CN110309399A - Scene information method for pushing and device based on date classification - Google Patents
Scene information method for pushing and device based on date classification Download PDFInfo
- Publication number
- CN110309399A CN110309399A CN201810118470.6A CN201810118470A CN110309399A CN 110309399 A CN110309399 A CN 110309399A CN 201810118470 A CN201810118470 A CN 201810118470A CN 110309399 A CN110309399 A CN 110309399A
- Authority
- CN
- China
- Prior art keywords
- date
- day
- information
- category
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2457—Query processing with adaptation to user needs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Abstract
Scene information method for pushing and device provided in an embodiment of the present invention based on date classification, pass through the date information and corresponding date category building maths modec disaggregated model according to history day, according to the date information of the date classification model and target day, determine the date category of target day, according to the date category of the target day, scene information corresponding with the date category of the target day is pushed to user, so that the date category of the target day according to determined by date classification model is to for pushing scene information corresponding with the date category, so that the scene information pushed and user demand matched, effectively improve online service quality.
Description
Technical field
The present invention relates to Internet technical field more particularly to a kind of scene information method for pushing based on date classification and
Device.
Background technique
With on line to (Online To Offline, the abbreviation O2O) development of technology and the expansion of industry under line, for
Family, which provides better online service, becomes enterprise development trend.
Under many scenes, enterprise can provide a user the service of the pushed information on some lines, in general, in enterprise
When pushed information, the behavioural habits of user can be considered.But since many behavioural habits of user are to cease manner of breathing with date category
It closes, therefore, how date category to be predicted and is obtained and provide the service of better pushed information as enterprise for user
When institute important technological problems to be solved.
Summary of the invention
For it is above-mentioned refer to how date category to be predicted and is obtained with for user provide better information push
The technical issues of service, the present invention provides a kind of scene information method for pushing and device based on date classification.
On the one hand, the present invention provides a kind of scene information method for pushing based on date classification, comprising:
According to the date information of history day and corresponding date category building maths modec disaggregated model;
According to the date information of the date classification model and target day, the date category of target day is determined;
According to the date category of the target day, scene letter corresponding with the date category of the target day is pushed to user
Breath.
It is described to be built according to the date information and corresponding date category of history day in a kind of wherein optional embodiment
Vertical date classification model, comprising:
It is established according to the loss function of definition to training pattern;Wherein, the loss function for characterize date information and
Loss between date category;
According to the date information of history day and corresponding date category, to it is described be trained to training pattern so that according to
Loss reduction between the date information and date category of the history day that model obtains after training, model is after the training
The date classification model.
In a kind of wherein optional embodiment, the date information includes gregorian calendar date, lunar date, solar term letter
Breath;The date category includes working day and day off.
In a kind of wherein optional embodiment, the date category according to the target day, to user push with
The corresponding scene information of date category of the target day, comprising:
Obtain the corresponding traffic monitoring strategy of preset each date category;
According to the date type of target day, the target day is flowed according to traffic monitoring strategy corresponding with the target day
Amount is monitored, and pushes monitoring information to user.
In a kind of wherein optional embodiment, the date category according to the target day, to user push with
The corresponding scene information of date category of the target day, comprising:
The corresponding consumer consumption behavior of each date type is determined according to user's history consumption information;
According to the corresponding consumer consumption behavior of date type of the target day, corresponding merchandise news is pushed to user.
On the other hand, the present invention provides a kind of scene information driving means based on date classification, comprising:
Unit is established, for the date information and corresponding date category building maths modec disaggregated model according to history day;
Determination unit determines the date of target day for the date information according to the date classification model and target day
Classification;
Interactive unit, for the date category according to the target day, to the date class of user's push and the target day
Not corresponding scene information.
It is described to establish unit in a kind of wherein optional embodiment, it is specifically used for:
It is established according to the loss function of definition to training pattern;Wherein, the loss function for characterize date information and
Loss between date category;
According to the date information of history day and corresponding date category, to it is described be trained to training pattern so that according to
Loss reduction between the date information and date category of the history day that model obtains after training, model is after the training
The date classification model.
In a kind of wherein optional embodiment, the date information includes gregorian calendar date, lunar date, solar term letter
Breath;The date category includes working day and day off.
