity combinations which do not meet the second preset condition according to the browsing frequency of each commodity combination.
Optionally, before the cutting the historical session data according to the timestamp, the method further includes:
removing noise data in the historical session data.
Optionally, the method further includes:
the method comprises the steps of obtaining a commodity browsing request, wherein the commodity browsing request is used for indicating a user to request to browse a first commodity;
acquiring the recommended commodity list corresponding to the first commodity according to the commodity browsing request;
recommending the commodities in the recommended commodity list to the user.
Optionally, the product browsing request further includes a timestamp;
after the commodity browsing request sent by the user is obtained, the method further comprises the following steps:
recording the first item and the timestamp into the historical session data.
In a second aspect, an embodiment of the present invention provides a product recommendation device, including:
a first obtaining module, configured to obtain historical session data, where the historical session data includes: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities;
a second obtaining module, configured to obtain at least one commodity combination according to the historical session data, where each commodity combination includes: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of commodities spaced between the second commodity and the first commodity is smaller than a preset value;
and the generating module is used for sequencing the second commodities corresponding to the first commodities according to the at least one commodity combination to obtain a recommended commodity list corresponding to the first commodities.
Optionally, the generating module is specifically configured to:
for each commodity combination, acquiring the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the commodity combination, the quantity of commodities spaced between the second commodity and the first commodity and the browsing stay time of each commodity between the first commodity and the second commodity;
and aiming at each first commodity, obtaining each second commodity corresponding to the first commodity according to each commodity combination, and sequencing each second commodity according to the skip relevance of each second commodity relative to the first commodity to obtain a recommended commodity list corresponding to the first commodity.
Optionally, the generating module is specifically configured to:
acquiring the skip probability of the second commodity relative to the first commodity according to the browsing frequency of the first commodity and the browsing frequency of the commodity combination;
acquiring a skip coefficient of the second commodity relative to the first commodity according to the quantity of the commodities spaced between the second commodity and the first commodity;
acquiring the skip time of the second commodity relative to the first commodity according to the browsing stay time of each commodity between the first commodity and the second commodity;
and acquiring the jump relevance of the second commodity relative to the first commodity according to the jump probability, the jump coefficient and the jump duration.
Optionally, the second obtaining module is further configured to: and acquiring the browsing stay time of each commodity in the historical conversation data according to the timestamp in the historical conversation data.
Optionally, the second obtaining module is specifically configured to: cutting the historical conversation data according to the timestamp to obtain at least one conversation, wherein each conversation comprises a sequence of commodities browsed by a user in the conversation;
and for each conversation, combining each commodity and N subsequent adjacent commodities pairwise to obtain the commodity combination, wherein N is the preset value.
Optionally, the second obtaining module is further configured to:
according to the historical conversation data, acquiring the browsing frequency of each commodity, and acquiring the browsing frequency of each commodity combination;
according to the browsing frequency of each commodity, deleting the commodity combination corresponding to the commodity which does not meet the first preset condition;
and deleting the commodity combinations which do not meet the second preset condition according to the browsing frequency of each commodity combination.
Optionally, the second obtaining module is further configured to:
removing noise data in the historical session data.
Optionally, the apparatus further comprises: a recommendation module;
the first obtaining module is further configured to: the method comprises the steps of obtaining a commodity browsing request, wherein the commodity browsing request is used for indicating a user to request to browse a first commodity;
and the recommending module is used for acquiring the recommended commodity list corresponding to the first commodity according to the commodity browsing request and recommending commodities in the recommended commodity list to the user.
Optionally, the product browsing request further includes a timestamp;
the first obtaining module is further configured to: recording the first item and the timestamp into the historical session data.
In a third aspect, an embodiment of the present invention provides a server, including: at least one processor and memory;
the memory stores computer-executable instructions;
the at least one processor executing the computer-executable instructions stored by the memory causes the at least one processor to perform the method of any one of the first aspects.
In a fourth aspect, the present invention provides a computer-readable storage medium, in which computer-executable instructions are stored, and when a processor executes the computer-executable instructions, the method according to any one of the first aspect is implemented.
The commodity recommendation method, the commodity recommendation device and the server provided by the embodiment of the invention acquire historical conversation data, wherein the historical conversation data comprises the following steps: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities; according to the historical conversation data, at least one commodity combination is obtained, and each commodity combination comprises: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of commodities spaced between the second commodity and the first commodity is smaller than a preset value; according to the at least one commodity combination, sequencing the second commodities corresponding to the first commodities to obtain a recommended commodity list corresponding to the first commodities; when a recommended commodity list corresponding to the first commodity is determined, the user browsing behavior in the historical conversation data is fully utilized, and the accuracy of a commodity recommendation result is improved; in addition, when the commodity combination is constructed, the quantity of the commodities at intervals between the second commodity and the first commodity is considered, namely, the adjacent jump relation and the interval jump relation are considered at the same time, so that the user browsing behavior information in the historical conversation data is utilized more fully, and the accuracy of the commodity recommendation result is further improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a diagram illustrating an application scenario in which embodiments of the present invention may be applied;
fig. 2 is a first flowchart of a commodity recommendation method according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a merchandise recommendation process according to an embodiment of the present invention;
fig. 4 is a schematic flowchart of a process of determining a recommended product list corresponding to a first product according to each product combination according to an embodiment of the present invention;
fig. 5 is a schematic flowchart of obtaining jump sameness according to an embodiment of the present invention;
fig. 6 is a schematic flowchart of acquiring a combination of commodities according to historical session data according to an embodiment of the present invention;
fig. 7 is a flowchart illustrating a second method for recommending a commodity according to an embodiment of the present invention;
fig. 8 is a first schematic structural diagram of a commodity recommendation device according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of a second commodity recommending device according to an embodiment of the present invention;
fig. 10 is a schematic diagram of a hardware structure of a server according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are, for example, capable of operation in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
The embodiment of the invention is suitable for the commodity recommendation scene in the Internet field, and particularly recommends other commodities which the user may be interested in to the user according to the current commodities browsed by the user. A possible application scenario of an embodiment of the present invention is described below with reference to fig. 1. Fig. 1 is a schematic diagram of an application scenario to which an embodiment of the present invention may be applied. As shown in fig. 1, the commodity recommendation system provided by the embodiment of the present invention includes a terminal and a server. The terminal is any electronic device for a user to input information and display an output result, and includes but is not limited to: computers, smart phones, notebook computers, platform computers, intelligent wearable devices, and the like. The server is an electronic device for executing commodity recommendation.
