CN108805594A - Information-pushing method and device - Google Patents
Information-pushing method and device Download PDFInfo
- Publication number
- CN108805594A CN108805594A CN201710286678.4A CN201710286678A CN108805594A CN 108805594 A CN108805594 A CN 108805594A CN 201710286678 A CN201710286678 A CN 201710286678A CN 108805594 A CN108805594 A CN 108805594A
- Authority
- CN
- China
- Prior art keywords
- category
- target
- product
- user
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
-
- 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
- G06Q30/0256—User search
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
Abstract
This application discloses information-pushing methods and device.One specific implementation mode of this method includes:Extract order data and access data of the target user in the first preset time period, in default website;The data extracted are parsed, the category for the product that the category for the product that the target user is placed an order is shown with the page accessed is determined respectively, based on the category determined respectively, the target category in the category for the product that the page that the target user is accessed is shown is determined;From access extracting data feature vector in the access data extracted, corresponding with the target category, this feature vector is input to lower single prediction model training in advance, corresponding with the target category, obtains lower single prediction result accordingly;It is not less than default value in response to the obtained lower single prediction result of determination, information to be pushed that is preset, matching with the target category is pushed to the target user.The embodiment, which realizes, is imbued with targetedly information push.
Description
Technical field
This application involves field of computer technology, and in particular to Internet technical field more particularly to information-pushing method
And device.
Background technology
Information pushes, and is by certain technical standard or agreement, on the internet by pushing away also known as " Web broadcast "
The information that user needs is sent to reduce a technology of information overload.Information advancing technique by active push information to user,
User can be reduced the time spent in being searched on network.By taking electric business platform as an example, it usually needs push some products to user
Information, to help user to carry out information browse more rapidly, more rich.
However, existing information push mode typically directly manually chooses information to be pushed and user to be pushed, in turn
Information to be pushed is directly pushed to selected each user, thus, the problem of being lack of pertinence there is information push.
Invention content
The purpose of the embodiment of the present application is to propose a kind of improved information-pushing method and device, to solve background above
The technical issues of technology segment is mentioned.
In a first aspect, the embodiment of the present application provides a kind of information-pushing method, this method includes:Extraction target user exists
Order data and access data in first preset time period, in default website;Order data and access data are parsed,
The category for determining the product that the category for the product that target user is placed an order is shown with the page accessed respectively is based on distinguishing
Determining category determines the target category in the category for the product that the page that target user is accessed is shown;From what is extracted
In data, corresponding with target category access extracting data feature vectors are accessed, feature vector is input to advance instruction
Lower single prediction model experienced, corresponding with target category obtains lower single prediction result corresponding with target category, wherein place an order
Correspondence of the prediction model for characteristic feature vector and lower single prediction result;In response to the obtained lower single prediction knot of determination
Fruit is not less than default value, and information to be pushed that is preset, matching with target category is pushed to target user.
In some embodiments, based on the category determined respectively, determine that the page that target user is accessed is shown
Target category in the category of product, including:Respectively determined by extraction, the category of the product that is placed an order of target user mark and
The category for the product that the page accessed is shown identifies;The category of product being extracted, being placed an order mark is determined as the
One category is identified to generate the first category identification list, and the category for the product that the page being extracted, being accessed is shown
Mark is determined as the second category mark to generate the second category identification list;For each in the second category identification list
Two categories identify, in response to not determining each first category mark in the first category identification list with second category mark not
Match, second category mark is determined as target category mark, and the indicated category of target category mark is determined as target
Category.
In some embodiments, feature vector includes number, the access target page for the product for searching for above-mentioned target category
Number and at least one of following:Average access duration in the target pages accessed, commenting in the target pages accessed
The browsing duration in valence area, the quantity of target pages of access, the number that shopping cart is added, inquiry customer service number, target pages institute
The average positive rating for the showpiece that average discount numerical value, the target pages of the product of displaying are shown, wherein target pages are to be used for
Show the page of the product of target category.
In some embodiments, the order data in extraction target user in the first preset time period, in default website
With before accessing data, this method further includes:Extract access number of multiple users in the first preset time period, in default website
According to;For each user in multiple users, access data being extracted, corresponding with the user are parsed, really
Searching times and access times of the fixed user to the product of each category;Based on identified searching times and access times,
Determine demand degree of the user to the product of each category;If existing more than preset first numerical value in identified demand degree
The user is then determined as target user by demand degree;Wherein, demand degree of the user to the product of each category is by following steps
It obtains:For each category, product of the user to the searching times and preset second value of the product of the category is determined;
By identified product and the user to the access times of the product of the category and be determined as product of the user to the category
Demand degree.
In some embodiments, if there is the demand degree more than preset first numerical value in identified demand degree,
After the user is determined as target user, this method further includes:Extract the category mark of each category;For being confirmed as mesh
Each user for marking user, in response to determining that the user is target user, according to the user to the need of the product of each category
The sequence of degree of asking from big to small is ranked up the category mark of each category to generate category mark corresponding with the user
List;The generated category identification list of storage.
In some embodiments, the order data in extraction target user in the first preset time period, in default website
With before accessing data, this method further includes the steps that the lower single prediction model of training, including:By user in the second preset time period
Access data interior, in default website are determined as history and access data;History is accessed into data, the first preset condition of satisfaction
History access data and be determined as the first history and access data, and will history access it is in data, meet second preset condition
History accesses data and is determined as the second history access data, wherein the first historical data is with the mark that placed an order, the second history number
According to the mark that do not place an order;Extracting data first eigenvector is accessed from the first history, and is accessed in data from the second history
Extract second feature vector;It will using first eigenvector and second feature vector as input using machine learning method
Respectively as output, training obtains lower single prediction model for the mark that placed an order and the mark that do not place an order.
Second aspect, the embodiment of the present application provide a kind of information push-delivery apparatus, which includes:First extraction unit,
It is configured to order data and access data of the extraction target user in the first preset time period, in default website;Parsing is single
Member is configured to order data and accesses data and parse, determine respectively the category of product that target user is placed an order and
The category for the product that the page accessed is shown determines the page that target user is accessed based on the category determined respectively
Target category in the category of the product shown;Input unit is configured to from the access data extracted and target
Feature vector is input to training in advance, corresponding with target category by the corresponding access extracting data feature vector of category
Lower single prediction model, obtain lower single prediction result corresponding with target category, wherein lower list prediction model is used for characteristic feature
The vectorial correspondence with lower single prediction result;Push unit is configured in response to the obtained lower single prediction result of determination
Not less than default value, information to be pushed that is preset, matching with target category is pushed to target user.
In some embodiments, resolution unit includes:Extraction module is configured to extract identified, target use respectively
The category mark for the product that the category mark for the product that family is placed an order and the page accessed are shown;Generation module, configuration are used
It is identified in the category mark of product being extracted, being placed an order is determined as the first category to generate the first category identification list,
And the category for the product that the page being extracted, being accessed is shown mark is determined as the second category and is identified to generate the second product
Class identification list;Determining module is configured to identify each second category in the second category identification list, in response to
Determine that each first category mark in the first category identification list is mismatched with second category mark, by the second category mark
Knowledge is determined as target category mark, and the indicated category of target category mark is determined as target category.
In some embodiments, feature vector includes number, the access target page for the product for searching for above-mentioned target category
Number and at least one of following:Average access duration in the target pages accessed, commenting in the target pages accessed
The browsing duration in valence area, the quantity of target pages of access, the number that shopping cart is added, inquiry customer service number, target pages institute
The average positive rating for the showpiece that average discount numerical value, the target pages of the product of displaying are shown, wherein target pages are to be used for
Show the page of the product of target category.