In a kind of wherein optional embodiment, the interactive unit is specifically used for:
Obtain the corresponding traffic monitoring strategy of preset each date category;
According to the date type of target day, the target day is flowed according to traffic monitoring strategy corresponding with the target day
Amount is monitored, and pushes monitoring information to user.
In a kind of wherein optional embodiment, the interactive unit is specifically used for:
The corresponding consumer consumption behavior of each date type is determined according to user's history consumption information;
According to the corresponding consumer consumption behavior of date type of the target day, corresponding merchandise news is pushed to user.
Scene information method for pushing and device provided by the invention based on date classification, passes through the date according to history day
Information and corresponding date category building maths modec disaggregated model, according to the date information of the date classification model and target day,
The date category for determining target day, according to the date category of the target day, to the date class of user's push and the target day
Not corresponding scene information, so that the date category of the target day according to determined by date classification model is to for pushing
Scene information corresponding with the date category, so that the scene information pushed and user demand matched, effectively mention
Online service quality is risen.
Detailed description of the invention
Through the above attached drawings, it has been shown that the specific embodiment of the disclosure will be hereinafter described in more detail.These attached drawings
It is not intended to limit the scope of this disclosure concept by any means with verbal description, but is by referring to specific embodiments
Those skilled in the art illustrate the concept of the disclosure.
Fig. 1 is a kind of process signal for scene information method for pushing based on date classification that the embodiment of the present invention one provides
Figure;
Fig. 2 is a kind of process signal of scene information method for pushing based on date classification provided by Embodiment 2 of the present invention
Figure;
Fig. 3 is that the process of another scene information method for pushing based on date classification provided by Embodiment 2 of the present invention is shown
It is intended to;
Fig. 4 is a kind of structural representation for scene information driving means based on date classification that the embodiment of the present invention three provides
Figure.
The drawings herein are incorporated into the specification and forms part of this specification, and shows the implementation for meeting the disclosure
Example, and together with specification for explaining the principles of this disclosure.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described.
With to (Online To Offline, the abbreviation O2O) development of technology and the expansion of industry under line, being utilized on line
Resource provides better online service as enterprise development trend for user in the Internet line.
Under many scenes, enterprise can provide a user the service of the pushed information on some lines, in order to enable enterprise pushes away
The information sent can be only possible with user demand matching, in general, the information of enterprise's push will with the behavior of user into
Row association.
By the analysis and observation to user behavior, it has been found that many behavioural habits of user with date category phase
It closes.By taking takeout service as an example, take-away order of the user on weekdays under classification is more than the take-away order under day off classification.This
Kind user behavior habit relevant to date category is also embodied in a series of business such as online shopping business, business of calling a taxi, scheduling business
In scene, in this case, enterprise, which can push corresponding information to user according to date category, then becomes raising online service
The essential link of quality.
Then, in this process, it can not obtain or predict in advance in the prior art the date category of target day, this will not
Conducive to the promotion of the information push accuracy of enterprise.
For it is above-mentioned refer to how date category to be predicted and is obtained with for user provide better information push
The technical issues of service, the scene information method for pushing and the device present invention that the present invention provides a kind of based on date classification provide
A kind of prediction technique with vehicle type, Fig. 1 are a kind of scene information based on date classification that the embodiment of the present invention one provides
The flow diagram of method for pushing.
As shown in Figure 1, the scene information method for pushing includes:
Step 101, date information and corresponding date category building maths modec disaggregated model according to history day.
Step 102, according to the date information of the date classification model and target day, determine the date category of target day.
Step 103, according to the date category of the target day, pushed to user corresponding with the date category of the target day
Scene information.
It should be noted that the executing subject of the scene information method for pushing provided by the invention based on date classification is specific
It may be based on the scene information driving means of date classification, which can pass through the side of hardware and/or software
Formula is realized.It can generally be integrated in the cloud server that the line upper mounting plate of enterprise is based on, the storage being based on line upper mounting plate
There is the data server of types of databases to be used cooperatively, in addition, the server that scene information driving means is based on can be with data
Server is same server, or to be under the jurisdiction of the different server of same server cluster, the present invention to this without
Limitation.