The commodity recommendation method provided by the embodiment of the invention can be suitable for any scene needing commodity recommendation, including but not limited to: commodity recommendation of an e-commerce platform, commodity recommendation of an entity store, and the like. In a possible application scenario, taking commodity recommendation of an e-commerce platform as an example, a user accesses the e-commerce platform through a terminal and browses commodities on the e-commerce platform, the terminal sends a commodity browsing request to a server, and the server predicts commodities which the user may be interested in according to the commodity browsing request and recommends the commodities to the user terminal so that the user can select and purchase the commodities.
The commodity in the embodiment of the present invention is a commodity in a broad sense, and the commodity may be a tangible commodity, an intangible commodity, or an electronic data commodity. The tangible goods may be, for example, goods sold by an e-commerce platform or physical goods; intangible goods may be, for example, service-class goods, such as insurance products, financial products, and the like; electronic data items include, but are not limited to, music items, video items, news items, and the like.
In the application scenario of the embodiment of the present invention, the scenario may be based on commodity shopping or may be based on commodity browsing. For a scene based on commodity shopping, for example, when a user purchases a commodity on the e-commerce platform and clicks and browses a certain commodity, the server of the e-commerce platform recommends other commodities that may be of interest to the user according to the historical browsing behavior of the user. For a scenario based on merchandise browsing, for example: when a user browses music data and selects to browse a certain piece of music, the recommendation server recommends other music which may be interested in the user to the user according to the historical browsing behavior of the user.
When the recommended commodity list corresponding to the first commodity is determined, the user browsing behavior in the historical conversation data is fully utilized, and the accuracy of the commodity recommendation result is improved; in addition, when the commodity combination is constructed, the quantity of the commodities at intervals between the second commodity and the first commodity is considered, namely, the adjacent jump relation and the interval jump relation are considered at the same time, so that the user browsing behavior information in the historical conversation data is utilized more fully, and the accuracy of the commodity recommendation result is further improved.
The technical solution of the present invention will be described in detail below with specific examples. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments.
Fig. 2 is a first flowchart of a product recommendation method according to an embodiment of the present invention, where the method of this embodiment may be executed by the server in fig. 1, as shown in fig. 2, the method of this embodiment includes:
s201: obtaining historical session data, the historical session data comprising: the sequence of the products browsed by the user and the time stamp corresponding to each product browsed by the user.
The historical conversation data is formed after the server identifies the behavior of the user browsing the commodities. Specifically, the server may create a session for each user, which is used to record the behavior of the user in browsing the goods; the server may also create different sessions for each user's browsing behavior over different time periods, such as: a session is created for the user's morning browsing behavior and a session is created for the user's afternoon browsing behavior. It may be understood that the server may also have other ways of creating a session, and this is not particularly limited in this embodiment of the present invention.
Specifically, the behavior of the user browsing the commodities may be a behavior that the user clicks a certain commodity at a client of the e-commerce platform, and when the user clicks a certain commodity at the client of the e-commerce platform, the server recognizes that the user is browsing the commodity and records the commodity browsing behavior of the user in the historical session data.
In this embodiment, the historical session data may include: the sequence of the commodities browsed by the user and the time stamp corresponding to each commodity browsed by the user.
Fig. 3 is a schematic diagram of a product recommendation process according to an embodiment of the present invention, and the following description is given by way of example with reference to fig. 3. In an alternative embodiment, as shown in fig. 3, the historical session data may be in the form { [ article a, time 1], [ article B, time 2], [ article C, time 3], [ article D, time 4], … … }, i.e., the user browses article a at time 1, article B at time 2, article C at time 3, article D at time 4, and so on.
It should be noted that the historical session data acquired in this step may be historical session data corresponding to all users, or may also be historical session data corresponding to a part of users, which is not specifically limited in this embodiment of the present invention. In an optional implementation manner, all users may be divided into groups in advance according to characteristics of the users, and when a certain user needs to be recommended for a commodity, only historical session data corresponding to the user in the group to which the user belongs may be acquired for analysis.
In an optional implementation manner, the historical session data is obtained by cutting the commodity browsing log information, each historical session data corresponds to a session process of the user, and each historical session data includes a sequence of commodities browsed by the user in the session process and a timestamp corresponding to browsing each commodity.
Specifically, when the user browses the commodities, the server records the commodity browsing behavior of the user to form browsing log information. Because the browsing log information includes a plurality of session processes of the user, in this embodiment, a plurality of historical session data can be obtained by cutting the browsing log information. More specifically, according to the commodity browsing timestamp of the user and a preset time period, the browsing behavior of the user is cut to generate historical conversation data. For example: the clicked item belongs to the same session within the set time threshold T1, and the value T1 is 60 min. It can be understood that the cut historical session data includes a sequence of the commodities browsed by the user in one session and a time stamp corresponding to browsing of each commodity.
