CN108446944A - A kind of the determination method, apparatus and electronic equipment in resident city - Google Patents
A kind of the determination method, apparatus and electronic equipment in resident city Download PDFInfo
- Publication number
- CN108446944A CN108446944A CN201810112757.8A CN201810112757A CN108446944A CN 108446944 A CN108446944 A CN 108446944A CN 201810112757 A CN201810112757 A CN 201810112757A CN 108446944 A CN108446944 A CN 108446944A
- Authority
- CN
- China
- Prior art keywords
- city
- user
- candidate
- probability
- resident
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
Landscapes
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides a kind of determination method, apparatus in resident city and electronic equipment, the method includes：According to the characteristic information of user, the characteristic information of each candidate city and user in the behavioural information of each candidate city, predict to obtain the Fitted probability of the user and each candidate city by city probabilistic model；According to the Fitted probability of the user and each candidate city, the first probability threshold value is determined；The resident city of the user is determined from each candidate city according to first probability threshold value.It solves and is counted in the prior art according to city classification, it is longer to expend the time, the problem of a resident city accuracy difference is determined for active user between multiple cities, behavioural information that can be by user, the characteristic information in city and user in city, the resident city for determining user, simplifies calculating process, improves accuracy.
Description
Technical field
The present embodiments relate to the determination method, apparatus in field of computer technology more particularly to a kind of resident city and
Electronic equipment.
Background technology
Resident city is that user lives throughout the year or the city of work can according to resident city to user's recommendation information, product
To effectively improve recommendation success rate.For example, the user for living or being operated in city A throughout the year, to the new of user's recommendation city A
The information such as information and ticket are heard, do not have to but recommend the information such as specialty and the tourist attractions of city A to user.
In the prior art, determine that the resident city algorithm steps of user include：First, it is counted respectively based on city specified
In history cycle, residence time of the user in the city；Wherein, the residence time can be indicated with number of days；Then, a user is existed
The residence time in each city sorts；Finally, using the most city of user's idle day as the resident city of user.
As can be seen that above process needs are classified according to city, when city numbers are larger, it is longer to expend the time；
And for user often active between multiple cities, determine that a resident city accuracy is poor.
Invention content
The present invention provides a kind of determination method, apparatus in resident city, electronic equipment, to solve to reside city in the prior art
The above problem that city determines.
According to the characteristic information of user, the characteristic information of each candidate city and user each candidate city row
For information, predict to obtain the Fitted probability of the user and each candidate city by city probabilistic model；
According to the Fitted probability of the user and each candidate city, the first probability threshold value is determined；
The resident city of the user is determined from each candidate city according to first probability threshold value.
According to the second aspect of the invention, a kind of determining device in resident city is provided, described device includes：
Probabilistic forecasting module is used for according to the characteristic information of user, the characteristic information of each candidate city and user in institute
The behavioural information for stating each candidate city, the fitting for predicting to obtain the user and each candidate city by city probabilistic model are general
Rate；
First probability threshold value determining module determines first for the Fitted probability according to the user and each candidate city
Probability threshold value；
Resident city determining module, for determining the use from each candidate city according to first probability threshold value
The resident city at family.
According to the third aspect of the invention we, a kind of electronic equipment is provided, including：
Processor, memory and it is stored in the computer journey that can be run on the memory and on the processor
Sequence, which is characterized in that the processor realizes the determination method in aforementioned resident city when executing described program.
According to the fourth aspect of the invention, a kind of readable storage medium storing program for executing is provided, which is characterized in that when the storage medium
In instruction by electronic equipment processor execute when so that electronic equipment is able to carry out the determination method in aforementioned resident city.
An embodiment of the present invention provides a kind of determination method, apparatus in resident city, electronic equipment, the method includes：
According to the characteristic information of user, the characteristic information of each candidate city and user in the behavioural information of each candidate city, lead to
City probabilistic model is crossed to predict to obtain the Fitted probability of the user and each candidate city；According to the user and each candidate city
Fitted probability, determine the first probability threshold value；The use is determined from each candidate city according to first probability threshold value
The resident city at family.It solves and is counted in the prior art according to city classification, calculating process is complicated, between multiple cities
Active user determines the problem of a resident city accuracy difference, can pass through user, the characteristic information in city and user
Behavioural information in city determines the resident city of user, simplifies calculating process, improves accuracy.
Description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below by institute in the description to the embodiment of the present invention
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the present invention
Example, for those of ordinary skill in the art, without having to pay creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is a kind of determination method specific steps flow in resident city under system architecture provided in an embodiment of the present invention
Figure；
Fig. 2 is the determination method specific steps stream in the resident city of another kind under system architecture provided in an embodiment of the present invention
Cheng Tu；
Fig. 3 is a kind of structure chart of the determining device in resident city provided in an embodiment of the present invention；
Fig. 4 is the structure chart of the determining device in another resident city provided in an embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are some of the embodiments of the present invention, instead of all the embodiments.Based on this hair
Embodiment in bright, the every other implementation that those of ordinary skill in the art are obtained without creative efforts
Example, shall fall within the protection scope of the present invention.