Specifically, the present invention is information and the date category of history day based on history day, to target to be predicted
The date category of day carries out prediction acquisition.In the information and the process analyzed of the date category of history day to history day
In, it has been found that there are certain relationships between the two.By taking on January 1st, 2018 as an example, since the date is New Year's Day
Holiday, corresponding date category are day off;Alternatively, by taking on January 6th, 2018 as an example, it is right since the date corresponds to Saturday
The date category answered also is day off, and by taking the lunar calendar junior one as an example, since the date is Chinese traditional festivals, date category is to stop
Cease day;For another example the Ching Ming Festival be a lunar calendar solar term, date category is also day off.In addition, based on habit over the years, one
The case where having multiple dates companies in a little special red-letter days not, such as 11 long holidays, long holidays of May Day etc., in these vacations,
Often there is the case where temporarily transferring a few thing day as day off, i.e., specified using original working day as day off, and
Using original day off as working day.
Situation based on above-mentioned analysis, the present invention are established according to the date information of above-mentioned history day and corresponding date category
Date classification model so that the date classification model can be used for the Dates Study to target day, and determines the date class of target day
Not.Wherein, furthermore, it is understood that in order to realize the Accurate Prediction to date category, the date involved in the date classification model
Classification specifically can be divided into two class of working day and day off according to whether for working day division.Certainly in other feasible realities
It applies in mode, day off can also be further divided into celebrate a festival day off and day off at weekend etc., in addition, can also be according to other factors
Carry out further division;And date information includes gregorian calendar date, lunar date, solar term information etc..
In a kind of wherein optional embodiment, a loss function can be defined first, which can be used for characterizing
Loss between date information and date category, and established according to the loss function to training pattern.Certainly, in loss function
Coefficient can refer to date continuity, and date category mobility etc. factor, date continuity therein can such as several work
Working day classification then for example can be become day off classification by the continuity of day or day off, date category mobility, or will rest
Day, classification became working day classification etc., these factors are contemplated that in the characterization to loss function.
Then, according to the date information of history day and corresponding date category, to it is described to training pattern be trained with
Make the loss reduction between the date information and date category of the history day obtained according to model after training, after the training
Model is the date classification model.It is wherein similar with the training to neural network model to the training of model, by several history
Day is trained model as sample, and using other history days as test, tests the model after training, with reality
Now the continuous training of model is adjusted, until it is loss reduction that model, which can go out the date category of output, model will after the training
As date classification model.
After completing the foundation to date classification model, the date letter of target day directly can be input to the date classification
In model, to obtain the date category of target day, that is to say, that the process can be regarded as model and calculate separately target day as work
Loss when day and day off;If the loss when target day is working day is less than the loss that the target day is day off,
Then the date category of the target day is working day;Otherwise, the date category of the target day is day off.
Finally, can be pushed to user corresponding with the date category of the target day according to the date category of the target day
Scene information, can be to if the date category in lower odd-numbered day when single under user is working day by taking takeout service above-mentioned as an example
User pushes the businessman near its job site, or pushes the letter meal information etc. for being suitble to working day to have dinner;If when single under user
The lower odd-numbered day date category be day off, then can to user push the place Qi Jia near businessman, or push be suitble to day off
The more rich food product information etc. having meal.
The scene information method for pushing based on date classification that the embodiment of the present invention one provides, passes through the day according to history day
Phase information and corresponding date category building maths modec disaggregated model are believed according to the date classification model and the date of target day
Breath, determines the date category of target day, according to the date category of the target day, to the date of user's push and the target day
The corresponding scene information of classification, so that the date category of the target day according to determined by date classification model is to for pushing away
Scene information corresponding with the date category is sent, so that the scene information pushed and user demand matched, effectively
Improve online service quality.
In order to preferably describe scene information method for pushing provided by the invention, on the basis of example 1, Fig. 2 is this
A kind of flow diagram for scene information method for pushing based on date classification that inventive embodiments two provide.
As shown in Fig. 2, the scene information method for pushing includes:
Step 201, date information and corresponding date category building maths modec disaggregated model according to history day.
Step 202, according to the date information of the date classification model and target day, determine the date category of target day.
Step 203 obtains the corresponding traffic monitoring strategy of preset each date category.
Step 204, according to the date type of target day, according to traffic monitoring strategy corresponding with the target day to described
Target daily flow is monitored, and pushes monitoring information to user.