S202: according to the historical conversation data, at least one commodity combination is obtained, and each commodity combination comprises: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of the commodities spaced between the second commodity and the first commodity is smaller than a preset value.
Specifically, at least one commodity combination is obtained according to the commodity browsing behaviors of the user recorded in the historical conversation data, wherein each commodity combination indicates a skip relation between commodities when the user browses the commodities.
In this embodiment, each commodity combination includes: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, that is, each commodity combination represents that the user browses the second commodity after browsing the first commodity.
Continuing with the description of fig. 3, in an alternative implementation, according to the historical conversation data shown in fig. 3, the commodity combination obtained in this embodiment may be { (commodity a, commodity B), (commodity B, commodity C), (commodity C, and commodity D) }.
It can be understood that, since the commodities browsed by the user in the same session usually have similarity or correlation, for example, although the historical session data records that the commodities browsed by the user are the commodity a, the commodity B, the commodity C and the commodity D in sequence, due to the similarity of the commodity B, the commodity C and the commodity D, the user is likely to browse the commodity C or the commodity D after browsing the commodity a.
Therefore, the embodiment of the present invention further mines possible commodity combinations according to the historical conversation data, and forms a commodity combination with the first commodity and the second commodity which satisfy the following conditions: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of the commodities spaced between the second commodity and the first commodity is smaller than a preset value. That is to say, when a commodity combination is constructed, not only adjacent jump relations but also interval jump relations are considered, so that the obtained commodity combination is richer, and further, a commodity recommendation result obtained according to the commodity combination is more accurate.
Continuing with the description of fig. 3, in an optional implementation manner, when the preset value is 3, as shown in fig. 3, the commodity combination obtained in the embodiment of the present invention may also be { (commodity a, commodity B), (commodity a, commodity C), (commodity a, commodity D), (commodity B, commodity C), (commodity B, commodity D), (commodity C, commodity D) }.
It should be noted that, in the embodiment of the present invention, a method for obtaining a product combination according to historical conversation data is not particularly limited, as long as the obtained product combination satisfies the following conditions: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of the commodities spaced between the second commodity and the first commodity is smaller than a preset value.
It can be understood that, in the embodiment of the present invention, specific values of the preset values are not specifically limited, and may be reasonably set according to actual situations, and the foregoing examples are only examples.
S203: and sequencing the second commodities corresponding to the first commodities according to the at least one commodity combination to obtain a recommended commodity list corresponding to the first commodities.
It can be understood that, since each product combination indicates the skip relationship between the products when the user browses the products, for example: (article a, article B) instructs the user to jump to article B after browsing article a, indicating that article B is the article of interest to the user, so that article B can be recommended to the user when the user browses article a.
On the basis of the above steps, after each commodity combination is obtained, all the second commodities corresponding to each first commodity can be obtained according to each commodity combination, and the second commodities are used as the recommended commodity list corresponding to the first commodities. When the user browses the first commodities, the second commodities are recommended to the user, so that the commodities recommended to the user are guaranteed to be interesting to the user, the commodity recommending accuracy is improved, and the user experience is improved.
As will be further described with reference to fig. 3, when the obtained product combination may be { (product a, product B), (product a, product C), (product a, product D), (product B, product C), (product B, product D), (product C, product D) }, as shown in fig. 3, it may be determined that each second product corresponding to product a is product B, product C, and product D, that is, the recommended product list corresponding to product a is { product B, product C, and product D }, and similarly, the recommended product list corresponding to product B is { product C, product D }, and the recommended product list corresponding to product C is { product D }.
Therefore, when the user browses the commodity A, the commodity B, the commodity C and the commodity D can be recommended to the user; when the user browses the commodity B, the commodity C and the commodity D can be recommended to the user; when the user browses the item C, the item D may be recommended to the user.
In this embodiment, after the second commodities corresponding to each first commodity are determined, the second commodities may be sorted, and the sorted second commodities are used as a recommended commodity list.
It should be noted that, in the embodiment of the present invention, a method for sorting the second commodities is not specifically limited, and the second commodities may be sorted according to a jump correlation between the second commodities and the first commodity, or sorted according to an attribute of the second commodities, for example: the attention of the commodity, the sales volume of the commodity, the good appraisal rate of the commodity, the price of the commodity and the like.
In addition, according to the commodity recommendation method provided by the embodiment of the present invention, the process of determining the recommended commodity list according to the historical session data may be performed online or offline. In an optional implementation manner, the server performs, according to a preset update cycle, steps S201 to S203 in this embodiment offline to obtain and store a recommended product list corresponding to each product, and when the server recognizes that the user browses the product, recommends to the user according to the stored recommended product list. In another optional implementation manner, when recognizing that the user browses a certain commodity, the server performs S201 to S203 in this embodiment on line to obtain a recommended commodity list corresponding to the commodity, and recommends the commodity to the user.
The commodity recommendation method provided by the embodiment of the invention acquires historical conversation data, wherein the historical conversation data comprises the following steps: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities; according to the historical conversation data, at least one commodity combination is obtained, and each commodity combination comprises: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of commodities spaced between the second commodity and the first commodity is smaller than a preset value; according to the at least one commodity combination, sequencing the second commodities corresponding to the first commodities to obtain a recommended commodity list corresponding to the first commodities; when a recommended commodity list corresponding to the first commodity is determined, the user browsing behavior in the historical conversation data is fully utilized, and the accuracy of a commodity recommendation result is improved; in addition, when the commodity combination is constructed, the quantity of the commodities at intervals between the second commodity and the first commodity is considered, namely, the adjacent jump relation and the interval jump relation are considered at the same time, so that the user browsing behavior information in the historical conversation data is utilized more fully, and the accuracy of the commodity recommendation result is further improved.