In some embodiments, which further includes:Second extraction unit is configured to extract multiple users first in advance
If the access data in the period, in default website;First determination unit is configured to for each use in multiple users
Family parses access data being extracted, corresponding with the user, determines the user searching to the product of each category
Rope number and access times;Based on identified searching times and access times, determine the user to the product of each category
Demand degree;If there is the demand degree more than preset first numerical value in identified demand degree, which is determined as target
User;Wherein, demand degree of the user to the product of each category is obtained by following steps:For each category, determining should
Product of the user to the searching times and preset second value of the product of the category;By identified product and the user to this
The access times of the product of category and be determined as demand degree of the user to the product of the category.
In some embodiments, which further includes:Third extraction unit is configured to extract the category mark of each category
Know;Sequencing unit is configured to each user for being targeted user, is used in response to determining that the user is target
Family, the sequence of demand degree from big to small according to the user to the product of each category identify the category of each category and carry out
Sequence is to generate category identification list corresponding with the user;Storage unit is configured to store generated category mark
List.
In some embodiments, which further includes:Second determination unit is configured to user in the second preset time
Access data in section, in default website are determined as history and access data;Third determination unit is configured to history accessing number
History in, meeting the first preset condition accesses data and is determined as the first history access data, and history is accessed data
In, meet the second preset condition history access data be determined as the second history access data, wherein the first historical data band
There is the mark that placed an order, the second historical data is with the mark that do not place an order;4th extraction unit is configured to access number from the first history
Extracting data second feature vector is accessed according to middle extraction first eigenvector, and from the second history;Training unit is configured to
Using machine learning method, using first eigenvector and second feature vector as input, the mark that will place an order and not under
For single mark respectively as output, training obtains lower single prediction model.
The third aspect, the embodiment of the present application provide a kind of server, including:One or more processors;Storage device,
For storing one or more programs, when one or more of programs are executed by one or more of processors so that institute
It states one or more processors and realizes the method such as any embodiment in above-mentioned information-pushing method.
Information-pushing method and device provided by the embodiments of the present application pass through the order data and access data to being extracted
Parsed, the target category in category to determine product that the page that target user is accessed is shown, then from
The corresponding access extracting data feature vector of target category, trained in advance so as to feature based vector sum and target product
The corresponding lower single prediction model of class determines lower single prediction result, is finally more than default value in response to lower single prediction result, to mesh
Mark user pushes information to be pushed that is preset, matching with target category.It is thus possible to realize based on under target user
Target category is chosen in one-state and the analysis for accessing situation, and based on identified target user to the product of target category
Lower single prediction result determine whether pushed information and determine information to be pushed, be imbued with targetedly information to realize and push away
It send.
Description of the drawings
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the application's is other
Feature, objects and advantages will become more apparent upon:
Fig. 1 is that this application can be applied to exemplary system architecture figures therein;
Fig. 2 is the flow chart according to one embodiment of the information-pushing method of the application;
Fig. 3 is the schematic diagram according to an application scenarios of the information-pushing method of the application;
Fig. 4 is the flow chart according to another embodiment of the information-pushing method of the application;
Fig. 5 is the structural schematic diagram according to one embodiment of the information push-delivery apparatus of the application;
Fig. 6 is adapted for the structural schematic diagram of the computer system of the server for realizing the embodiment of the present application.
Specific implementation mode
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, is illustrated only in attached drawing and invent relevant part with related.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 shows the exemplary system architecture of the information-pushing method or information push-delivery apparatus that can apply the application
100。
As shown in Figure 1, system architecture 100 may include terminal device 101,102,103, network 104 and server 105.
Network 104 between terminal device 101,102,103 and server 105 provide communication link medium.Network 104 can be with
Including various connection types, such as wired, wireless communication link or fiber optic cables etc..
User can be interacted by network 104 with server 105 with using terminal equipment 101,102,103, to receive or send out
Send message etc..Various telecommunication customer end applications, such as the application of shopping class, net can be installed on terminal device 101,102,103
The application of page browsing device, searching class application, instant messaging tools, mailbox client, social platform software etc..
Terminal device 101,102,103 can be the various electronic equipments with display screen and supported web page browsing, packet
Include but be not limited to smart mobile phone, tablet computer, E-book reader, pocket computer on knee and desktop computer etc..
Server 105 can be to provide the server of various services, such as to terminal device 101,102,103 in certain website
In the data analytics server analyzed of behavioral data (such as lower forms data, access data etc.).Data analytics server
A variety of data can be extracted, and the data to being extracted carry out the processing such as analyzing, obtain corresponding handling result (such as referring to
Show the lower single prediction result whether certain user places an order to the product of certain category), it is determined whether to certain user's pushed information and really
Surely it is pushed to the information of the user.
It should be noted that the information-pushing method that the embodiment of the present application is provided generally is executed by server 105, accordingly
Ground, information push-delivery apparatus are generally positioned in server 105.
It should be pointed out that server 105 can be single server, it can also be by multiple servers or multiple servers
Cluster is constituted.
It should be understood that the number of the terminal device, network and server in Fig. 1 is only schematical.According to realization need
It wants, can have any number of terminal device, network and server.
With continued reference to Fig. 2, the flow 200 of one embodiment of the information-pushing method according to the application is shown.It is described
Information-pushing method, include the following steps:
Step 201, order data and access number of the extraction target user in the first preset time period, in default website
According to.
The electronic equipment (such as server 105 shown in FIG. 1) of information-pushing method operation thereon can be by wired
Connection type or radio connection, from the operation data for storing default website (such as some e-commerce website)
Another server (not shown in figure 1) extract target user in the first preset time period (on the day of such as or 24 hours it is interior,
In 48 hours etc.), above-mentioned default website order data and access data.In addition, the operation data of above-mentioned default website
The local of above-mentioned electronic equipment can also be stored in.At this point, above-mentioned electronic equipment directly can obtain the above order number from local
According to above-mentioned access data.Wherein, above-mentioned target user can be some pre-set user list or some user set
In user, can also be meet certain conditions (such as the big Mr. Yu's designated value of access times of website is preset to this, it is default at this
Big Mr. Yu's designated value of searching times in website etc.) user.It should be pointed out that above-mentioned radio connection may include but
It is not limited to 3G/4G connections, WiFi connections, bluetooth connection, WiMAX connections, Zigbee connections, UWB (ultra wideband) even
Connect and other it is currently known or in the future exploitation radio connections.
It should be noted that the above order data can be related in the order in presetting website to above-mentioned target user
Data.For example, the above order data can include but is not limited to order settling time, the product for the product that order is recorded is believed
Cease (such as the title of product, category mark, model, price, positive rating, discount numerical value etc.), order status (such as it is shipped,
Etc. to be dispensed) etc. data.Wherein, above-mentioned category identifies the character string that can be made of various characters, can serve to indicate that production
The category (such as electrical type, luggage class, toiletries etc.) of product.
It should also be noted that, above-mentioned access data, which can be above-mentioned target user, carries out the operations such as page access, browsing
When the various data that generate and store, wherein each page accessed can show a product, the access extracted
Can include product information (such as the title of product, category mark, model, the valence of the product that is shown of each page in data
Lattice, positive rating, discount numerical value etc.).In addition, above-mentioned access data can also include with above-mentioned target user to the operation phase of the page
The information of pass, for example, search term transmitted before can include but is not limited to the network address of the accessed page, accession page, searching
Product indicated by rope time, the search term category mark, each page access time and stop access time, accessed
The page evaluation area access time and stop access time, the page accessed quantity, inquiry the customer service time, inquiry visitor
The information such as the content of clothes.