Similarly with embodiment one, above embodiment provide scene information method for pushing in, the present invention according to
The date information of above-mentioned history day and corresponding date category building maths modec disaggregated model, so that the date classification model is available
In the Dates Study to target day, and determine the date category of target day.Wherein, furthermore, it is understood that in order to realize to date class
Other Accurate Prediction, date category involved in the date classification model specifically can be according to whether divide for working day
It is divided into two class of working day and day off.Certainly in other feasible embodiments, day off can also be further divided into and celebrate a festival
Day off and day off at weekend etc., in addition, also further division can be carried out according to other factors;And date information includes calendar day
Phase, lunar date, solar term information etc..In a kind of wherein optional embodiment, a loss function, the damage can be defined first
Losing function can be used for characterizing the loss between date information and date category, and be established according to the loss function to training pattern.
Certainly, the coefficient in loss function can refer to date continuity, date category mobility etc. factor, date continuity therein
Can such as several working days or day off continuity, working day classification then for example can become resting by date category mobility
Day classification, or day off classification is become into working day classification etc., these factors are contemplated that in the characterization to loss function.
What is different from the first embodiment is that in the present embodiment, will also obtain the corresponding stream of preset each date category
Monitoring strategies are measured, and according to the date type of target day, according to traffic monitoring strategy corresponding with the target day to the mesh
Mark daily flow is monitored, and pushes monitoring information to user.
For example, the application scenarios of present embodiment specifically can be such as the monitoring to flow information.With the business of calling a taxi
For the monitoring of flow, the order flow of business of calling a taxi differs greatly with day off on weekdays: when on weekdays, calling a taxi and orders
Single flow is generally concentrated at morning peak and the two periods of evening peak, and at day off, order flow of calling a taxi generally only collects
In 7 points to 10 points of period at night.Therefore, during monitoring, the difference in view of date category is needed, using difference
Monitoring strategies.Correspondingly, being provided through the invention during monitoring device or monitoring server are monitored flow
The method of determination of date category then supervised again using corresponding monitoring strategies with first determining the date category of target day
Control, situations such as to avoid the wrong report generated using the monitoring strategies of mistake so as to be used for or maintenance personnel push it is quasi-
The high monitoring information of exactness.
Certainly, it should be noted that in the present embodiment, the monitoring of flow is generally to real time monitoring, i.e. target
Day can also be the simulative surveillance for the flow to future anticipation to work as the day before yesterday, i.e. target day is the non-future, monitor principle class
Seemingly, then this is not repeated.In addition, monitoring strategies can set acquisition modes as the prior art, not do herein to be preset
It repeats.
On the basis of example 1, Fig. 3 is another scene based on date classification provided by Embodiment 2 of the present invention
The flow diagram of information-pushing method, as shown in figure 3, the scene information method for pushing includes:
Step 301, date information and corresponding date category building maths modec disaggregated model according to history day.
Step 302, according to the date information of the date classification model and target day, determine the date category of target day.
Step 303 determines the corresponding consumer consumption behavior of each date type according to user's history consumption information.
Step 304, the corresponding consumer consumption behavior of date type according to the target day push corresponding quotient to user
Product information.
Similarly with embodiment one, above embodiment provide scene information method for pushing in, the present invention according to
The date information of above-mentioned history day and corresponding date category building maths modec disaggregated model, so that the date classification model is available
In the Dates Study to target day, and determine the date category of target day.Wherein, furthermore, it is understood that in order to realize to date class
Other Accurate Prediction, date category involved in the date classification model specifically can be according to whether divide for working day
It is divided into two class of working day and day off.Certainly in other feasible embodiments, day off can also be further divided into and celebrate a festival
Day off and day off at weekend etc., in addition, also further division can be carried out according to other factors;And date information includes calendar day
Phase, lunar date, solar term information etc..In a kind of wherein optional embodiment, a loss function, the damage can be defined first
Losing function can be used for characterizing the loss between date information and date category, and be established according to the loss function to training pattern.
Certainly, the coefficient in loss function can refer to date continuity, date category mobility etc. factor, date continuity therein
Can such as several working days or day off continuity, working day classification then for example can become resting by date category mobility
Day classification, or day off classification is become into working day classification etc., these factors are contemplated that in the characterization to loss function.
What is different from the first embodiment is that in the present embodiment, each day will be also determined according to user's history consumption information
The corresponding consumer consumption behavior of phase type is pushed away according to the corresponding consumer consumption behavior of date type of the target day to user
Send corresponding merchandise news.