On the basis of the above embodiments, an alternative implementation of S203 is described in detail below with reference to a specific embodiment. Fig. 4 is a schematic flowchart of a process of determining a recommended product list corresponding to a first product according to each product combination according to an embodiment of the present invention, and as shown in fig. 4, the method includes:
s401: and aiming at each commodity combination, acquiring the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the commodity combination, the quantity of the commodities spaced between the second commodity and the first commodity and the browsing stay time of each commodity from the first commodity to the second commodity.
Specifically, in this embodiment, according to each commodity combination, a jump correlation of the second commodity relative to the first commodity is determined, where the jump correlation is used to represent a possibility that the user jumps to browse the second commodity after browsing the first commodity. In this embodiment, the skip correlation may be obtained according to the following parameters: browsing frequency of the first commodity, browsing frequency of the commodity combination, commodity quantity of the interval between the second commodity and the first commodity, and browsing stay time of each commodity from the first commodity to the second commodity.
An alternative embodiment of obtaining jump dependencies is described in detail below with reference to a specific example. Fig. 5 is a schematic flowchart of obtaining jump sameness according to an embodiment of the present invention, and as shown in fig. 5, the method includes:
s4011: and acquiring the skip probability of the second commodity relative to the first commodity according to the browsing frequency of the first commodity and the browsing frequency of the commodity combination.
For convenience of description, the present embodiment is described by taking the product combination (product a, product B) as an example, that is, the first product is product a, and the second product is product B.
Specifically, the historical session data may be historical session data corresponding to a plurality of users, and each user may correspond to a plurality of historical session data. According to all historical conversation data, the browsing frequency of the commodity A can be obtained through statistics; then, after all possible commodity combinations are obtained according to the historical conversation data, the browsing frequency of the commodity combinations (commodity A and commodity B) can be obtained through statistics.
Assuming that the browsing frequency of the article a is count (a) and the browsing frequency of the article combination (article a, article B) is count (ab), in an optional embodiment, the skipping probability of the article B with respect to the article a may be: p (B | a) ═ count (ab)/count (a).
S4012: and acquiring the jump coefficient of the second commodity relative to the first commodity according to the quantity of the commodities spaced between the second commodity and the first commodity.
Still taking the combination of commodities (commodity a, commodity B) as an example, assuming that the quantity of commodities separated from the commodity B in the historical session data is λ, that is, after browsing the commodity a, the user browses λ commodities and then browses the commodity B, the skip coefficient of the commodity B relative to the commodity a can be represented by F (λ).
Where F (-) represents a functional transformation, in an alternative embodiment,
s4013: and acquiring the jump duration of the second commodity relative to the first commodity according to the browsing stay duration of each commodity between the first commodity and the second commodity.
Optionally, before S4013, the method may further include: and acquiring the browsing stay time of each commodity in the historical conversation data according to the timestamp in the historical conversation data.
Specifically, the browsing stay time of each commodity may be the difference dt between the browsing time stamps of two adjacent commodities in the historical conversation datai＝ti+1-tiTo identify. Wherein, tiAnd the browsing time stamp of the ith commodity in the historical conversation data.
In a specific implementation process, for the last commodity in the historical conversation data, since the commodity has no subsequent browsed commodity, the browsing stay time of the commodity may be set to a fixed value C1. In addition, in order to avoid the situation that the browsing stay time of each commodity is too long due to special reasons of users, a threshold value C2 can be set for the browsing stay time of each commodity, and the browsing stay time dt of a certain commodity isiWhen the threshold value C2 is exceeded, the browsing stay time of the commodity is directly set to be C2. That is, the browsing stay time of each product can be expressed by the following formula:
it can be understood that, for the values of C1 and C2, the values can be reasonably set in combination with actual situations, for example: c1 ═ 10s, C2 ═ 300 s.
After the browsing stay time of each commodity in the historical conversation data is obtained, the browsing stay time of each commodity between the first commodity and the second commodity in the commodity combination can be determined according to the browsing stay time of each commodity.
For example, assume that the historical session data is { [ merchandise A, time 1 { [ product A { ]]And [ commodity C, time 2]Time 3, [ product B]Get the browsing retention time corresponding to the commodity A dtAThe browsing retention time corresponding to the commodity C is dt when the time is 2-time 1CTime 3-time 2. For the product combination (product a, product B), t may be dtA+dtCAnd determining the browsing stay time of each commodity from the commodity A to the commodity B.
In addition, the obtained browsing duration may also be modeled, and in an optional implementation, T ═ H (T), where H (·) represents a modeling function, for example:
or h (x) log (x).
S4014: and acquiring the jump relevance of the second commodity relative to the first commodity according to the jump probability, the jump coefficient and the jump duration.
Specifically, for each commodity combination, the jump correlation of the second commodity relative to the first commodity can be obtained according to the jump probability P (B | a), the jump coefficient F (λ), and the jump duration T of the second commodity relative to the first commodity, which are obtained in the above steps.
In an alternative embodiment, the jump correlation can be expressed by the following formula:
s(B|A)＝P(B|A)*F(λ)*T
s402: and aiming at each first commodity, obtaining each second commodity corresponding to the first commodity according to each commodity combination, and sequencing each second commodity according to the skip relevance of each second commodity relative to the first commodity to obtain a recommended commodity list corresponding to the first commodity.