In general, user can utilize the webpage installed in client (such as terminal device shown in FIG. 1 101,102,103)
Browser accesses each webpage of above-mentioned default website.In the present embodiment, above-mentioned webpage may include html formats,
Xhtml formats, asp formats, php formats, jsp formats, shtml formats, nsp formats, the webpage of xml formats or other futures
By the webpage of the format of exploitation (as long as the web page files of this format can use browser or opening and browse it includes figure
The contents such as piece, word).
Step 202, order data and access data are parsed, determines the product for the product that target user is placed an order respectively
The category for the product that class and the page accessed are shown determines what target user was accessed based on the category determined respectively
Target category in the category for the product that the page is shown.
In the present embodiment, the above order data can include the category mark for each product that above-mentioned target user is placed an order
Know, the category mark for the product that each webpage that above-mentioned access data can be accessed comprising above-mentioned target user is shown.On
Stating electronic equipment first can parse the above order data and above-mentioned access data, to determine above-mentioned target user respectively
The category for the product that the category of the product to be placed an order and the page accessed are shown.Then, above-mentioned electronic equipment can be based on
The category determined respectively determines the target category in the category for the product that the page that above-mentioned target user is accessed is shown.
As an example, in the category for the product that above-mentioned electronic equipment can be shown the page that above-mentioned target user is accessed, on
The number for stating target user's access is more than the categories of preset times (such as 6 times, 10 inferior) and is determined as target category.As another
Example, mesh in the category for the product that above-mentioned electronic equipment can be shown the page that above-mentioned target user is accessed, above-mentioned
The category for the product that mark user does not place an order is determined as target category.As another example, above-mentioned electronic equipment can also will be above-mentioned
The number that in the category for the product that the page that target user is accessed is shown, above-mentioned target user accesses is more than preset times
And the category that above-mentioned target user does not place an order is determined as target category.
Step 203, from access extracting data features in the access data extracted, corresponding with target category to
Feature vector is input to lower single prediction model training in advance, corresponding with target category, obtained corresponding with target category by amount
Lower single prediction result.
In the present embodiment, above-mentioned electronic equipment can first parse the access data extracted, and determine above-mentioned
The category of product in the page that target user is accessed, being shown is the page of above-mentioned target category;Later, it extracts above-mentioned
Target user's data for generating and the storing during page determined by accessing and browsing, by the data be determined as with it is upper
State the corresponding access data of target category.Then, above-mentioned electronic equipment can be from identified, opposite with above-mentioned target category
The access extracting data feature vector answered.Herein, features described above vector can include for the access to above-mentioned target user
The various information characterized are operated, can also be characterized comprising the content of the page for being accessed above-mentioned target user
Various information.As an example, can include above-mentioned target user access to each page total duration, accessed it is each
The average price of product shown in a page, the information such as average sales volume.
In the present embodiment, it can be stored with lower single prediction model of multiple training in advance in above-mentioned electronic equipment, deposited
The lower single prediction model of each of storage is corresponding with a category.After extracting feature vector, above-mentioned electronic equipment can will be upper
It states feature vector and is input to lower single prediction model training in advance, corresponding with above-mentioned target category, obtain and above-mentioned target
The corresponding lower single prediction result of category.Herein, obtained lower single prediction result can serve to indicate that above-mentioned target user to upper
The prediction case whether product of target category places an order is stated, lower list prediction result can be a numerical value.Under it should be noted that
Single prediction model can be used for the correspondence of characteristic feature vector and lower single prediction result.As an example, lower list prediction model
Can to be technical staff be pre-established based on the statistics to a large amount of feature vector and lower single prediction result, it is multiple to be stored with
The mapping table of feature vector and the correspondence of lower single prediction result;Can also be technical staff based on to mass data
Count and pre-set and store it is into above-mentioned electronic equipment, to one or more of feature vector numerical value carry out numerical value meter
It calculates to obtain the calculation formula of the result of calculation for characterizing lower single prediction result, for example, the calculation formula can be by feature
The formula that the total duration to access to each page in vector is multiplied with average sales volume, obtained product can be used for
The lower single prediction result of characterization.
In some optional realization methods of the present embodiment, above-mentioned electronic equipment can train according to the following steps in advance
Lower list prediction model:
First, above-mentioned electronic equipment can extract user (such as first 30 days of current date in the second preset time period
In interior, current date first 60 days etc.) access data, and by the access data extracted be determined as history access data.
Later, above-mentioned history can be accessed history in data, meeting the first preset condition and visited by above-mentioned electronic equipment
It asks that data are determined as the first history and access data, and above-mentioned history is accessed into history in data, meeting the second preset condition
It accesses data and is determined as the second history access data.Herein, above-mentioned first preset condition can be following condition:It is user under
What category generate and store when (such as in 3 days) accession page in the preset duration before list and corresponding placed an order with user
The category of product is identical.Above-mentioned second preset condition can be following condition:Corresponding category is that user is accessing certain page
The category that does not place an order to the product of same category in above-mentioned preset duration afterwards and be that user accesses displaying in above-mentioned preset duration
There is the page of same category product to generate and stores.It has placed an order it should be pointed out that above-mentioned first historical data can carry
Mark, above-mentioned second historical data can carry the mark that do not place an order.Here, as an example, being electrical type phase in training and category
When corresponding lower single prediction model, above-mentioned electronic equipment can be by user in first 3 days of single-electron class product, access electronics
The history for generating and storing when the page of class product accesses data and is determined as the first history access data, and user can exist
Access electronic product the page after do not place an order to electronic product in 3 days in the case of, access electronic product in this 3 days
The page when generate and the history that stores accesses data and is determined as the second history and accesses data.
Then, above-mentioned electronic equipment accesses extracting data first eigenvector from above-mentioned first history, and from above-mentioned the
Two history access extracting data second feature vector.Herein, the base of extraction first eigenvector and extraction second feature vector
This method accesses extracting data feature vector with described above from identified, with above-mentioned target category corresponding
Method is essentially identical, and details are not described herein.
Finally, above-mentioned electronic equipment can utilize machine learning method, by above-mentioned first eigenvector and above-mentioned second spy
Sign vector is respectively as input, and using above-mentioned mark and the above-mentioned mark that do not place an order of having placed an order as output, training is placed an order
Prediction model.Specifically, above-mentioned electronic equipment can use model-naive Bayesian (Naive Bayesian Model, NBM)
Or the model for classification such as support vector machines (Support Vector Machine, SVM), above-mentioned first eigenvector is made
For the input of model, the above-mentioned mark that placed an order is exported as corresponding model, while using above-mentioned second feature vector as mould
Type inputs, and the above-mentioned mark that do not place an order is exported as corresponding model, using machine learning method, is trained to the model,
Obtain lower single prediction model.It should be noted that each category can correspond to lower single prediction model of a training in advance, often
The training method of lower single prediction model corresponding to one category is identical.For each category, placing an order corresponding to the category
Prediction model first eigenvector and second feature limitation respectively from the first history corresponding with the category access data and
The second history corresponding with the category accesses extracting data.
Step 204, it is not less than default value in response to the obtained lower single prediction result of determination, it is pre- to target user's push
If, the information to be pushed to match with target category.