Embodiment one
Referring to Fig.1, it illustrates a kind of step flow charts of the determination method in resident city, including：
Step 101, according to the characteristic information of user, the characteristic information of each candidate city and user in each candidate
The behavioural information in city is predicted to obtain the Fitted probability of the user and each candidate city by city probabilistic model.
Wherein, the characteristic information of user includes but not limited to：Gender, the age, occupation, the level of consumption, income level, whether
Travel intelligent.
The characteristic information of candidate city includes but not limited to：City level, whether tourist city, city local user's number
The average daily order numbers for the business such as traffic of travelling with strange land number of users, hotel.
User includes but not limited in the behavioural information of each candidate city：User browsed within the specified historical time period and
It positions the number of the candidate city and whether maximum time window that accounting, user occur in the candidate city, the candidate city
The local city of user.It is appreciated that the specified historical time period can be past half a year, one month or one week etc., the present invention
Embodiment does not limit the historical time period.
As can be seen that above- mentioned information includes continuous with discrete two kinds of information, wherein continuous information includes：Age, city
Local user's number and strange land number of users, the average daily order numbers of hotel's travelling business such as traffic, user it is all in specified historical time
Browsing and the number of the candidate city is positioned and maximum time window that accounting, user occur in the candidate city in phase, it is discrete
Information includes：Gender, occupation, the level of consumption, income level, the intelligent that whether travels, city level, whether tourist city, the time
Select city whether the local city of user.
In practical applications, the discrete features in above- mentioned information can be indicated with discrete values.For example, for the property of user
Other information：Men's 1 indicates, lady's 2 indicate, the income level of user：Low income is indicated with 1, medium income is indicated with 2, high
Etc. incomes with 3 mark etc..
It is appreciated that for a user, a candidate city, each characteristic information of the user, the candidate city it is each
Characteristic information and the user respectively correspond to a variable and are input to city probability in each behavioural information of the candidate city
Model obtains the Fitted probability of the user and the candidate city, hence for multiple users, multiple candidate cities, obtains each use
The Fitted probability at family and each candidate city.
Step 102, according to the Fitted probability of the user and each candidate city, the first probability threshold value is determined.
The embodiment of the present invention is by two adjacent probability values of probability value disparity, as the ginseng for determining the first probability threshold value
Examine value.
Specifically, can be using the average value of two adjacent probability values as the first probability threshold value, it can also be adjacent by two
Other weighted averages of probability threshold value are as the first probability threshold value.
Step 103, the resident city of the user is determined from each candidate city according to first probability threshold value.
Specifically, for the candidate city using probability more than the first probability threshold value as resident city, probability is less than the first probability
The candidate city of threshold value is as non-resident city.To resident city, there may be multiple.
The embodiment of the present invention determines that multiple resident cities are more in line with practical application, in practical applications, small part user
It can be active between multiple cities.For example, a user frequently goes on business or one in big cities such as Beijing and Shanghai Guangzhou back and forth for a long time
A user lives in the Yanjiao town in city of Langfang in Hebei Province, but goes to work in Beijing City.
Determining that user after resident city, can carry out personalized recommendation, to improve recommendation success rate.For example, certain
User is when resident city accesses U.S. group's platform, the hot item that local user can be recommended to buy, and local hot spot, spy
Production etc. is without recommending；When certain user is in the U.S. group's platform of non-resident city access, the fast sale of strange land user purchase can be recommended
Commodity, strange land travelling products, resident city to the train ticket air ticket etc. between strange land.
In conclusion an embodiment of the present invention provides a kind of determination method in resident city, the method includes：According to
The characteristic information at family, the characteristic information of each candidate city and user pass through city in the behavioural information of each candidate city
Probabilistic model is predicted to obtain the Fitted probability of the user and each candidate city；According to the fitting of the user and each candidate city
Probability determines the first probability threshold value；Determine that the user's is normal from each candidate city according to first probability threshold value
In city.It solves and is counted in the prior art according to city classification, the consuming time is longer, for active between multiple cities
User determines the problem of a resident city accuracy difference, can be by user, the characteristic information in city and user in city
Behavioural information, determine the resident city of user, simplify calculating process, improve accuracy.
Embodiment two
The embodiment of the present application is described the determination method for optionally residing city from the level of system architecture.
With reference to Fig. 2, it illustrates the specific steps flow charts of the determination method in another resident city.
Step 201, train to obtain city probabilistic model based on labeled data sample set, in the labeled data sample set
Each labeled data sample includes at least：The characteristic information of user, the characteristic information of candidate city and user are in candidate city
Behavioural information.