In this embodiment, for each second commodity corresponding to each first commodity, the second commodities can be sorted according to the jump relevance of each second commodity relative to the first commodity, so as to obtain the recommended commodity list corresponding to the first commodity, and the rationality and accuracy of the commodity sequence in the recommended commodity list are ensured.
In the following description, it is assumed that the recommended products corresponding to the product a obtained from the product combination { (product a, product B), (product a, product C), (product a, product D) } are product B, product C, and product D, respectively. If the jump correlations of the product B, the product C, and the product D with respect to the product a determined in S401 are S (B | a), S (C | a), and S (D | a), respectively, the products B, the products C, and the products D may be sorted in descending order from S (B | a), S (C | a), and S (D | a), and the sorted products may be used as a recommended product list.
In this embodiment, when determining the skip relevance of the second commodity with respect to the first commodity, the skip probability and the skip system are considered, and the skip duration, that is, the stay duration when the user browses the commodities, is also considered. It can be understood that the stay time information of the user browsing the commodities can reflect the preference degree of the user to the commodities, therefore, the stay time of the user browsing the commodities is comprehensively considered, the determined skip relevance is more in line with the preference of the user, the accuracy of the commodity recommendation result is improved, and the user experience is improved.
On the basis of the above embodiments, an alternative implementation of S202 is described in detail below with reference to a specific embodiment. Fig. 6 is a schematic flowchart of a process of acquiring a combination of commodities according to historical session data according to an embodiment of the present invention, and as shown in fig. 6, the method includes:
s601: and cutting the historical conversation data according to the timestamp to obtain at least one conversation, wherein each conversation comprises a sequence of commodities browsed by the user in the conversation.
Specifically, each historical conversation data may include a plurality of conversation processes of the user, and therefore, in this embodiment, the historical conversation data may be cut to generate a conversation sequence of the user. When the commodity is specifically cut, the user browsing behavior is cut according to the commodity browsing timestamp of the user to generate a user conversation sequence, the commodities clicked within the set time threshold T1 belong to the same conversation, and the acquirable value T1 is 60 min.
Wherein, each cut conversation comprises the sequence of commodities browsed by the user in one conversation.
Optionally, before the cutting the historical session data, the method may further include: removing noise data in the historical session data. Specifically, commodity data which does not meet preset conditions in the historical conversation data is deleted, for example: the commodity number is not in accordance with the rule, the browsing time stamp is abnormal, and the like.
S602: and for each conversation, combining each commodity and N subsequent adjacent commodities pairwise to obtain the commodity combination, wherein N is the preset value.
Specifically, for the cut commodity sequence in each session, each commodity and the subsequent adjacent N commodities are combined pairwise to obtain the commodity combination.
For example, assuming that the commodity sequence in a certain session is { commodity a, commodity B, commodity C, commodity D, commodity E, and commodity F }, and the value of N is 4, the process of obtaining the commodity combination is as follows:
for the commodity a, the commodity a and the following 4 commodities (commodity B, commodity C, commodity D, commodity E) are combined in pairs to obtain commodity combinations (commodity a, commodity B), (commodity a, commodity C), (commodity a, commodity D), (commodity a, commodity E).
For the commodity B, the commodity B and the subsequent 4 commodities (commodity C, commodity D, commodity E, commodity F) are combined in pairs to obtain commodity combinations (commodity B, commodity C), (commodity B, commodity D), (commodity B, commodity E), (commodity B, commodity F).
The acquisition processes of the commodity combinations corresponding to the commodity C, the commodity D, and the commodity E are similar, and are not described herein again.
S603: and acquiring the browsing frequency of each commodity according to the historical conversation data, and acquiring the browsing frequency of each commodity combination.
S604: according to the browsing frequency of each commodity, deleting the commodity combination corresponding to the commodity which does not meet the first preset condition; and deleting the commodity combinations which do not meet the second preset condition according to the browsing frequency of each commodity combination.
In this embodiment, after the commodity combination is obtained, the commodity combination with the low frequency may be deleted, so that the commodity combination with the high frequency remains. Specifically, a threshold T2 is set, a low-frequency commodity with a browsing frequency less than T2 is determined, and a commodity combination including the low-frequency commodity is deleted. Then, the threshold T3 is set, and the combinations of commodities whose browsing frequency is less than T3 are deleted. Therefore, the rest commodity combinations are high-frequency commodity combinations, and the recommended commodity list determined according to the high-frequency commodity combinations is more accurate.
It should be noted that, the method for acquiring the commodity frequency and the commodity combination frequency in the embodiment of the present invention is not specifically limited, and in an optional implementation, a frequent itemset algorithm may be adopted, for example: and acquiring the browsing frequency of each commodity by adopting a frequent one-item set algorithm, and acquiring the browsing frequency of each commodity combination by adopting a frequency two-item set algorithm.
Fig. 7 is a flowchart illustrating a second method for recommending a commodity according to an embodiment of the present invention, and as shown in fig. 7, the method according to the embodiment of the present invention includes:
s701: the method comprises the steps of obtaining a commodity browsing request, wherein the commodity browsing request is used for indicating a user to request to browse a first commodity.
Specifically, the commodity browsing request may be obtained according to a behavior of clicking the commodity picture by the user, and the commodity browsing request may also be obtained in other manners, which is not limited in the embodiment of the present invention.
S702: and acquiring the recommended commodity list corresponding to the first commodity according to the commodity browsing request.
S703: recommending the commodities in the recommended commodity list to the user.
According to the embodiment of the invention, when the user is identified to browse the first commodity, the commodity recommendation is carried out on the user according to the recommended commodity list corresponding to the first commodity obtained in the embodiment, so that the accuracy of the commodity recommendation result is ensured, and the user experience is favorably improved.