In the present embodiment, in response to the obtained lower single prediction result of determination not less than default value (such as 0.5 or 1
Deng), above-mentioned electronic equipment can be pushed to above-mentioned target user it is preset, with above-mentioned target category match wait for push letter
Breath.Wherein, above-mentioned information to be pushed can be the product of one or more above-mentioned target categories information (such as summary info,
Picture, link etc.).
It is a schematic diagram according to the application scenarios of the information-pushing method of the present embodiment with continued reference to Fig. 3, Fig. 3.?
In the application scenarios of Fig. 3, target user's using terminal equipment 301 carries out the access operation of the page and the behaviour that places an order in default website
Make.First, server 302 extract target user in the first preset time period, this preset website order data 303 and visit
Ask data 304.Then, server 302 parses order data 303 and access data 304, determines target user institute respectively
The category for the product that the category of the product to place an order and the page accessed are shown, and then determine target category 305.Later, it takes
Device 302 be engaged in from access extracting data feature vector in the access data 304 extracted, corresponding with target category 305
306, feature vector 306 is input to lower single prediction model training in advance, corresponding with above-mentioned target category 305, is obtained down
Single prediction result 307.Finally, in response to determining that lower single prediction result 307 is not less than default value, preset and target is extracted
The information to be pushed 308 that category 305 matches, and information to be pushed 308 is pushed into terminal device used in target user
301。
The method that above-described embodiment of the application provides is parsed by the order data to being extracted with data are accessed,
The target category in category to determine product that the page that target user is accessed is shown, then from target category phase
Corresponding access extracting data feature vector, under the training in advance of feature based vector sum, corresponding with target category
Single prediction model determines lower single prediction result, is finally more than default value in response to lower single prediction result, is pushed to target user
Information to be pushed that is preset, matching with target category.It is thus possible to realize based on to the lower one-state of target user and visit
The analysis of situation is asked to choose target category, and based on identified target user to the lower single prediction of the product of target category
As a result determine whether pushed information and determine information to be pushed, be imbued with targetedly information push to realize.
With further reference to Fig. 4, it illustrates the flows 400 of another embodiment of information-pushing method.The information pushes
The flow 400 of method, includes the following steps:
Step 401, access data of multiple users in the first preset time period, in default website are extracted.
In the present embodiment, the electronic equipment (such as server 105 shown in FIG. 1) of information-pushing method operation thereon
Access data of multiple users in the first preset time period (on the day of such as), in default website can be extracted.
Step 402, for each user in multiple users, to access number being extracted, corresponding with the user
According to being parsed, searching times and access times of the user to the product of each category are determined;Based on identified search time
Number and access times, determine demand degree of the user to the product of each category;It is more than in advance if existing in identified demand degree
If the first numerical value demand degree, then the user is determined as target user.
In the present embodiment, for each user in above-mentioned multiple users, above-mentioned electronic equipment executes following steps:
The first step can parse access data being extracted, corresponding with the user, determine the user to each
The searching times and access times of the product of a category.Wherein, which can refer to the access times of the product of each category
The user has displaying the access times of the page of the product of each category.For example, there is the page of the product of electrical type to displaying
Access times be 6 times, it is 8 inferior to have the access times of the page of the product of clothing to displaying.In practice, access in data
Can include with the user to the relevant information of the access operation of the page, for example, can include but is not limited to the accessed page
Network address, inputted before accession page search term, search time, the category mark of product indicated by the search term, each
Page access time and stopping access time, the access time in the evaluation area of the page accessed and stopping access time, institute
The quantity of the page of access inquires the information such as customer service time, the content for inquiring customer service.Above-mentioned electronic equipment can carry out above-mentioned visit
Searching times and access times of the user to the product of each category are determined in statistics and the calculating for asking data.
Second step can determine product of the user to each category based on identified searching times and access times
Demand degree.Number, demand degree can be for characterizing user to the interest level of product or the numerical value of desirability.Its
In, demand degree of the user to the product of each category is obtained by following steps:Firstly, for each category, the use is determined
Product of the family to the searching times and preset second value (such as 5) of the product of the category;Later, by identified product with
The user to the access times of the product of the category and be determined as demand degree of the user to the product of the category.As showing
Example, which is 2 times to the searching times for the product that category is electrical type, and access times are 6 times, and above-mentioned second value is 5, then
The user is 16 to the demand degree for the product that category is electrical type.
Third walks, and can be compared identified demand degree and default first numerical value (such as 6), if identified need
There is the demand degree more than above-mentioned first numerical value in degree of asking, then the user can be determined as target user.
In some optional realization methods of the present embodiment, above-mentioned electronic equipment determines the operation of target user executing
Later, above-mentioned electronic equipment can extract the category mark of each category first;Later, for being targeted the every of user
One user, in response to determining that the user is target user, the production that above-mentioned electronic equipment can be according to the user to each category
The sequence of the demand degree of product from big to small is ranked up to generate product corresponding with the user category mark of each category
Class identification list;Finally, above-mentioned electronic equipment can store generated category identification list.
Step 403, order data and access number of the extraction target user in the first preset time period, in default website
According to.
In the present embodiment, above-mentioned electronic equipment can extract target user in above-mentioned first preset time period, pre-
If the order data of website and access data.
Step 404, order data and access data are parsed, determines the product for the product that target user is placed an order respectively
The category for the product that class and the page accessed are shown determines what target user was accessed based on the category determined respectively
Target category in the category for the product that the page is shown.
In the present embodiment, the above order data can include the category mark for each product that above-mentioned target user is placed an order
Know, the category mark for the product that each webpage that above-mentioned access data can be accessed comprising above-mentioned target user is shown.On
Stating electronic equipment first can parse the above order data and above-mentioned access data, to determine above-mentioned target user respectively
The category for the product that the category of the product to be placed an order and the page accessed are shown.Then, above-mentioned electronic equipment can be according to
Following steps determine target category:It is possible, firstly, to extract the category for the product that identified, above-mentioned target user is placed an order respectively
The category mark for the product that mark and the page accessed are shown.It later, can be by the product of product being extracted, being placed an order
Class mark is determined as the first category mark to generate the first category identification list, and by page institute being extracted, being accessed exhibition
The category mark of the product shown is determined as the second category and identifies to generate the second category identification list.Finally, for above-mentioned second
Each second category mark in category identification list, above-mentioned electronic equipment can identify second category and above-mentioned first
Each first category mark in category identification list is matched.In response to each in the above-mentioned first category identification list of determination
A first category mark, which is identified with second category, to be mismatched, and can be identified second category and is determined as target category mark
Know, and the indicated category of above-mentioned target category mark is determined as target category.
Step 405, from access extracting data features in the access data extracted, corresponding with target category to
Feature vector is input to lower single prediction model training in advance, corresponding with target category, obtained corresponding with target category by amount
Lower single prediction result.
In the present embodiment, above-mentioned electronic equipment can first parse the access data extracted, and determine above-mentioned
The category of product in the page that target user is accessed, being shown is the page of above-mentioned target category;It later, can be by institute
The determining page extracts above-mentioned target user and generates and store during accessing above-mentioned target pages as target pages
Data, which is determined as access data corresponding with above-mentioned target category.Then, above-mentioned electronic equipment can be from institute
Access extracting data feature vector determining, corresponding with above-mentioned target category.Features described above vector may include search
The number of the product of above-mentioned target category, the number of the access target page and at least one of following:In the target pages accessed
Average access duration, the target pages accessed evaluation area browsing duration, access target pages quantity, be added
Average discount numerical value, the target pages for the product that number, inquiry customer service number, the target pages of shopping cart are shown are shown
The average positive rating of showpiece.