Wherein, every data sample standard deviation of labeled data sample set has been marked to whether Yingcheng City is the normal of corresponding user
In city.In practical applications, a field can be increased to indicate that data sample is resident city sample or non-resident city
City's sample.For example, sample can be marked by field ResidentCity, when ResidentCity=1 indicates that sample is resident
City sample；When ResidentCity=0 indicates that sample is non-resident city sample.
Labeled data sample set can be by contact staff to similar platform users such as hotel occupancy client, U.S. group's platforms
Call-on back by phone, questionnaire acquire.
The characteristic information of user can be obtained by user to being parsed in the usage log of application.For example, user rolls into a ball in U.S.
When platform register account number, the personal information of input includes：Gender, age, occupation, the level of consumption, income level can be used as and use
The characteristic information at family, further, it is also possible to which analysis obtains whether user is tourism intelligent from the usage log of user.When user is closed
When the information of note is mostly the information such as tourist attractions, hotel, it may be determined that user is tourism intelligent；Otherwise, user is not that tourism reaches
People.It is appreciated that in practical applications, can also be provided in registration and whether like travel option for user's selection；To select
It selects and likes the user of tourism for tourism intelligent；Otherwise the user for not liking tourism is not tourism intelligent.
The characteristic information of candidate city can be obtained from database.For example, U.S. group's platform can do city on one basis
Data.
User candidate city behavioural information can by user to the access log of application in parse and obtain, for example,
It can determine that user browses and position number and accounting, the user of the candidate city within the specified historical time period by daily record
The candidate city occur maximum time window and the candidate city whether the local city of user.
In embodiments of the present invention, Logic Regression Models may be used or decision-tree model is trained, so that
Parameter group in model is optimized parameter group.
Optionally, in another embodiment of the invention, step 201 includes sub-step 2011 to 2014：
Sub-step 2011 initializes the parameter group of city probabilistic model.
After choosing probabilistic model, the parameter group of initialization probability model.For example, for Logic Regression Models, model
Formula is as follows：
Wherein, u is user, and c is candidate city；
N is the number of non-constant parameter, is existed according to the characteristic information of user, the characteristic information of candidate city and user
The number of the behavioural information of candidate city determines.For example, if the characteristic information of user includes gender, age, occupation, consumption water
It is flat, whether income level, travel six kinds of intelligent etc., the characteristic information in city include city level, whether tourist city, city
Five kinds of average daily order numbers of business such as local user's number and strange land number of users, hotel's travelling traffic etc., user is in candidate city
Behavioural information includes user browse and position within the specified historical time period candidate city number and accounting, user at this
The maximum time window and the candidate city that candidate city occurs whether three kinds of the local city of user etc., to parameter group
Size be 14.
xiBelieve in the behavior of candidate city c for the characteristic information of user u, the characteristic information of candidate city c and user u
Cease the value of the ith feature information in the characteristic information of composition；
wiFor xiParameter, b is constant parameter.
It is appreciated that training is just to determine wiWith the value of b.
Specifically, can initiation parameter group based on experience value, other values can also be initialized as by analysis.It can manage
Solution, when the value for initializing parameter group is inappropriate, it will increase the training time；When the value of initialization parameter group is close to optimal ginseng
When array, it will reduce the training time.
Sub-step 2012, for each labeled data sample in the labeled data sample set, by the characteristic information of user,
The characteristic information of candidate city and user are input to default city probabilistic model in the behavioural information of candidate city, are used
The Fitted probability at family and corresponding candidate city.
Specifically, for user u and candidate city c, by the characteristic information of user u, the characteristic information of candidate city c, use
Family u is sequentially input into probabilistic model according to specified in the behavioural information of candidate city c, user u and candidate city is calculated
The Fitted probability of c.
It is appreciated that the behavior of the characteristic information of user u, the characteristic information of candidate city c, user u in candidate city c is believed
The sequence of breath can be set according to practical application scene, and the embodiment of the present invention does not limit it.
In practical applications, labeled data sample set includes the corresponding multiple candidate cities of a large number of users, for every
A labeled data sample, one of a corresponding user and the user correspondence candidate city.So as to according to mark
Set of data samples obtains Fitted probability of a large number of users respectively with multiple candidate cities.
Sub-step 2013, the user and the Fitted probability of corresponding candidate city determined by each labeled data sample are true
Determine penalty values.
The loss function of the embodiment of the present invention resides the probability interval in city and non-resident city by maximization, with tradition
Loss function in model can not only intend user and city if the cross entropy loss function of Logic Regression Models is compared
It closes probability to be ranked up, it is also possible that user is more apparent than with the Fitted probability gap in non-resident city with resident city.
Optionally, in another embodiment of the invention, step 2013 includes sub-step 20131 to 20132：
Sub-step 20131, for each user in labeled data sample set, respectively by the quasi- of user and each non-resident city
The Fitted probability that probability subtracts user and each resident city is closed, preset protection value is added, obtain the user each first is poor
Value.