For example, present embodiment applies also for the industry in the fields O2O such as the business of calling a taxi, online shopping business, takeout service
Business.For the business of calling a taxi, it can determine that user is corresponding with day off on weekdays according to user's history order information first
Call a taxi habit, such as on weekdays when habit use trip of taxi, in day off habit share-car trip, for another example, practise on weekdays
It is used to travel to and fro between job site and family, commercial circle and family etc. are travelled to and fro between in day off habit.Then, it opens in user and calls a taxi accordingly
When software, called a taxi the information that is consistent of habit according to the date type when the day before yesterday to user's push and user: if when the day before yesterday is work
Day, then can call a taxi interface to user's push taxi, and/or, the information such as push " whether you go to company ";It was rest when the day before yesterday
Day, then can call a taxi interface to user's push express, and/or, the information such as push " whether you will go to certain market ".
And with online shopping business and takeout service for, can be first according to outside user's history online shopping order information or history
Sell order information and determine user's corresponding habit that places an order with day off on weekdays, such as on weekdays when the place of receipt be accustomed to or
Receiving meal place is company, day off habit place of receipt or to receive meal place be that family for another example is accustomed to doing to order on weekdays
Utility or working lunch were accustomed to ordering amusement article or luxurious family meal etc. on day off.
Then, when user opens corresponding online shopping software or takes out software, according to when receiving in the day before yesterday or order
Day date type place an order the order information that is consistent of habit to user's push and user: if target day is working day, can to
Family pushes restaurant information or office appliance information near working lunch or company, or directly fills out the place of receipt in order information
It is written as company etc.;If the restaurant that can be pushed near family near luxurious family meal or family to user is believed when the day before yesterday is day off
Breath or amusement article information, or the place of receipt place in order information is directly extended this as into family etc..
On the basis of example 1, embodiment two by using date classification described in aforementioned present invention determination side
Formula, to realize to user's push and the high scene information of the more matched accuracy of user demand, so as to be taken on the line of enterprise
Business quality is effectively promoted.
Fig. 4 is a kind of structural representation of the scene information driving means based on date classification provided by Embodiment 2 of the present invention
Figure, as shown in figure 4, the scene information driving means, comprising:
Model training unit 10, for obtaining date classification model according to the training of the date category of history day;
Date classification unit 20, for determining whether target day is working day according to the date classification model after training, and
Obtain the date category of the target day;
Interactive unit 20, for the date category according to the target day, to the date of user's push and the target day
The corresponding scene information of classification.
In a kind of wherein optional embodiment, the date classification unit 20 was specifically used for: according to the day after training
Loss function in phase disaggregated model, calculate separately the target day be working day and day off when loss;If the target
Loss when day is working day is less than the loss that the target day is day off, then the date category of the target day is work
Day;Otherwise, the date category of the target day is day off.
In a kind of wherein optional embodiment, the model training unit 10, for being determined in the interactive unit
Information pushed scene and before target day to be sorted, according to the gregorian calendar date information of history consecutive days, lunar date information, section
Gas information and history consecutive days corresponding date category are trained preset loss function, are divided with the date after being trained
Class model.
In a kind of wherein optional embodiment, the interactive unit 30 is specifically used for: obtaining preset each date
The corresponding traffic monitoring strategy of classification;According to the date type of target day, according to traffic monitoring plan corresponding with the target day
Slightly the target daily flow is monitored, and pushes monitoring information to user.
In a kind of wherein optional embodiment, the interactive unit 30 is specifically used for: being consumed and is believed according to user's history
Breath determines the corresponding consumer consumption behavior of each date type;According to the corresponding customer consumption row of the date type of the target day
To push corresponding merchandise news to user.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description
Specific work process and corresponding beneficial effect, can refer to corresponding processes in the foregoing method embodiment, herein no longer
It repeats.
The scene information driving means based on date classification that the embodiment of the present invention three provides, passes through the day according to history day
Phase information and corresponding date category building maths modec disaggregated model are believed according to the date classification model and the date of target day
Breath, determines the date category of target day, according to the date category of the target day, to the date of user's push and the target day
The corresponding scene information of classification, so that the date category of the target day according to determined by date classification model is to for pushing away
Scene information corresponding with the date category is sent, so that the scene information pushed and user demand matched, effectively
Improve online service quality.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above-mentioned each method embodiment can lead to
The relevant hardware of program instruction is crossed to complete.Program above-mentioned can be stored in a computer readable storage medium.The journey
When being executed, execution includes the steps that above-mentioned each method embodiment to sequence;And storage medium above-mentioned include: ROM, RAM, magnetic disk or
The various media that can store program code such as person's CD.
Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations;To the greatest extent
Pipe present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: its according to
So be possible to modify the technical solutions described in the foregoing embodiments, or to some or all of the technical features into
Row equivalent replacement;And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (10)
1. a kind of scene information method for pushing based on date classification characterized by comprising
According to the date information of history day and corresponding date category building maths modec disaggregated model;
According to the date information of the date classification model and target day, the date category of target day is determined;
According to the date category of the target day, scene information corresponding with the date category of the target day is pushed to user.
2. scene information method for pushing according to claim 1, which is characterized in that the date information according to history day
With corresponding date category building maths modec disaggregated model, comprising:
It is established according to the loss function of definition to training pattern;Wherein, the loss function is for characterizing date information and date
Loss between classification;
According to the date information of history day and corresponding date category, it is trained to training pattern so that according to training to described
Loss reduction between the date information and date category of the history day that model obtains afterwards, model is described after the training
Date classification model.
3. scene information method for pushing according to claim 1 or 2, which is characterized in that the date information includes the Gregorian calendar
Date, lunar date, solar term information;The date category includes working day and day off.
4. scene information method for pushing according to claim 1, which is characterized in that the date according to the target day
Classification pushes scene information corresponding with the date category of the target day to user, comprising:
Obtain the corresponding traffic monitoring strategy of preset each date category;
According to the date type of target day, according to traffic monitoring strategy corresponding with the target day to the target daily flow into
Row monitoring, and monitoring information is pushed to user.
5. scene information method for pushing according to claim 1, which is characterized in that the date according to the target day
Classification pushes scene information corresponding with the date category of the target day to user, comprising:
The corresponding consumer consumption behavior of each date type is determined according to user's history consumption information;
According to the corresponding consumer consumption behavior of date type of the target day, corresponding merchandise news is pushed to user.
6. a kind of scene information driving means based on date classification characterized by comprising
Unit is established, for the date information and corresponding date category building maths modec disaggregated model according to history day;
Determination unit determines the date category of target day for the date information according to the date classification model and target day;
Interactive unit, for the date category according to the target day, to the date category pair of user's push and the target day
The scene information answered.
7. scene information driving means according to claim 6, which is characterized in that it is described to establish unit, it is specifically used for:
It is established according to the loss function of definition to training pattern;Wherein, the loss function is for characterizing date information and date
Loss between classification;
According to the date information of history day and corresponding date category, it is trained to training pattern so that according to training to described
Loss reduction between the date information and date category of the history day that model obtains afterwards, model is described after the training
Date classification model.
8. scene information driving means according to claim 6 or 7, which is characterized in that the date information includes the Gregorian calendar
Date, lunar date, solar term information;The date category includes working day and day off.
9. scene information driving means according to claim 6, which is characterized in that the interactive unit is specifically used for:
Obtain the corresponding traffic monitoring strategy of preset each date category;
According to the date type of target day, according to traffic monitoring strategy corresponding with the target day to the target daily flow into
Row monitoring, and monitoring information is pushed to user.