In the present embodiment, it can be stored with lower single prediction model of multiple training in advance in above-mentioned electronic equipment, deposited
The lower single prediction model of each of storage is corresponding with a category.After extracting feature vector, above-mentioned electronic equipment can will be upper
It states feature vector and is input to lower single prediction model training in advance, corresponding with above-mentioned target category, obtain and above-mentioned target
The corresponding lower single prediction result of category.
Step 406, it is not less than default value in response to the obtained lower single prediction result of determination, it is pre- to target user's push
If, the information to be pushed to match with target category.
In the present embodiment, it is not less than default value in response to the obtained lower single prediction result of determination, above-mentioned electronics is set
It is standby that information to be pushed that is preset, matching with above-mentioned target category can be pushed to above-mentioned target user.
Figure 4, it is seen that compared with the corresponding embodiments of Fig. 2, the flow of the information-pushing method in the present embodiment
400 highlight to determine target user the step of.The scheme of the present embodiment description can be based on the analysis to accessing data as a result,
And determine target user, need not manually choose or determine, to realize be imbued with targetedly information push while, drop
Low human cost.
With further reference to Fig. 5, as the realization to method shown in above-mentioned each figure, this application provides a kind of push of information to fill
The one embodiment set, the device embodiment is corresponding with embodiment of the method shown in Fig. 2, which specifically can be applied to respectively
In kind electronic equipment.
As shown in figure 5, the information push-delivery apparatus 500 described in the present embodiment includes:First extraction unit 501, is configured to
Extract order data and access data of the target user in the first preset time period, in default website;Resolution unit 502, matches
It sets for being parsed to the above order data and above-mentioned access data, determines the product that above-mentioned target user is placed an order respectively
The category for the product that category and the page accessed are shown determines above-mentioned target user institute based on the category determined respectively
Target category in the category for the product that the page of access is shown;Input unit 503 is configured to from the access number extracted
Access extracting data feature vector in, corresponding with above-mentioned target category, features described above vector is input in advance
Lower single prediction model trained, corresponding with above-mentioned target category obtains lower single prediction result corresponding with above-mentioned target category,
Wherein, correspondence of the above-mentioned lower single prediction model for characteristic feature vector and lower single prediction result;Push unit 504, matches
It sets for being not less than default value in response to the obtained lower single prediction result of determination, preset to above-mentioned target user push,
The information to be pushed to match with above-mentioned target category.
In the present embodiment, above-mentioned first extraction unit 501 can extract target user in the first preset time period,
The order data and access data of above-mentioned default website.
In the present embodiment, above-mentioned resolution unit 502 first can carry out the above order data and above-mentioned access data
Parsing, with the product for the product that the category for the product that determining above-mentioned target user is placed an order respectively is shown with the page accessed
Class.Then, the product for the product that the page that above-mentioned target user is accessed is shown can be determined based on the category determined respectively
Target category in class.
In some optional realization methods of the present embodiment, above-mentioned resolution unit 502 can also include extraction module, life
At module and determining module (not shown).Wherein, said extracted module may be configured to extract respectively determined by, on
The category for stating the product that target user is placed an order identifies the category mark of the product shown with the page accessed.Above-mentioned generation
Module may be configured to the category mark of product being extracted, being placed an order being determined as the first category mark to generate first
Category identification list, and the category for the product that the page being extracted, being accessed is shown mark is determined as the second category mark
Know to generate the second category identification list.Above-mentioned determining module may be configured in above-mentioned second category identification list
Each second category identify, in response in the above-mentioned first category identification list of determination each first category mark with this second
Category mark mismatches, and second category mark is determined as target category mark, and will be indicated by above-mentioned target category mark
Category be determined as target category.
In the present embodiment, above-mentioned input unit 503 can first parse the access data extracted, in determination
The category for stating product in the page that target user is accessed, being shown is the page of above-mentioned target category;Later, in extraction
State target user's data for generating and the storing during page determined by accessing and browsing, which is determined as and
The above-mentioned corresponding access data of target category.It then, can be from access number identified, corresponding with above-mentioned target category
According to middle extraction feature vector.After extracting feature vector, features described above vector can be input to train in advance and above-mentioned mesh
The corresponding lower single prediction model of category is marked, lower single prediction result corresponding with above-mentioned target category is obtained.
In some optional realization methods of the present embodiment, features described above vector includes that may search for above-mentioned target category
The number of product, the number of the access target page and at least one of following:In the average access of the target pages accessed
The quantity of long target pages in the browsing duration in the evaluation area of the target pages accessed, access, time that shopping cart is added
The showpiece that average discount numerical value, the target pages for the product that number, inquiry customer service number, target pages are shown are shown is averaged
Positive rating.
In the present embodiment, above-mentioned push unit 504 is not less than default in response to the obtained lower single prediction result of determination
Numerical value can push information to be pushed that is preset, matching with above-mentioned target category to above-mentioned target user.Wherein, above-mentioned
Information to be pushed can be the information (such as summary info, picture, link etc.) of the product of one or more above-mentioned target categories.
In some optional realization methods of the present embodiment, above- mentioned information pusher 500 can also be carried including second
Take unit and the first determination unit (not shown).Wherein, above-mentioned second extraction unit may be configured to extract multiple use
Access data of the family in the first preset time period, in default website.Above-mentioned first determination unit may be configured to for upper
Each user in multiple users is stated, access data being extracted, corresponding with the user are parsed, determine the use
Searching times and access times of the family to the product of each category;Based on identified searching times and access times, determining should
Demand degree of the user to the product of each category;If there is the demand more than preset first numerical value in identified demand degree
Degree, then be determined as target user by the user;Wherein, demand degree of the user to the product of each category is obtained by following steps
?:For each category, product of the user to the searching times and preset second value of the product of the category is determined;It will
Identified product and the user to the access times of the product of the category and be determined as the user to the product of the category
Demand degree.
In some optional realization methods of the present embodiment, above- mentioned information pusher 500 can also be carried including third
Take unit, sequencing unit and storage unit (not shown).Wherein, it is each to may be configured to extraction for above-mentioned third extraction unit
The category of a category identifies.Above-mentioned sequencing unit may be configured to each user for being targeted user, ring
Ying Yu determines that the user is target user, the sequence of demand degree from big to small according to the user to the product of each category, right
The category mark of each category is ranked up to generate category identification list corresponding with the user.Said memory cells can be with
It is configured to store generated category identification list.
In some optional realization methods of the present embodiment, above- mentioned information pusher 500 can also include second true
Order member and third determination unit (not shown).Wherein, above-mentioned second determination unit may be configured to user
Access data in two preset time periods, in default website are determined as history and access data.Above-mentioned third determination unit can match
It sets and is determined as the access of the first history for above-mentioned history to be accessed history access data in data, meeting the first preset condition
Data, and above-mentioned history is accessed into history access data in data, meeting the second preset condition and is determined as the visit of the second history
Ask data, wherein above-mentioned first historical data is with the mark that placed an order, and above-mentioned second historical data is with the mark that do not place an order;The
Four extraction units are configured to access extracting data first eigenvector from above-mentioned first history, and from above-mentioned second history
Access extracting data second feature vector;Training unit, be configured to utilize machine learning method, by above-mentioned fisrt feature to
Amount and above-mentioned second feature vector identify above-mentioned placed an order mark and above-mentioned do not place an order as defeated respectively as input
Go out, training obtains lower single prediction model.