In embodiments of the present invention, for all users in labeled data sample set, all marks based on each user
Data sample is noted, according to the Fitted probability in the user and each non-resident city, the Fitted probability with each resident city, calculates the use
Each first difference at family.
Specifically, for user u, resident city c ', non-resident city c, the calculation formula of the first difference is as follows：
M1=f (φ (u, c))-f (φ (u, c'))+ε (2)
Wherein, f (φ (u, c)) is the Fitted probability of user u and non-resident city c, and f (φ (u, c')) is for user u and often
The Fitted probability of c in city '.In practical applications, the Logic Regression Models formula as shown in formula (1) may be used, also may be used
To use other model formations.
ε is preset protection value, can influence training result, can be according to being adjusted in the training process according to prediction result
Whole, the embodiment of the present invention does not limit it.
Sub-step 20132 takes each first difference of the user to be worth to second with zero maximum each user
Difference, and the summation of each second difference is counted, obtain the third difference of the user.
Specifically, for user u, resident city c ', non-resident city c, the calculation formula of the second difference is as follows：
M2=max (0, M1)=max (0, f (φ (u, c))-f (φ (u, c'))+ε) (3)
Then, the calculation formula of third difference is as follows：
Wherein,CuFor the candidate city set of user u, it is divided into resident city and non-resident city.It is appreciated that here often
It all can be multiple in city and non-resident city.
Sub-step 20133 counts the summation of the third difference of each user, obtains penalty values.
Specifically, the calculation formula of penalty values is as follows：
Wherein, U gathers for all users.
Sub-step 2014 adjusts the parameter group if the penalty values are unsatisfactory for preset condition, until the penalty values
Meet preset condition.
Specifically, when penalty values are less than or equal to preset value, penalty values meet preset condition, and training terminates, corresponds at this time
Parameter group be target component group；When penalty values are more than preset value, penalty values are unsatisfactory for preset condition, adjusting parameter group, with
Continue to train, until penalty values meet preset condition.
It is appreciated that preset value can be set according to practical application scene, the embodiment of the present invention does not limit it.In advance
If value is smaller, training result is more accurate, and the training time is longer；Preset value is bigger, and training result is more coarse, and the training time is shorter.
Step 202, according to the characteristic information of user, the characteristic information of each candidate city and user in each candidate
The behavioural information in city is predicted to obtain the Fitted probability of the user and each candidate city by city probabilistic model.
The step is referred to the detailed description of step 101, and details are not described herein.
Step 203, the Fitted probability of the user and each candidate city is sorted.
Specifically, can be arranged in decreasing order can also ascending order arrangement.
Step 204, the difference between two adjacent probability is calculated separately.
Specifically, descending is arranged, the latter probability is subtracted with previous probability；Ascending order is arranged, the latter is used
Probability subtracts previous probability.
For in ascending order, for i-th of probability and i+1 probability, difference MiIt can be calculated according to following formula：
Mi=Pi-Pi+1 (6)
Wherein, PiFor i-th of probability, Pi+1For i+1 probability.
It is appreciated that in practical applications, difference can also be taken absolute value, to ensure obtained difference as positive value.
Step 205, the weighted average of the maximum two adjacent probability of calculating difference, obtains the first probability threshold value.
Based on formula (6), judgement obtains the maximum two adjacent probability Ps of differenceIAnd PI+1, then the first probability threshold value PsIt can be with
It is calculated according to following formula：
Ps=C1·PI+C2·PI+1 (7)
Wherein, C1For probability PIWeighting parameters, C2For probability PI+1Weighting parameters, C1+C2=1, C1≠0,C2≠0.It is special
Not, work as C1=C2When=0.5, the first probability threshold value is PIAnd PI+1Average value.
Step 206, for each candidate city, if the Fitted probability of user and the candidate city is more than or equal to described first
Probability threshold value, then the candidate city is the resident city of user.
It is appreciated that the first probability threshold value that formula (7) is calculated, probability PIAnd come PIThe probability pair of front
The candidate city answered is the resident city of user；Probability PI+1And come PI+1The corresponding candidate city of subsequent probability is user's
Non-resident city.
In conclusion an embodiment of the present invention provides a kind of determination method in resident city, the method includes：According to
The characteristic information at family, the characteristic information of each candidate city and user pass through city in the behavioural information of each candidate city
Probabilistic model is predicted to obtain the Fitted probability of the user and each candidate city；According to the fitting of the user and each candidate city
Probability determines the first probability threshold value；Determine that the user's is normal from each candidate city according to first probability threshold value
In city.It solves and is counted in the prior art according to city classification, the consuming time is longer, for active between multiple cities
User determines the problem of a resident city accuracy difference, can be by user, the characteristic information in city and user in city
Behavioural information, determine the resident city of user, simplify calculating process, improve accuracy.