10. scene information driving means according to claim 6, which is characterized in that the interactive unit is specifically used for:
The corresponding consumer consumption behavior of each date type is determined according to user's history consumption information;
According to the corresponding consumer consumption behavior of date type of the target day, corresponding merchandise news is pushed to user.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810118470.6A CN110309399A (en) | 2018-02-06 | 2018-02-06 | Scene information method for pushing and device based on date classification |
PCT/CN2019/073855 WO2019154208A1 (en) | 2018-02-06 | 2019-01-30 | Systems and methods for determining operation strategy for service platform |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810118470.6A CN110309399A (en) | 2018-02-06 | 2018-02-06 | Scene information method for pushing and device based on date classification |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110309399A true CN110309399A (en) | 2019-10-08 |
Family
ID=67549280
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810118470.6A Pending CN110309399A (en) | 2018-02-06 | 2018-02-06 | Scene information method for pushing and device based on date classification |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN110309399A (en) |
WO (1) | WO2019154208A1 (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111382359B (en) * | 2020-03-09 | 2024-01-12 | 北京京东振世信息技术有限公司 | Service policy recommendation method and device based on reinforcement learning, and electronic equipment |
CN111931038A (en) * | 2020-06-18 | 2020-11-13 | 珠海格力电器股份有限公司 | Method, device, equipment and storage medium for determining recommended starting time |
Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2002065327A1 (en) * | 2001-02-12 | 2002-08-22 | New York University | System, process and software arrangement for providing multidimensional recommendation/suggestions |
US20130226856A1 (en) * | 2012-02-23 | 2013-08-29 | Palo Alto Research Center Incorporated | Performance-efficient system for predicting user activities based on time-related features |
CN103905978A (en) * | 2012-12-27 | 2014-07-02 | 中国电信股份有限公司 | Information pushing method based on position, pushing device and network |
CN104318322A (en) * | 2014-10-11 | 2015-01-28 | 中国南方电网有限责任公司 | Lunar calendar date based holiday load predication method |
CN105069525A (en) * | 2015-07-30 | 2015-11-18 | 广西大学 | All-weather 96-point daily load curve prediction and optimization correction system |
CN105976161A (en) * | 2016-04-29 | 2016-09-28 | 随身云(北京)信息技术有限公司 | Time axis-based intelligent recommendation calendar and user-based presentation method |
CN106383882A (en) * | 2016-09-13 | 2017-02-08 | 北京三快在线科技有限公司 | Information recommendation method and device and server |
CN106971240A (en) * | 2017-03-16 | 2017-07-21 | 河海大学 | The short-term load forecasting method that a kind of variables choice is returned with Gaussian process |
CN106980914A (en) * | 2017-06-05 | 2017-07-25 | 厦门美柚信息科技有限公司 | Forecasting Methodology and device, the terminal of determining female physiological periodicity |
CN106997577A (en) * | 2016-01-26 | 2017-08-01 | 滴滴(中国)科技有限公司 | A kind of order strategy dynamic adjusting method and device |
CN106997497A (en) * | 2017-06-13 | 2017-08-01 | 中南大学 | A kind of bank's excess reserve Forecasting Methodology based on time series and holiday information |
CN107145985A (en) * | 2017-05-09 | 2017-09-08 | 北京城建设计发展集团股份有限公司 | A kind of urban track traffic for passenger flow Regional Linking method for early warning |
CN107203579A (en) * | 2016-03-18 | 2017-09-26 | 滴滴(中国)科技有限公司 | The day off sorting technique and device for data of being called a taxi based on user |
CN107249041A (en) * | 2017-07-17 | 2017-10-13 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103533393B (en) * | 2013-09-17 | 2017-08-01 | 上海交通大学 | The family's analysis noted down based on home audience and program commending method |
US10762538B2 (en) * | 2014-04-24 | 2020-09-01 | DataSpark, PTE. LTD. | Knowledge model for personalization and location services |
CN107220899B (en) * | 2016-03-21 | 2020-06-26 | 阿里巴巴集团控股有限公司 | Social network construction method, information recommendation method, device and server |
US10147335B2 (en) * | 2016-07-15 | 2018-12-04 | Lakshmi Arthi Krishnaswami | Education data platform to support a holistic model of a learner |
-
2018
- 2018-02-06 CN CN201810118470.6A patent/CN110309399A/en active Pending
-
2019
- 2019-01-30 WO PCT/CN2019/073855 patent/WO2019154208A1/en active Application Filing
Patent Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2002065327A1 (en) * | 2001-02-12 | 2002-08-22 | New York University | System, process and software arrangement for providing multidimensional recommendation/suggestions |