The device that above-described embodiment of the application provides, extracts the first extraction unit 501 by resolution unit 502
It order data and accesses data and is parsed, in the category to determine product that the page that target user is accessed is shown
Target category, then input unit 503 is from access extracting data feature vector corresponding with target category, to be based on spy
It levies lower single prediction model that vector sum is trained in advance, corresponding with target category and determines lower single prediction result, last push unit
504 are more than default value in response to lower single prediction result, to target user push it is preset, wait pushing away with what target category matched
It delivers letters breath.It is thus possible to realize based on to target user lower one-state and access the analysis of situation and choose target category, with
And pushed information and determination, which wait, to be determined whether to lower single prediction result of the product of target category based on identified target user
Pushed information is imbued with targetedly information push to realize.
Below with reference to Fig. 6, it illustrates the computer systems 600 suitable for the server for realizing the embodiment of the present application
Structural schematic diagram.Server shown in Fig. 6 is only an example, should not be to the function and use scope band of the embodiment of the present application
Carry out any restrictions.
As shown in fig. 6, computer system 600 includes central processing unit (CPU) 601, it can be read-only according to being stored in
Program in memory (ROM) 602 or be loaded into the program in random access storage device (RAM) 603 from storage section 608 and
Execute various actions appropriate and processing.In RAM 603, also it is stored with system 600 and operates required various programs and data.
CPU 601, ROM 602 and RAM 603 are connected with each other by bus 604.Input/output (I/O) interface 605 is also connected to always
Line 604.
It is connected to I/O interfaces 605 with lower component:Importation 606 including keyboard, mouse etc.;It is penetrated including such as cathode
The output par, c 607 of spool (CRT), liquid crystal display (LCD) etc. and loud speaker etc.;Storage section 608 including hard disk etc.;
And the communications portion 609 of the network interface card including LAN card, modem etc..Communications portion 609 via such as because
The network of spy's net executes communication process.Driver 610 is also according to needing to be connected to I/O interfaces 605.Detachable media 611, such as
Disk, CD, magneto-optic disk, semiconductor memory etc. are mounted on driver 610, as needed in order to be read from thereon
Computer program be mounted into storage section 608 as needed.
Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description
Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium
On computer program, which includes the program code for method shown in execution flow chart.In such reality
It applies in example, which can be downloaded and installed by communications portion 609 from network, and/or from detachable media
611 are mounted.When the computer program is executed by central processing unit (CPU) 601, limited in execution the present processes
Above-mentioned function.It should be noted that computer-readable medium described herein can be computer-readable signal media or
Computer readable storage medium either the two arbitrarily combines.Computer readable storage medium for example can be --- but
Be not limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or arbitrary above combination.
The more specific example of computer readable storage medium can include but is not limited to:Electrical connection with one or more conducting wires,
Portable computer diskette, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only deposit
Reservoir (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory
Part or above-mentioned any appropriate combination.In this application, computer readable storage medium can any be included or store
The tangible medium of program, the program can be commanded the either device use or in connection of execution system, device.And
In the application, computer-readable signal media may include the data letter propagated in a base band or as a carrier wave part
Number, wherein carrying computer-readable program code.Diversified forms may be used in the data-signal of this propagation, including but not
It is limited to electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be computer
Any computer-readable medium other than readable storage medium storing program for executing, the computer-readable medium can send, propagate or transmit use
In by instruction execution system, device either device use or program in connection.Include on computer-readable medium
Program code can transmit with any suitable medium, including but not limited to:Wirelessly, electric wire, optical cable, RF etc., Huo Zheshang
Any appropriate combination stated.
Flow chart in attached drawing and block diagram, it is illustrated that according to the system of the various embodiments of the application, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part for a part for one module, program segment, or code of table, the module, program segment, or code includes one or more uses
The executable instruction of the logic function as defined in realization.It should also be noted that in some implementations as replacements, being marked in box
The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually
It can be basically executed in parallel, they can also be executed in the opposite order sometimes, this is depended on the functions involved.Also it to note
Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding
The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
Being described in unit involved in the embodiment of the present application can be realized by way of software, can also be by hard
The mode of part is realized.Described unit can also be arranged in the processor, for example, can be described as:A kind of processor packet
Include the first extraction unit, resolution unit, input unit and push unit.Wherein, the title of these units is under certain conditions simultaneously
The restriction to the unit itself is not constituted, for example, extraction unit is also described as, " extraction target user is when first is default
Between in section, default website order data and access the units of data ".
As on the other hand, present invention also provides a kind of computer-readable medium, which can be
Included in device described in above-described embodiment;Can also be individualism, and without be incorporated the device in.Above-mentioned calculating
Machine readable medium carries one or more program, when said one or multiple programs are executed by the device so that should
Device:Extract order data and access data of the target user in the first preset time period, in default website;To what is extracted
Data are parsed, and determine the product that the category for the product that the target user is placed an order is shown with the page accessed respectively
Category determines the mesh in the category for the product that the page that the target user is accessed is shown based on the category determined respectively
Mark category;From access extracting data feature vector in the access data extracted, corresponding with the target category, by this
Feature vector is input to lower single prediction model training in advance, corresponding with the target category, obtains descending single prediction knot accordingly
Fruit;It is not less than default value in response to the obtained lower single prediction result of determination, the preset and mesh is pushed to the target user
The information to be pushed that mark category matches.
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.People in the art
Member should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from foregoing invention design, it is carried out by above-mentioned technical characteristic or its equivalent feature
Other technical solutions of arbitrary combination and formation.Such as features described above has similar work(with (but not limited to) disclosed herein
Can technical characteristic replaced mutually and the technical solution that is formed.
Claims (14)
1. a kind of information-pushing method, which is characterized in that the method includes:
Extract order data and access data of the target user in the first preset time period, in default website;
The order data and the access data are parsed, determine the product for the product that the target user is placed an order respectively
The category for the product that class and the page accessed are shown determines that the target user is visited based on the category determined respectively
Target category in the category for the product that the page asked is shown;
From access extracting data feature vector in the access data extracted, corresponding with the target category, by institute
It states feature vector and is input to lower single prediction model training in advance, corresponding with the target category, obtain and the target product
The corresponding lower single prediction result of class, wherein pair of the lower single prediction model for characteristic feature vector and lower single prediction result
It should be related to;
Be not less than default value in response to the obtained lower single prediction result of determination, to the target user push it is preset, with
The information to be pushed that the target category matches.
2. information-pushing method according to claim 1, which is characterized in that it is described based on the category determined respectively, really
Target category in the category for the product that the page that the fixed target user is accessed is shown, including:
Respectively determined by extraction, the category for the product that the target user is placed an order mark and the page accessed are shown
The category of product identifies;
The category mark of product being extracted, being placed an order is determined as the first category to identify to generate the first category identity column
Table, and the category of product that the page being extracted, being accessed is shown mark is determined as the second category mark to generate the
Two category identification lists;
For each second category mark in the second category identification list, identified in response to determination first category
Each first category mark in list is mismatched with second category mark, and second category mark is determined as target category
Mark, and the indicated category of target category mark is determined as target category.
3. information-pushing method according to claim 1, which is characterized in that described eigenvector includes searching for above-mentioned target
The number of the product of category, the number of the access target page and at least one of following:In the average visit of the target pages accessed
Ask duration, the browsing duration in the evaluation area of the target pages that are accessed, the target pages of access quantity, shopping cart is added
The showpiece that average discount numerical value, the target pages for the product that number, inquiry customer service number, target pages are shown are shown is put down
Equal positive rating, wherein target pages are the pages of the product for showing the target category.