Embodiment three
It is specific as follows it illustrates a kind of structure chart of the determining device in resident city with reference to Fig. 3.
Probabilistic forecasting module 301, for according to the characteristic information of user, the characteristic information of each candidate city and user
In the behavioural information of each candidate city, predict to obtain the fitting of the user and each candidate city by city probabilistic model
Probability.
First probability threshold value determining module 302, for according to the Fitted probability of the user and each candidate city, determining the
One probability threshold value.
Resident city determining module 303, for determining institute from each candidate city according to first probability threshold value
State the resident city of user.
In conclusion an embodiment of the present invention provides a kind of determining device in resident city, described device includes：Probability is pre-
Module is surveyed, is used for according to the characteristic information of user, the characteristic information of each candidate city and user in each candidate city
Behavioural information is predicted to obtain the Fitted probability of the user and each candidate city by city probabilistic model；First probability threshold value
Determining module determines the first probability threshold value for the Fitted probability according to the user and each candidate city；Resident city determines
Module, the resident city for determining the user from each candidate city according to first probability threshold value.It solves
It is counted in the prior art according to city classification, the consuming time is longer, and one is determined for active user between multiple cities
The problem of resident city accuracy difference, behavioural information that can be by user, the characteristic information in city and user in city,
The resident city for determining user, simplifies calculating process, improves accuracy.
Example IV
It is specific as follows it illustrates the structure chart of the determining device in another resident city with reference to Fig. 4.
Probabilistic model training module 401 trains to obtain city probabilistic model, the mark for being based on labeled data sample set
Each labeled data sample that note data sample is concentrated includes at least：The characteristic information of user, the characteristic information of candidate city and
Behavioural information of the user in candidate city.
Probabilistic forecasting module 402, for according to the characteristic information of user, the characteristic information of each candidate city and user
In the behavioural information of each candidate city, predict to obtain the fitting of the user and each candidate city by city probabilistic model
Probability.
First probability threshold value determining module 403, for according to the Fitted probability of the user and each candidate city, determining the
One probability threshold value.Optionally, in another embodiment of the invention, above-mentioned first probability threshold value determining module 403, including：
Sorting sub-module 4031, for the Fitted probability of the user and each candidate city to sort.
Probability difference computational submodule 4032, for calculating separately the difference between two adjacent probability.
First probability threshold value determination sub-module 4033 is used for the weighted average of the maximum two adjacent probability of calculating difference
Value, obtains the first probability threshold value.
Resident city determining module 404, for determining institute from each candidate city according to first probability threshold value
State the resident city of user.Optionally, in another embodiment of the invention, above-mentioned resident city determining module, including：
Resident city determination sub-module 4041, is used for for each candidate city, if the fitting of user and the candidate city
Probability is more than or equal to first probability threshold value, then the candidate city is the resident city of user.
Optionally, in another embodiment of the invention, above-mentioned probabilistic model training module 401, including：
Parameter group initialization submodule, the parameter group for initializing city probabilistic model.
Probability calculation submodule is used for for each labeled data sample in the labeled data sample set, by user's
Characteristic information, the characteristic information of candidate city and user are input to default city probability mould in the behavioural information of candidate city
Type obtains the Fitted probability of user and corresponding candidate city.
Penalty values determination sub-module, the user for being determined by each labeled data sample and corresponding candidate city
Fitted probability determines penalty values.
Continue that submodule is trained to adjust the parameter group if being unsatisfactory for preset condition for the penalty values, until institute
It states penalty values and meets preset condition.
Optionally, in another embodiment of the invention, above-mentioned penalty values determination sub-module, including：
First difference computational unit, for for each user in labeled data sample set, respectively by user and it is each very
Fitted probability in city subtracts the Fitted probability of user and each resident city, adds preset protection value, obtains the user
Each first difference.
Second difference computational unit, for for each user, each first difference of the user being taken to be obtained with zero maximum value
To the second difference, and the summation of each second difference is counted, obtains the third difference of the user.
Penalty values determination unit, the summation of the third difference for counting each user, obtains penalty values.
In conclusion an embodiment of the present invention provides a kind of determining device in resident city, described device includes：Probability is pre-
Module is surveyed, is used for according to the characteristic information of user, the characteristic information of each candidate city and user in each candidate city
Behavioural information is predicted to obtain the Fitted probability of the user and each candidate city by city probabilistic model；First probability threshold value
Determining module determines the first probability threshold value for the Fitted probability according to the user and each candidate city；Resident city determines
Module, the resident city for determining the user from each candidate city according to first probability threshold value.It solves
It is counted in the prior art according to city classification, the consuming time is longer, and one is determined for active user between multiple cities
The problem of resident city accuracy difference, behavioural information that can be by user, the characteristic information in city and user in city,
The resident city for determining user, simplifies calculating process, improves accuracy.