US20130226856A1 (en) * | 2012-02-23 | 2013-08-29 | Palo Alto Research Center Incorporated | Performance-efficient system for predicting user activities based on time-related features |
CN103905978A (en) * | 2012-12-27 | 2014-07-02 | 中国电信股份有限公司 | Information pushing method based on position, pushing device and network |
CN104318322A (en) * | 2014-10-11 | 2015-01-28 | 中国南方电网有限责任公司 | Lunar calendar date based holiday load predication method |
CN105069525A (en) * | 2015-07-30 | 2015-11-18 | 广西大学 | All-weather 96-point daily load curve prediction and optimization correction system |
CN106997577A (en) * | 2016-01-26 | 2017-08-01 | 滴滴(中国)科技有限公司 | A kind of order strategy dynamic adjusting method and device |
CN107203579A (en) * | 2016-03-18 | 2017-09-26 | 滴滴(中国)科技有限公司 | The day off sorting technique and device for data of being called a taxi based on user |
CN105976161A (en) * | 2016-04-29 | 2016-09-28 | 随身云(北京)信息技术有限公司 | Time axis-based intelligent recommendation calendar and user-based presentation method |
CN106383882A (en) * | 2016-09-13 | 2017-02-08 | 北京三快在线科技有限公司 | Information recommendation method and device and server |
CN106971240A (en) * | 2017-03-16 | 2017-07-21 | 河海大学 | The short-term load forecasting method that a kind of variables choice is returned with Gaussian process |
CN107145985A (en) * | 2017-05-09 | 2017-09-08 | 北京城建设计发展集团股份有限公司 | A kind of urban track traffic for passenger flow Regional Linking method for early warning |
CN106980914A (en) * | 2017-06-05 | 2017-07-25 | 厦门美柚信息科技有限公司 | Forecasting Methodology and device, the terminal of determining female physiological periodicity |
CN106997497A (en) * | 2017-06-13 | 2017-08-01 | 中南大学 | A kind of bank's excess reserve Forecasting Methodology based on time series and holiday information |
CN107249041A (en) * | 2017-07-17 | 2017-10-13 | 北京京东尚科信息技术有限公司 | Information-pushing method and device |
Non-Patent Citations (2)
Title |
---|
ROYDEN S. S. FERNANDES 等: "Day ahead load forecasting models for holidays in Indian context", 《2011 INTERNATIONAL CONFERENCE ON POWER AND ENERGY SYSTEMS》 * |
张娟 等: "基于SVM的城市快速路行程时间预测研究", 《交通运输系统工程与信息》 * |
Also Published As
Publication number | Publication date |
---|---|
WO2019154208A1 (en) | 2019-08-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Dantas et al. | Air transportation demand forecast through Bagging Holt Winters methods | |
CN110782064B (en) | Visualization method and system integrating vehicle scheduling optimization and task allocation | |
CN106156115B (en) | A kind of resource regulating method and device | |
Ibrahim et al. | On the modeling and forecasting of call center arrivals | |
Kurtz et al. | Predicting demand for hydrogen station fueling | |
CN107580073A (en) | Realize the method and device of targeted customer's message push | |
Chakrabarti | The demand for reliable transit service: New evidence using stop level data from the Los Angeles Metro bus system | |
CN112215448A (en) | Method and device for distributing customer service | |
WO2021072128A1 (en) | Systems and methods for big data analytics | |
CN110309399A (en) | Scene information method for pushing and device based on date classification | |
CN110889092A (en) | Short-time large-scale activity peripheral track station passenger flow volume prediction method based on track transaction data | |
Ma et al. | Public transportation big data mining and analysis | |
CN115860800A (en) | Festival and holiday commodity sales volume prediction method and device and computer storage medium | |
CN109886737A (en) | Needing forecasting method, device, electronic equipment and readable storage medium storing program for executing | |
JP2020112847A (en) | Water demand prediction method and system | |
Cortés et al. | A simulation-based approach for fleet design in a technician dispatch problem with stochastic demand | |
US20160267399A1 (en) | Analysis device, analysis method, program, and recording medium | |
CN111353642A (en) | Telephone traffic scheduling prediction method, device, electronic equipment and storage medium | |
US11334790B1 (en) | System and method for recurrent neural networks for forecasting of consumer goods' sales and inventory | |
Rausch et al. | Beyond the beaten paths of forecasting call center arrivals: on the use of dynamic harmonic regression with predictor variables | |
Du et al. | MetroViz: Visual analysis of public transportation data | |
US20180211199A1 (en) | Decision engine for generating interfaces to simulate and optimize employee scheduling in work locations | |
Cao et al. | Forecasting long-term call traffic based on seasonal dependencies | |
CN105989022B (en) | Data rearrangement and system | |
JP2000123096A (en) | Scheduled forwarding quantity calculating method |
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 | ||
AD01 | Patent right deemed abandoned | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20220603 |