CN107194412A - A kind of method of processing data, device, equipment and computer-readable storage medium - Google Patents
A kind of method of processing data, device, equipment and computer-readable storage medium Download PDFInfo
- Publication number
- CN107194412A CN107194412A CN201710260035.2A CN201710260035A CN107194412A CN 107194412 A CN107194412 A CN 107194412A CN 201710260035 A CN201710260035 A CN 201710260035A CN 107194412 A CN107194412 A CN 107194412A
- Authority
- CN
- China
- Prior art keywords
- label
- characteristic data
- user characteristic
- user
- tags
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a kind of method of processing data, device, equipment and computer-readable storage medium, the method for wherein processing data includes：Obtain the user characteristic data of the relevant information containing customer location；According to division of the customer location relevant information on geographical position, the set of tags corresponding to user characteristic data is determined；The user characteristic data is predicted using the set of tags corresponding classification submodel, the label of user is obtained；Wherein, each set of tags corresponds to a classification submodel respectively.The present invention is by obtaining with after the set of tags corresponding to user characteristic data, being predicted using with the classification submodel corresponding to set of tags to user characteristic data, so as to improve predetermined speed and prediction accuracy to user characteristic data.
Description
【Technical field】
The present invention relates to Map Services technical field, more particularly to a kind of method of processing data, device, equipment and calculating
Machine storage medium.
【Background technology】
Existing disaggregated model, algorithmically there is SVM (support vector machine, SVMs), LR
The multiple choices such as (logistic regression, logistic regression), decision tree, but depended on when realizing training and predicting
Single model covers all tag along sort and feature.Therefore, prior art works as classification when carrying out the training of disaggregated model
When larger, the involved tag along sort of problem and many features, then the scale of required training data also can be in disastrously
Increase, so as to cause disaggregated model in finite time it is difficult to complete training, and then influence the ageing of model application and change
For development efficiency.In addition, prior art is when carrying out the prediction of disaggregated model, when the tag along sort in single model and feature rule
When mould is larger, predetermined speed of model can be influenceed, if in the scene of real-time estimate is carried out using disaggregated model, the prediction of model
Speed can influence the response speed of real-time system.
【The content of the invention】
In view of this, the invention provides a kind of method of processing data, device, equipment and computer-readable storage medium, energy
Enough lift the predetermined speed and prediction accuracy of disaggregated model to user data.
The present invention there is provided a kind of method of processing data, the side to solve the technical scheme that technical problem is used
Method includes：Obtain the user characteristic data of the relevant information containing customer location；According to the customer location relevant information in geographical position
The division put, determines the set of tags corresponding to user characteristic data；Using the corresponding classification submodel of the set of tags to institute
State user characteristic data to be predicted, obtain the label of user；Wherein, each set of tags corresponds to a classification submodel respectively.
According to one preferred embodiment of the present invention, the classification submodel is previously obtained using following training method：Obtain
Take label and the user characteristic data associated with label；According to division of each label on geographical position, to the label
It is grouped；Each set of tags is included into label and the user characteristic data associated with label is as training data, point
The corresponding classification submodel of each set of tags is not trained.
According to one preferred embodiment of the present invention, the division according to each label on geographical position, to the label
When being grouped, the label in boundaries of packets is divided to the multiple set of tags closed on.
According to one preferred embodiment of the present invention, it is described in the label for being included each set of tags and associated with label
User characteristic data as training data when, further comprise：Confidence level is less than to the user characteristics of default confidence threshold value
Data are filtered out from training data.
According to one preferred embodiment of the present invention, the label includes point-of-interest, or area-of-interest.
According to one preferred embodiment of the present invention, the customer location relevant information includes gps data, Wifi information and IP
At least one of address.
According to one preferred embodiment of the present invention, the drawing on geographical position according to the customer location relevant information
Point, determine that the set of tags corresponding to user characteristic data includes：The geographical position letter included according to the user characteristic data
Breath, spatial index or polymerization are carried out to the user characteristic data；According to spatial index or polymerization result, determine that the user is special
Levy the corresponding set of tags of data.