The embodiment of the present invention additionally provides a kind of electronic equipment, including：Processor, memory and it is stored in the storage
On device and the computer program that can run on the processor, which is characterized in that the processor executes real when described program
The determination method in the resident city of existing previous embodiment.
The embodiment of the present invention additionally provides a kind of readable storage medium storing program for executing, when the instruction in the storage medium is by electronic equipment
Processor execute when so that electronic equipment is able to carry out the determination method in the resident city of previous embodiment.
For device embodiments, since it is basically similar to the method embodiment, so fairly simple, the correlation of description
Place illustrates referring to the part of embodiment of the method.
Algorithm and display be not inherently related to any certain computer, virtual system or miscellaneous equipment provided herein.
Various general-purpose systems can also be used together with teaching based on this.As described above, it constructs required by this kind of system
Structure be obvious.In addition, the present invention is not also directed to any certain programmed language.It should be understood that can utilize various
Programming language realizes the content of invention described herein, and the description done above to language-specific is to disclose this hair
Bright preferred forms.
In the instructions provided here, numerous specific details are set forth.It is to be appreciated, however, that the implementation of the present invention
Example can be put into practice without these specific details.In some instances, well known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this description.
Similarly, it should be understood that in order to simplify the disclosure and help to understand one or more of each inventive aspect,
Above in the description of exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes
In example, figure or descriptions thereof.However, the method for the disclosure should be construed to reflect following intention：It is i.e. required to protect
Shield the present invention claims the more features of feature than being expressly recited in each claim.More precisely, as following
Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore,
Thus the claims for following specific implementation mode are expressly incorporated in the specific implementation mode, wherein each claim itself
All as a separate embodiment of the present invention.
Those skilled in the art, which are appreciated that, to carry out adaptively the module in the equipment in embodiment
Change and they are arranged in the one or more equipment different from the embodiment.It can be the module or list in embodiment
Member or component be combined into a module or unit or component, and can be divided into addition multiple submodule or subelement or
Sub-component.Other than such feature and/or at least some of process or unit exclude each other, it may be used any
Combination is disclosed to all features disclosed in this specification (including adjoint claim, abstract and attached drawing) and so to appoint
Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification (including adjoint power
Profit requires, abstract and attached drawing) disclosed in each feature can be by providing the alternative features of identical, equivalent or similar purpose come generation
It replaces.
The all parts embodiment of the present invention can be with hardware realization, or to run on one or more processors
Software module realize, or realized with combination thereof.It will be understood by those of skill in the art that can use in practice
Microprocessor or digital signal processor (DSP) realize resident city according to the ... of the embodiment of the present invention really in locking equipment
The some or all functions of some or all components.The present invention is also implemented as executing method as described herein
Some or all equipment or program of device.The program of such realization present invention can be stored in computer-readable
On medium, or can be with the form of one or more signal.Such signal can be downloaded from internet website
It arrives, either provided on carrier signal or provides in any other forms.
It should be noted that the present invention will be described rather than limits the invention for above-described embodiment, and ability
Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference mark between bracket should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" before element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of several different elements and being come by means of properly programmed computer real
It is existing.In the unit claims listing several devices, several in these devices can be by the same hardware branch
To embody.The use of word first, second, and third does not indicate that any sequence.These words can be explained and be run after fame
Claim.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
All any modification, equivalent and improvement etc., should all be included in the protection scope of the present invention made by within refreshing and principle.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, protection scope of the present invention should be subject to the protection scope in claims.
Claims (10)
1. a kind of determination method in resident city, which is characterized in that the method includes：
According to the characteristic information of target user, the characteristic information of each candidate city and the target user in each candidate
The behavioural information in city is predicted to obtain the Fitted probability of the target user and each candidate city by city probabilistic model；
According to the Fitted probability of the target user and each candidate city, the first probability threshold value is determined；
The resident city of the target user is determined from each candidate city according to first probability threshold value.
2. according to the method described in claim 1, it is characterized in that, it is described according to first probability threshold value from each candidate
The step of resident city of the target user is determined in city, including：
For each candidate city, if the Fitted probability of the target user and the candidate city is more than or equal to first probability
Threshold value, then the candidate city is the resident city of the target user.
3. according to the method described in claim 1, it is characterized in that, described according to the quasi- of the target user and each candidate city
The step of closing probability, determining the first probability threshold value, including：
The Fitted probability of the target user and each candidate city is sorted；
Calculate separately the difference between two adjacent probability；
The weighted average of the maximum two adjacent probability of calculating difference, obtains the first probability threshold value.
4. according to the method described in claim 1, it is characterized in that, the method further includes：
It trains to obtain city probabilistic model based on labeled data sample set, each labeled data sample in the labeled data sample set
Originally it includes at least：The behavioural information of the characteristic information of user, the characteristic information of candidate city and user in candidate city.