In addition, in an optional implementation manner, the article browsing request further includes a timestamp, and after the article browsing request is acquired, the method may further include:
s704: recording the first item and the timestamp into the historical session data.
By recording the commodity browsing behavior of the user, the recommended commodity list is updated according to continuously updated historical conversation data, the real-time performance and the accuracy of the recommended commodity list are guaranteed, and the user experience is improved.
Fig. 8 is a first schematic structural diagram of a product recommendation device according to an embodiment of the present invention, and as shown in fig. 8, a product recommendation device 800 according to an embodiment of the present invention includes: a first acquisition module 801, a second acquisition module 802 and a generation module 803.
The first obtaining module 801 is configured to obtain historical session data, where the historical session data includes: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities;
a second obtaining module 802, configured to obtain at least one commodity combination according to the historical session data, where each commodity combination includes: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of commodities spaced between the second commodity and the first commodity is smaller than a preset value;
the generating module 803 is configured to sort, according to the at least one product combination, the second products corresponding to each first product, so as to obtain a recommended product list corresponding to the first product.
The commodity recommendation device of this embodiment may be used to execute the commodity recommendation method shown in fig. 2, and the implementation principle and the technical effect are similar, which are not described herein again.
Fig. 9 is a schematic structural diagram of a second product recommendation device according to an embodiment of the present invention, and based on the embodiment shown in fig. 8, the product recommendation device according to this embodiment further includes: a recommendation module 804.
Optionally, the generating module 803 is specifically configured to:
for each commodity combination, acquiring the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the commodity combination, the quantity of commodities spaced between the second commodity and the first commodity and the browsing stay time of each commodity between the first commodity and the second commodity;
and aiming at each first commodity, obtaining each second commodity corresponding to the first commodity according to each commodity combination, and sequencing each second commodity according to the skip relevance of each second commodity relative to the first commodity to obtain a recommended commodity list corresponding to the first commodity.
Optionally, the generating module 803 is specifically configured to:
acquiring the skip probability of the second commodity relative to the first commodity according to the browsing frequency of the first commodity and the browsing frequency of the commodity combination;
acquiring a skip coefficient of the second commodity relative to the first commodity according to the quantity of the commodities spaced between the second commodity and the first commodity;
acquiring the skip time of the second commodity relative to the first commodity according to the browsing stay time of each commodity between the first commodity and the second commodity;
and acquiring the jump relevance of the second commodity relative to the first commodity according to the jump probability, the jump coefficient and the jump duration.
Optionally, the second obtaining module 802 is further configured to: and acquiring the browsing stay time of each commodity in the historical conversation data according to the timestamp in the historical conversation data.
Optionally, the second obtaining module 802 is specifically configured to: cutting the historical conversation data according to the timestamp to obtain at least one conversation, wherein each conversation comprises a sequence of commodities browsed by a user in the conversation;
and for each conversation, combining each commodity and N subsequent adjacent commodities pairwise to obtain the commodity combination, wherein N is the preset value.
Optionally, the second obtaining module 802 is further configured to:
according to the historical conversation data, acquiring the browsing frequency of each commodity, and acquiring the browsing frequency of each commodity combination;
according to the browsing frequency of each commodity, deleting the commodity combination corresponding to the commodity which does not meet the first preset condition;
and deleting the commodity combinations which do not meet the second preset condition according to the browsing frequency of each commodity combination.
Optionally, the second obtaining module 802 is further configured to:
removing noise data in the historical session data.
Optionally, the first obtaining module 801 is further configured to: the method comprises the steps of obtaining a commodity browsing request, wherein the commodity browsing request is used for indicating a user to request to browse a first commodity;
the recommending module 804 is configured to obtain the recommended commodity list corresponding to the first commodity according to the commodity browsing request, and recommend the commodities in the recommended commodity list to the user.
Optionally, the product browsing request further includes a timestamp;
the first obtaining module 801 is further configured to: recording the first item and the timestamp into the historical session data.
The commodity recommendation device provided in this embodiment may be configured to execute the commodity recommendation method in any of the above method embodiments, and the implementation principle and the technical effect are similar, which are not described herein again.
Fig. 10 is a schematic diagram of a hardware structure of a server according to an embodiment of the present invention, and as shown in fig. 10, a server 1000 according to the embodiment includes: at least one processor 1001 and memory 1002. The processor 1001 and the memory 1002 are connected to each other via a bus 1003.
In a specific implementation process, the at least one processor 1001 executes the computer-executable instructions stored in the memory 1002, so that the at least one processor 1001 executes the commodity recommendation method in any one of the method embodiments.
For a specific implementation process of the processor 1001, reference may be made to the above method embodiments, which have similar implementation principles and technical effects, and details of this embodiment are not described herein again.
In the embodiment shown in fig. 10, it should be understood that the Processor may be a Central Processing Unit (CPU), other general purpose processors, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor, or in a combination of the hardware and software modules within the processor.
The memory may comprise high speed RAM memory and may also include non-volatile storage NVM, such as at least one disk memory.
The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, the buses in the figures of the present application are not limited to only one bus or one type of bus.
The embodiment of the present invention further provides a computer-readable storage medium, where a computer execution instruction is stored in the computer-readable storage medium, and when a processor executes the computer execution instruction, the commodity recommendation method in any of the above method embodiments is implemented.
The computer-readable storage medium may be implemented by any type of volatile or non-volatile memory device or combination thereof, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk. Readable storage media can be any available media that can be accessed by a general purpose or special purpose computer.
An exemplary readable storage medium is coupled to the processor such the processor can read information from, and write information to, the readable storage medium. Of course, the readable storage medium may also be an integral part of the processor. The processor and the readable storage medium may reside in an Application Specific Integrated Circuits (ASIC). Of course, the processor and the readable storage medium may also reside as discrete components in the apparatus.