4. information-pushing method according to claim 1, which is characterized in that default first in the extraction target user
In period, in the order data for presetting website and before accessing data, the method further includes:
Extract access data of multiple users in the first preset time period, in default website;
For each user in the multiple user, access data being extracted, corresponding with the user are solved
Analysis, determines searching times and access times of the user to the product of each category;Based on identified searching times and access
Number determines demand degree of the user to the product of each category;If existing in identified demand degree and being more than preset first
The demand degree of numerical value, then be determined as target user by the user;
Wherein, demand degree of the user to the product of each category is obtained by following steps:For each category, the use is determined
Product of the family to the searching times and preset second value of the product of the category;By identified product and the user to the product
The access times of the product of class and be determined as demand degree of the user to the product of the category.
5. information-pushing method according to claim 4, which is characterized in that if existing in the identified demand degree
More than the demand degree of preset first numerical value, then after the user being determined as target user, the method further includes:
Extract the category mark of each category;
For being targeted each user of user, in response to determining that the user is target user, according to the user couple
The sequence of the demand degree of the product of each category from big to small is ranked up to generate and the use category mark of each category
The corresponding category identification list in family;
The generated category identification list of storage.
6. information-pushing method according to claim 1, which is characterized in that default first in the extraction target user
In period, in the order data for presetting website and before accessing data, the method further includes the lower single prediction model of training
Step, including:
Access data by user in the second preset time period, in default website are determined as history and access data;
The history is accessed into history access data in data, meeting the first preset condition and is determined as the first history access number
According to, and the history is accessed into history access data in data, meeting the second preset condition and is determined as the access of the second history
Data, wherein first historical data is with the mark that placed an order, and second historical data is with the mark that do not place an order;
Extracting data first eigenvector is accessed from first history, and extracting data the is accessed from second history
Two feature vectors;
It will be described using the first eigenvector and the second feature vector as input using machine learning method
Respectively as output, training obtains lower single prediction model for the mark that placed an order and the mark that do not place an order.
7. a kind of information push-delivery apparatus, which is characterized in that described device includes:
First extraction unit is configured to order data of the extraction target user in the first preset time period, in default website
With access data;
Resolution unit is configured to parse the order data and the access data, determines that the target is used respectively
The category for the product that the category for the product that family is placed an order and the page accessed are shown, based on the category determined respectively, really
Target category in the category for the product that the page that the fixed target user is accessed is shown;
Input unit is configured to from access data in the access data extracted, corresponding with the target category
Feature vector is extracted, described eigenvector is input to lower single prediction model training in advance, corresponding with the target category,
Obtain lower single prediction result corresponding with the target category, wherein lower single prediction model for characteristic feature vector with
The correspondence of lower list prediction result;
Push unit is configured to be not less than default value in response to the obtained lower single prediction result of determination, to the target
User pushes information to be pushed that is preset, matching with the target category.
8. information push-delivery apparatus according to claim 7, which is characterized in that the resolution unit includes:
Extraction module is configured to determined by extracting respectively, the category for the product that the target user is placed an order identifies and institute
The category for the product that the page of access is shown identifies;
Generation module is configured to the category mark of product being extracted, being placed an order being determined as the first category mark with life
It is determined as second at the first category identification list, and by the category for the product that the page being extracted, being accessed is shown mark
Category is identified to generate the second category identification list;
Determining module is configured to for each second category mark in the second category identification list, in response to true
Each first category mark in the fixed first category identification list is mismatched with second category mark, by second category
Mark is determined as target category mark, and the indicated category of target category mark is determined as target category.
9. information push-delivery apparatus according to claim 7, which is characterized in that described eigenvector includes searching for above-mentioned target
The number of the product of category, the number of the access target page and at least one of following:In the average visit of the target pages accessed
Ask duration, the browsing duration in the evaluation area of the target pages that are accessed, the target pages of access quantity, shopping cart is added
The showpiece that average discount numerical value, the target pages for the product that number, inquiry customer service number, target pages are shown are shown is put down
Equal positive rating, wherein target pages are the pages of the product for showing the target category.
10. information push-delivery apparatus according to claim 7, which is characterized in that described device further includes:
Second extraction unit is configured to extract access data of multiple users in the first preset time period, in default website;
First determination unit, is configured to for each user in the multiple user, to it is being extracted, with user's phase
Corresponding access data are parsed, and determine searching times and access times of the user to the product of each category;Based on institute
Determining searching times and access times determine demand degree of the user to the product of each category;If identified demand degree
The middle demand degree existed more than preset first numerical value, then be determined as target user by the user;Wherein, the user is to each product
The demand degree of the product of class is obtained by following steps:For each category, search of the user to the product of the category is determined
The product of number and preset second value;By identified product and the user to the access times of the product of the category and
It is determined as demand degree of the user to the product of the category.
11. information push-delivery apparatus according to claim 10, which is characterized in that described device further includes:
Third extraction unit is configured to extract the category mark of each category;
Sequencing unit is configured to each user for being targeted user, in response to determining that the user is target
User, the sequence of demand degree from big to small according to the user to the product of each category, to the category of each category identify into
Row sequence is to generate category identification list corresponding with the user;
Storage unit is configured to store generated category identification list.
12. information push-delivery apparatus according to claim 7, which is characterized in that described device further includes:
Second determination unit is configured to user in the second preset time period, in the access data of default website being determined as
History accesses data;
Third determination unit is configured to the history accessing history access number in data, meeting the first preset condition
Data are accessed according to the first history is determined as, and the history is accessed into history in data, meeting the second preset condition and is accessed
Data are determined as the second history and access data, wherein first historical data is with the mark that placed an order, the second history number
According to the mark that do not place an order;
4th extraction unit is configured to access extracting data first eigenvector from first history, and from described the
Two history access extracting data second feature vector;
Training unit is configured to utilize machine learning method, by the first eigenvector and the second feature vector point
It Zuo Wei not input, using mark and the mark that do not place an order of having placed an order as output, training obtains lower single prediction model.
13. a kind of server, including:
One or more processors;
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors so that one or more of processors are real
The now method as described in any in claim 1-6.
14. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
The method as described in any in claim 1-6 is realized when execution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710286678.4A CN108805594B (en) | 2017-04-27 | 2017-04-27 | Information pushing method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710286678.4A CN108805594B (en) | 2017-04-27 | 2017-04-27 | Information pushing method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108805594A true CN108805594A (en) | 2018-11-13 |
CN108805594B CN108805594B (en) | 2022-04-12 |
Family
ID=64070178
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710286678.4A Active CN108805594B (en) | 2017-04-27 | 2017-04-27 | Information pushing method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108805594B (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109903126A (en) * | 2019-02-01 | 2019-06-18 | 北京字节跳动网络技术有限公司 | Comment on processing method, device and its relevant device that list is commented in region |
CN109993583A (en) * | 2019-04-02 | 2019-07-09 | 深圳市腾讯信息技术有限公司 | Information-pushing method and device, storage medium and electronic device |
CN110266805A (en) * | 2019-06-28 | 2019-09-20 | 京东数字科技控股有限公司 | Information-pushing method, device, electronic equipment and readable medium |
CN110675217A (en) * | 2019-09-05 | 2020-01-10 | 广州亚美信息科技有限公司 | Personalized background image generation method and device |
CN110910180A (en) * | 2019-12-02 | 2020-03-24 | 北京嘀嘀无限科技发展有限公司 | Information pushing method and device, electronic equipment and storage medium |
CN111242713A (en) * | 2018-11-29 | 2020-06-05 | 北京京东尚科信息技术有限公司 | Information pushing method and device |
CN111506643A (en) * | 2019-01-31 | 2020-08-07 | 北京沃东天骏信息技术有限公司 | Method, device and system for generating information |
CN111598597A (en) * | 2019-02-21 | 2020-08-28 | 北京京东尚科信息技术有限公司 | Method and apparatus for transmitting information |
WO2020220757A1 (en) * | 2019-04-29 | 2020-11-05 | 阿里巴巴集团控股有限公司 | Method and device for pushing object to user based on reinforcement learning model |
US10902298B2 (en) | 2019-04-29 | 2021-01-26 | Alibaba Group Holding Limited | Pushing items to users based on a reinforcement learning model |
CN112330059A (en) * | 2020-11-24 | 2021-02-05 | 北京沃东天骏信息技术有限公司 | Method, apparatus, electronic device, and medium for generating prediction score |
CN112950304A (en) * | 2019-12-11 | 2021-06-11 | 北京沃东天骏信息技术有限公司 | Information pushing method, device, equipment and storage medium |
CN113191797A (en) * | 2021-04-20 | 2021-07-30 | 北京异乡旅行网络科技有限公司 | Broadcast promotion method, device and storage medium suitable for peripheral supermarket |
CN113469767A (en) * | 2020-03-31 | 2021-10-01 | 珠海优特智厨科技有限公司 | Order information processing method, device, equipment and computer readable storage medium |
CN113763112A (en) * | 2021-02-25 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Information pushing method and device |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140046939A1 (en) * | 2012-08-08 | 2014-02-13 | Tencent Technology (Shenzhen) Company Limited; | User recommendation method and a user recommendation system using the same |
CN105224623A (en) * | 2015-09-22 | 2016-01-06 | 北京百度网讯科技有限公司 | The training method of data model and device |
CN105528374A (en) * | 2014-10-21 | 2016-04-27 | 苏宁云商集团股份有限公司 | A commodity recommendation method in electronic commerce and a system using the same |
CN106202516A (en) * | 2016-07-24 | 2016-12-07 | 广东聚联电子商务股份有限公司 | A kind of e-commerce platform merchandise display method according to timing node |
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 |
CN106339393A (en) * | 2015-07-09 | 2017-01-18 | 阿里巴巴集团控股有限公司 | Information push method and device |
-
2017
- 2017-04-27 CN CN201710286678.4A patent/CN108805594B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140046939A1 (en) * | 2012-08-08 | 2014-02-13 | Tencent Technology (Shenzhen) Company Limited; | User recommendation method and a user recommendation system using the same |
CN105528374A (en) * | 2014-10-21 | 2016-04-27 | 苏宁云商集团股份有限公司 | A commodity recommendation method in electronic commerce and a system using the same |
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 |
CN106339393A (en) * | 2015-07-09 | 2017-01-18 | 阿里巴巴集团控股有限公司 | Information push method and device |
CN105224623A (en) * | 2015-09-22 | 2016-01-06 | 北京百度网讯科技有限公司 | The training method of data model and device |
CN106202516A (en) * | 2016-07-24 | 2016-12-07 | 广东聚联电子商务股份有限公司 | A kind of e-commerce platform merchandise display method according to timing node |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111242713A (en) * | 2018-11-29 | 2020-06-05 | 北京京东尚科信息技术有限公司 | Information pushing method and device |
CN111506643B (en) * | 2019-01-31 | 2023-09-29 | 北京沃东天骏信息技术有限公司 | Method, device and system for generating information |
CN111506643A (en) * | 2019-01-31 | 2020-08-07 | 北京沃东天骏信息技术有限公司 | Method, device and system for generating information |
CN109903126A (en) * | 2019-02-01 | 2019-06-18 | 北京字节跳动网络技术有限公司 | Comment on processing method, device and its relevant device that list is commented in region |
CN111598597A (en) * | 2019-02-21 | 2020-08-28 | 北京京东尚科信息技术有限公司 | Method and apparatus for transmitting information |
CN109993583A (en) * | 2019-04-02 | 2019-07-09 | 深圳市腾讯信息技术有限公司 | Information-pushing method and device, storage medium and electronic device |
WO2020220757A1 (en) * | 2019-04-29 | 2020-11-05 | 阿里巴巴集团控股有限公司 | Method and device for pushing object to user based on reinforcement learning model |
US10902298B2 (en) | 2019-04-29 | 2021-01-26 | Alibaba Group Holding Limited | Pushing items to users based on a reinforcement learning model |
CN110266805A (en) * | 2019-06-28 | 2019-09-20 | 京东数字科技控股有限公司 | Information-pushing method, device, electronic equipment and readable medium |
CN110675217A (en) * | 2019-09-05 | 2020-01-10 | 广州亚美信息科技有限公司 | Personalized background image generation method and device |
CN110910180A (en) * | 2019-12-02 | 2020-03-24 | 北京嘀嘀无限科技发展有限公司 | Information pushing method and device, electronic equipment and storage medium |
CN112950304A (en) * | 2019-12-11 | 2021-06-11 | 北京沃东天骏信息技术有限公司 | Information pushing method, device, equipment and storage medium |
CN113469767A (en) * | 2020-03-31 | 2021-10-01 | 珠海优特智厨科技有限公司 | Order information processing method, device, equipment and computer readable storage medium |
CN112330059B (en) * | 2020-11-24 | 2023-05-30 | 北京沃东天骏信息技术有限公司 | Method, apparatus, electronic device, and medium for generating predictive score |
CN112330059A (en) * | 2020-11-24 | 2021-02-05 | 北京沃东天骏信息技术有限公司 | Method, apparatus, electronic device, and medium for generating prediction score |
CN113763112A (en) * | 2021-02-25 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Information pushing method and device |
CN113191797A (en) * | 2021-04-20 | 2021-07-30 | 北京异乡旅行网络科技有限公司 | Broadcast promotion method, device and storage medium suitable for peripheral supermarket |
Also Published As
Publication number | Publication date |
---|---|
CN108805594B (en) | 2022-04-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108805594A (en) | Information-pushing method and device | |
CN105320766B (en) | Information-pushing method and device | |
CN109460513A (en) | Method and apparatus for generating clicking rate prediction model | |
CN107908789A (en) | Method and apparatus for generating information | |
CN107172151A (en) | Method and apparatus for pushed information | |
CN107908740A (en) | Information output method and device | |
CN108520324A (en) | Method and apparatus for generating information | |
CN108804450A (en) | The method and apparatus of information push | |
CN105117491B (en) | Page push method and apparatus | |
CN107295095A (en) | The method and apparatus for pushing and showing advertisement | |
CN107944481A (en) | Method and apparatus for generating information | |
CN109976997A (en) | Test method and device | |
CN107426328A (en) | Information-pushing method and device | |
CN110298716A (en) | Information-pushing method and device | |
CN110020162A (en) | User identification method and device | |
CN109388548A (en) | Method and apparatus for generating information | |
CN106897905A (en) | Method, device and electronic equipment for pushed information | |
CN108280200A (en) | Method and apparatus for pushed information | |
US20220222306A1 (en) | System, method and apparatus for presenting information | |
CN108388563A (en) | Information output method and device | |
CN108595448A (en) | Information-pushing method and device | |
CN110413872A (en) | Method and apparatus for showing information | |
CN107911449A (en) | Method and apparatus for pushed information | |
CN108810047A (en) | For determining that information pushes the method, apparatus and server of accuracy rate | |
CN107784076A (en) | The method and apparatus of visualization structure user behavior data |
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 |