5. according to the method described in claim 4, it is characterized in that, it is described train to obtain city based on labeled data sample set it is general
The step of rate model, including：
Initialize the parameter group of city probabilistic model；
For each labeled data sample in the labeled data sample set, by the feature of the characteristic information of user, candidate city
Information and user are input to default city probabilistic model in the behavioural information of candidate city, obtain user and corresponding candidate city
The Fitted probability in city；
The user determined by each labeled data sample determines penalty values with the Fitted probability of corresponding candidate city；
If the penalty values are unsatisfactory for preset condition, the parameter group is adjusted, until the penalty values meet preset condition.
6. according to the method described in claim 5, it is characterized in that, the user determined by each labeled data sample
The step of penalty values being determined with the Fitted probability of corresponding candidate city, including：
For each user in labeled data sample set, respectively by the Fitted probability of user and each non-resident city subtract user with
The Fitted probability in each resident city adds preset protection value, obtains each first difference of the user；
For each user, each first difference of the user is taken to be worth to the second difference with zero maximum, and it is poor to count each second
The summation of value obtains the third difference of the user；
The summation for counting the third difference of each user, obtains penalty values.
7. a kind of determining device in resident city, which is characterized in that described device includes：
Probabilistic forecasting module is used for according to the characteristic information of user, the characteristic information of each candidate city and user described each
The behavioural information of candidate city is predicted to obtain the Fitted probability of the user and each candidate city by city probabilistic model；
First probability threshold value determining module determines the first probability for the Fitted probability according to the user and each candidate city
Threshold value；
Resident city determining module, for determining the user's from each candidate city according to first probability threshold value
Resident city.
8. the apparatus according to claim 1, which is characterized in that the resident city determining module, including：
Resident city determination sub-module, is used for for each candidate city, if the Fitted probability of user and the candidate city is more than
Equal to first probability threshold value, then the candidate city is the resident city of user.
9. a kind of electronic equipment, which is characterized in that including：
Processor, memory and it is stored in the computer program that can be run on the memory and on the processor,
It is characterized in that, the processor realizes the resident city as described in one or more in claim 1-6 when executing described program
Determination method.
10. a kind of readable storage medium storing program for executing, which is characterized in that when the instruction in the storage medium is held by the processor of electronic equipment
When row so that electronic equipment is able to carry out the determination side in the resident city as described in one or more in claim to a method 1-6
Method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810112757.8A CN108446944B (en) | 2018-02-05 | 2018-02-05 | Resident city determination method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810112757.8A CN108446944B (en) | 2018-02-05 | 2018-02-05 | Resident city determination method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108446944A true CN108446944A (en) | 2018-08-24 |
| CN108446944B CN108446944B (en) | 2020-03-17 |
Family
ID=63191751
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810112757.8A Active CN108446944B (en) | 2018-02-05 | 2018-02-05 | Resident city determination method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108446944B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109195096A (en) * | 2018-10-15 | 2019-01-11 | 北京创鑫旅程网络技术有限公司 | Determine the method and device of travelling state |
| CN109242580A (en) * | 2018-11-28 | 2019-01-18 | 北京腾云天下科技有限公司 | Determine the method that market is entered in recommendation of the target brand in given city |
| CN110177152A (en) * | 2019-06-11 | 2019-08-27 | 秒针信息技术有限公司 | A kind of message push method and device |
| CN110956351A (en) * | 2018-09-27 | 2020-04-03 | 北京嘀嘀无限科技发展有限公司 | Order allocation method and device, server and computer readable storage medium |
| CN111241225A (en) * | 2020-01-10 | 2020-06-05 | 北京百度网讯科技有限公司 | Resident area change judgment method, resident area change judgment device, resident area change judgment equipment and storage medium |
| CN111311193A (en) * | 2020-02-26 | 2020-06-19 | 百度在线网络技术（北京）有限公司 | Configuration method and device of public service resources |
| CN111935646A (en) * | 2020-07-22 | 2020-11-13 | 北京明略昭辉科技有限公司 | Method and system for estimating common address of mobile equipment user |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150074021A1 (en) * | 2013-09-12 | 2015-03-12 | International Business Machines Corporation | Generating a training model based on feedback |
| CN105142104A (en) * | 2015-06-19 | 2015-12-09 | 北京奇虎科技有限公司 | Method, device and system for providing recommendation information |