Those of ordinary skill in the art will understand that: all or a portion of the steps of implementing the above-described method embodiments may be performed by hardware associated with program instructions. The program may be stored in a computer-readable storage medium. When executed, the program performs steps comprising the method embodiments described above; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (11)
1. A method for recommending an article, comprising:
obtaining historical session data, the historical session data comprising: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities;
according to the historical conversation data, at least one commodity combination is obtained, and each commodity combination comprises: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of commodities spaced between the second commodity and the first commodity is smaller than a preset value;
for each commodity combination, acquiring the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the commodity combination, the quantity of commodities spaced between the second commodity and the first commodity and the browsing stay time of each commodity between the first commodity and the second commodity;
and aiming at each first commodity, obtaining each second commodity corresponding to the first commodity according to each commodity combination, and sequencing each second commodity according to the skip relevance of each second commodity relative to the first commodity to obtain a recommended commodity list corresponding to the first commodity.
2. The method as claimed in claim 1, wherein the obtaining the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the combination of commodities, the quantity of commodities spaced between the second commodity and the first commodity, and the browsing stay duration of each commodity from the first commodity to the second commodity comprises:
acquiring the skip probability of the second commodity relative to the first commodity according to the browsing frequency of the first commodity and the browsing frequency of the commodity combination;
acquiring a skip coefficient of the second commodity relative to the first commodity according to the quantity of the commodities spaced between the second commodity and the first commodity;
acquiring the skip time of the second commodity relative to the first commodity according to the browsing stay time of each commodity between the first commodity and the second commodity;
and acquiring the jump relevance of the second commodity relative to the first commodity according to the jump probability, the jump coefficient and the jump duration.
3. The method according to claim 2, wherein before acquiring the jump duration of the second product relative to the first product according to the browsing stay duration from the first product to each of the second products, further comprising:
and acquiring the browsing stay time of each commodity in the historical conversation data according to the timestamp in the historical conversation data.
4. The method of claim 1, wherein said obtaining at least one combination of items from said historical session data comprises:
cutting the historical conversation data according to the timestamp to obtain at least one conversation, wherein each conversation comprises a sequence of commodities browsed by a user in the conversation;
and for each conversation, combining each commodity and N subsequent adjacent commodities pairwise to obtain the commodity combination, wherein N is the preset value.
5. The method of claim 4, wherein after combining each commodity with the N subsequent adjacent commodities, pairwise, to obtain the commodity combination, further comprising:
according to the historical conversation data, acquiring the browsing frequency of each commodity, and acquiring the browsing frequency of each commodity combination;
according to the browsing frequency of each commodity, deleting the commodity combination corresponding to the commodity which does not meet the first preset condition;
and deleting the commodity combinations which do not meet the second preset condition according to the browsing frequency of each commodity combination.
6. The method of claim 4, wherein before the slicing the historical session data according to the time stamps, further comprising:
removing noise data in the historical session data.
7. The method of claim 1, further comprising:
the method comprises the steps of obtaining a commodity browsing request, wherein the commodity browsing request is used for indicating a user to request to browse a first commodity;
acquiring the recommended commodity list corresponding to the first commodity according to the commodity browsing request;
recommending the commodities in the recommended commodity list to the user.
8. The method of claim 1, wherein the merchandise browsing request further comprises a timestamp;
after the commodity browsing request sent by the user is obtained, the method further comprises the following steps:
recording the first item and the timestamp into the historical session data.
9. An article recommendation device, comprising:
a first obtaining module, configured to obtain historical session data, where the historical session data includes: the method comprises the steps of browsing a sequence of commodities by a user and browsing timestamps corresponding to the commodities;
a second obtaining module, configured to obtain at least one commodity combination according to the historical session data, where each commodity combination includes: the browsing sequence of the second commodity is behind the browsing sequence of the first commodity, and the quantity of commodities spaced between the second commodity and the first commodity is smaller than a preset value;
the generation module is used for sequencing the second commodities corresponding to the first commodities according to the at least one commodity combination to obtain a recommended commodity list corresponding to the first commodities;
the generation module is specifically configured to:
for each commodity combination, acquiring the skip relevance of the second commodity relative to the first commodity according to the browsing frequency of the first commodity, the browsing frequency of the commodity combination, the quantity of commodities spaced between the second commodity and the first commodity and the browsing stay time of each commodity between the first commodity and the second commodity;
and aiming at each first commodity, obtaining each second commodity corresponding to the first commodity according to each commodity combination, and sequencing each second commodity according to the skip relevance of each second commodity relative to the first commodity to obtain a recommended commodity list corresponding to the first commodity.
10. A server, comprising: at least one processor and memory;
the memory stores computer-executable instructions;
the at least one processor executing the computer-executable instructions stored by the memory causes the at least one processor to perform the method of any of claims 1-8.
11. A computer-readable storage medium having computer-executable instructions stored thereon which, when executed by a processor, implement the method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811549358.4A CN110413870B (en) | 2018-12-18 | 2018-12-18 | Commodity recommendation method and device and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811549358.4A CN110413870B (en) | 2018-12-18 | 2018-12-18 | Commodity recommendation method and device and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110413870A CN110413870A (en) | 2019-11-05 |
| CN110413870B true CN110413870B (en) | 2021-12-31 |
Family
ID=68357479
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811549358.4A Active CN110413870B (en) | 2018-12-18 | 2018-12-18 | Commodity recommendation method and device and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110413870B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112783842B (en) * | 2019-11-07 | 2024-03-01 | 北京沃东天骏信息技术有限公司 | Log collection method and device |
| CN111339829B (en) * | 2020-01-19 | 2021-04-06 | 海通证券股份有限公司 | User identity authentication method, device, computer equipment and storage medium |
| CN111311353B (en) * | 2020-01-19 | 2023-06-30 | 河北工业职业技术学院 | User purchase intention evaluation method, system and terminal equipment |
| CN111415219B (en) * | 2020-03-02 | 2023-08-25 | 青岛聚好联科技有限公司 | Commodity recommendation method and device based on big data of home and community shopping |
| CN112102002A (en) * | 2020-09-16 | 2020-12-18 | 珠海格力电器股份有限公司 | Commodity recommendation method and device based on knowledge graph, computer and storage medium |
| CN112258264A (en) * | 2020-09-30 | 2021-01-22 | 中智关爱通（上海）科技股份有限公司 | Welfare commodity selection method and system |
| CN113763110A (en) * | 2021-02-08 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Article recommendation method and device |
| CN113240489B (en) * | 2021-05-18 | 2024-02-09 | 广州卓铸网络科技有限公司 | Article recommendation method and device based on big data statistical analysis |
| CN113362143B (en) * | 2021-07-01 | 2023-06-16 | 北京民融惠民科技有限公司 | Internet sales recommendation method and system based on big data |
| CN116629984B (en) * | 2023-07-24 | 2024-02-06 | 中信证券股份有限公司 | Product information recommendation method, device, equipment and medium based on embedded model |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011122575A1 (en) * | 2010-03-30 | 2011-10-06 | 楽天株式会社 | Product recommendation device, product recommendation method, program, and recording medium |
| CN102629360A (en) * | 2012-03-13 | 2012-08-08 | 浙江大学 | Effective dynamic commodity recommendation method and commodity recommendation system |
| CN103839167A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Commodity candidate set recommendation method |
| CN103839169A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Personalized commodity recommendation method based on frequency matrix and text similarity |
| CN103942193A (en) * | 2013-01-17 | 2014-07-23 | 腾讯科技（深圳）有限公司 | Information pushing method and device |
| CN105894332A (en) * | 2016-04-22 | 2016-08-24 | 深圳市永兴元科技有限公司 | Commodity recommendation method, device and system based on user behavior analysis |
| CN107103487A (en) * | 2017-03-02 | 2017-08-29 | 浙江兰德纵横网络技术股份有限公司 | A kind of advertisement sending method based on big data |
-
2018
- 2018-12-18 CN CN201811549358.4A patent/CN110413870B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011122575A1 (en) * | 2010-03-30 | 2011-10-06 | 楽天株式会社 | Product recommendation device, product recommendation method, program, and recording medium |
| CN102629360A (en) * | 2012-03-13 | 2012-08-08 | 浙江大学 | Effective dynamic commodity recommendation method and commodity recommendation system |
| CN103839167A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Commodity candidate set recommendation method |
| CN103839169A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Personalized commodity recommendation method based on frequency matrix and text similarity |
| CN103942193A (en) * | 2013-01-17 | 2014-07-23 | 腾讯科技（深圳）有限公司 | Information pushing method and device |
| CN105894332A (en) * | 2016-04-22 | 2016-08-24 | 深圳市永兴元科技有限公司 | Commodity recommendation method, device and system based on user behavior analysis |
| CN107103487A (en) * | 2017-03-02 | 2017-08-29 | 浙江兰德纵横网络技术股份有限公司 | A kind of advertisement sending method based on big data |
Non-Patent Citations (1)
| Title |
|---|
| 基于移动用户浏览行为的推荐模型;丁哲;《电子科技大学学报》;20171130;第46卷(第6期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110413870A (en) | 2019-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110413870B (en) | Commodity recommendation method and device and server | |
| CN109657138B (en) | Video recommendation method and device, electronic equipment and storage medium | |
| US8990241B2 (en) | System and method for recommending queries related to trending topics based on a received query | |
| CN110543598B (en) | Information recommendation method and device and terminal | |
| CN105989004B (en) | Information delivery preprocessing method and device | |
| TWI539305B (en) | Personalized information push method and device | |
| US10348550B2 (en) | Method and system for processing network media information | |
| US20140032264A1 (en) | Data refining engine for high performance analysis system and method | |
| CN109840782B (en) | Click rate prediction method, device, server and storage medium | |
| US11176586B2 (en) | Data analysis method and system thereof | |
| US11768908B2 (en) | System and method for collection of a website in a past state and retroactive analysis thereof | |
| US9330071B1 (en) | Tag merging | |
| US20120311140A1 (en) | Method of processing web access information and server implementing same | |
| CN112052394B (en) | Professional content information recommendation method, system, terminal equipment and storage medium | |
| CN112149003B (en) | Commodity community recommendation method and device and computer equipment | |
| CN111967914A (en) | User portrait based recommendation method and device, computer equipment and storage medium | |
| CN110570271A (en) | information recommendation method and device, electronic equipment and readable storage medium | |
| CN113220974A (en) | Click rate prediction model training and search recall method, device, equipment and medium | |
| CN111177564B (en) | Product recommendation method and device | |
| CN110347922B (en) | Recommendation method, device, equipment and storage medium based on similarity | |
| CN110796505B (en) | Business object recommendation method and device | |
| CN108664550B (en) | Funnel analysis method and device for user behavior data | |
| CN110674404A (en) | Link information generation method, device, system, storage medium and electronic equipment | |
| CN114742573A (en) | Marketing analysis method and device, electronic equipment and storage medium | |
| CN114020980A (en) | Method for generating and creating content list, electronic 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 |