| CN105740401A (en) * | 2016-01-28 | 2016-07-06 | 北京理工大学 | Individual behavior and group interest-based interest place recommendation method and device |
| 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 |
| US20170364933A1 (en) * | 2014-12-09 | 2017-12-21 | Beijing Didi Infinity Technology And Development Co., Ltd. | User maintenance system and method |
-
2018
- 2018-02-05 CN CN201810112757.8A patent/CN108446944B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150074021A1 (en) * | 2013-09-12 | 2015-03-12 | International Business Machines Corporation | Generating a training model based on feedback |
| US20170364933A1 (en) * | 2014-12-09 | 2017-12-21 | Beijing Didi Infinity Technology And Development Co., Ltd. | User maintenance system and method |
| CN105142104A (en) * | 2015-06-19 | 2015-12-09 | 北京奇虎科技有限公司 | Method, device and system for providing recommendation information |
| CN105740401A (en) * | 2016-01-28 | 2016-07-06 | 北京理工大学 | Individual behavior and group interest-based interest place recommendation method and device |
| 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 |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110956351A (en) * | 2018-09-27 | 2020-04-03 | 北京嘀嘀无限科技发展有限公司 | Order allocation method and device, server and computer readable storage medium |
| CN110956351B (en) * | 2018-09-27 | 2022-05-17 | 北京嘀嘀无限科技发展有限公司 | Order allocation method and device, server and computer readable storage medium |
| CN109195096A (en) * | 2018-10-15 | 2019-01-11 | 北京创鑫旅程网络技术有限公司 | Determine the method and device of travelling state |
| CN109242580A (en) * | 2018-11-28 | 2019-01-18 | 北京腾云天下科技有限公司 | Determine the method that market is entered in recommendation of the target brand in given city |
| CN109242580B (en) * | 2018-11-28 | 2020-12-29 | 北京腾云天下科技有限公司 | Method for determining recommended emporium of target brand in specified city |
| CN110177152A (en) * | 2019-06-11 | 2019-08-27 | 秒针信息技术有限公司 | A kind of message push method and device |
| CN110177152B (en) * | 2019-06-11 | 2022-03-29 | 秒针信息技术有限公司 | Information pushing method and device |
| CN111241225A (en) * | 2020-01-10 | 2020-06-05 | 北京百度网讯科技有限公司 | Resident area change judgment method, resident area change judgment device, resident area change judgment equipment and storage medium |
| CN111241225B (en) * | 2020-01-10 | 2023-08-08 | 北京百度网讯科技有限公司 | Method, device, equipment and storage medium for judging change of resident area |
| CN111311193A (en) * | 2020-02-26 | 2020-06-19 | 百度在线网络技术（北京）有限公司 | Configuration method and device of public service resources |
| CN111311193B (en) * | 2020-02-26 | 2023-09-22 | 百度在线网络技术（北京）有限公司 | Method and device for configuring public service resources |
| CN111935646A (en) * | 2020-07-22 | 2020-11-13 | 北京明略昭辉科技有限公司 | Method and system for estimating common address of mobile equipment user |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108446944B (en) | 2020-03-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108446944A (en) | A kind of the determination method, apparatus and electronic equipment in resident city | |
| Zhang et al. | Can consumer-posted photos serve as a leading indicator of restaurant survival? Evidence from Yelp | |
| KR101871747B1 (en) | Similarity tendency based user-sightseeing recommendation system and method thereof | |
| CN109993652B (en) | Loan credit risk assessment method and device | |
| Batsell et al. | A model and measurement methodology for predicting individual consumer choice | |
| CN110532479A (en) | A kind of information recommendation method, device and equipment | |
| CN109934619A (en) | User's portrait tag modeling method, apparatus, electronic equipment and readable storage medium storing program for executing | |
| CN108335137A (en) | Sort method and device, electronic equipment, computer-readable medium | |
| CN108550068A (en) | A kind of individual commodity recommendation method and system based on user behavior analysis | |
| Moital et al. | A cross national study of golf tourists' satisfaction | |
| CN107146089A (en) | The single recognition methods of one kind brush and device, electronic equipment | |
| Eckhardt | Similarity of users’(content-based) preference models for Collaborative filtering in few ratings scenario | |
| CN109522372A (en) | The prediction technique of civil aviaton field passenger value | |
| CN106779808A (en) | Consumer space's behavior analysis system and method in a kind of commercial circle | |
| CN107909392A (en) | Stimulation behavior susceptibility Forecasting Methodology, device, equipment and readable storage medium storing program for executing | |
| CN110880127B (en) | Consumption level prediction method and device, electronic equipment and storage medium | |
| Mano et al. | A conceptual model of the antecedents and consequences of tourist destination image | |
| CN112288554B (en) | Commodity recommendation method and device, storage medium and electronic device | |
| Zhang et al. | Forecasting hotel room demand amid COVID-19 | |
| Gu et al. | Consumer online search with partially revealed information | |
| CN110990717B (en) | Interest point recommendation method based on cross-domain association | |
| CN115952355A (en) | Next interest point recommendation system based on space-time power law attention | |
| Ibrahim et al. | Recognising intrinsic motivation using smartphone trajectories | |
| Lee et al. | A generalized Bass model for predicting the sales patterns of motion pictures having seasonality and herd behavior | |
| CN116762070A (en) | System, method, computing platform, and storage medium for providing image recommendations |
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 |