According to one preferred embodiment of the present invention, the classification submodel using corresponding to the set of tags is to the user
Characteristic is predicted, and is obtained the label of user and is included：If the classification submodel only one of which, by the user characteristics
Data are sent to the classification submodel, according to predicting the outcome for the classification submodel, obtain the label of user；Or, if described
Classification submodel has multiple, then sends the characteristic of each classification submodel of correspondence in user characteristic data to corresponding point
In class submodel, according to the prediction amalgamation result of multiple classification submodels, the label of user is obtained.
According to one preferred embodiment of the present invention, the prediction amalgamation result according to multiple classification submodels obtains user's
Label includes：The confidence level of user characteristic data determines to predict the outcome according to corresponding to each classification submodel, obtains the mark of user
Label；Or, determine to predict the outcome according to the confidence level predicted the outcome obtained by each classification submodel, obtain the label of user.
According to one preferred embodiment of the present invention, the confidence level of the user characteristic data is according to geographical position, the frequency of occurrences
Or at least one of signal intensity etc. is determined.
The present invention is to provide a kind of device of processing data, described device to solve the technical scheme that technical problem is used
Including：Acquiring unit, the user characteristic data for obtaining the relevant information containing customer location；Determining unit, for according to described
Division of the customer location relevant information on geographical position, determines the set of tags corresponding to user characteristic data；Predicting unit, is used
The user characteristic data is predicted in using the set of tags corresponding classification submodel, the label of user is obtained；Its
In, each set of tags corresponds to a classification submodel respectively.
According to one preferred embodiment of the present invention, described device also includes training unit, for using following training method instruction
Get classification submodel：Obtain label and the user characteristic data associated with label；According to each label in geographical position
On division, the label is grouped；Each set of tags is included into label and the user characteristics associated with label
The corresponding classification submodel of each set of tags is respectively trained as training data in data.
According to one preferred embodiment of the present invention, the training unit is for the division according to each label on geographical position
When being grouped to the label, the label in boundaries of packets is divided to the multiple set of tags closed on.
According to one preferred embodiment of the present invention, the training unit in the label for each set of tags to be included and
It is specific to perform when the user characteristic data associated with label is as training data：Confidence level is less than default confidence threshold value
User characteristic data filtered out from training data.
According to one preferred embodiment of the present invention, the label includes point-of-interest, or area-of-interest.
According to one preferred embodiment of the present invention, the customer location relevant information includes gps data, Wifi information and IP
At least one of address.
According to one preferred embodiment of the present invention, the determining unit for according to the customer location relevant information on ground
The division on position is managed, it is specific to perform when determining the set of tags corresponding to user characteristic data：According to the user characteristic data
Comprising geographical location information, spatial index or polymerization are carried out to the user characteristic data；According to spatial index or polymerization
As a result, the corresponding set of tags of the user characteristic data is determined.
According to one preferred embodiment of the present invention, the predicting unit is for utilizing classification corresponding to the set of tags
Model is predicted to the user characteristic data, when obtaining the label of user, specific to perform：If the classification submodel only has
One, then the user characteristic data is sent to the classification submodel, according to predicting the outcome for the classification submodel, used
The label at family；Or, if the classification submodel has multiple, by the spy of each classification submodel of correspondence in user characteristic data
Levy data to send into corresponding classification submodel, according to the prediction amalgamation result of multiple classification submodels, obtain the mark of user
Label.
According to one preferred embodiment of the present invention, the predicting unit according to the prediction of multiple classification submodels for merging
It is specific to perform when as a result obtaining the label of user：The confidence level of user characteristic data is determined according to corresponding to each classification submodel
Predict the outcome, obtain the label of user；Or, prediction knot is determined according to the confidence level predicted the outcome obtained by each classification submodel
Really, the label of user is obtained.
According to one preferred embodiment of the present invention, the confidence level of the user characteristic data is according to geographical position, the frequency of occurrences
Or at least one of signal intensity etc. is determined.
As can be seen from the above technical solutions, after the present invention is by the set of tags corresponding to acquisition and user characteristic data,
User characteristic data is predicted using with the classification submodel corresponding to set of tags, so as to improve to user characteristic data
Predetermined speed and prediction accuracy.
【Brief description of the drawings】
The method flow diagram that Fig. 1 provides for one embodiment of the invention.
The labeled packet schematic diagram that Fig. 2 provides for one embodiment of the invention.
The structure drawing of device that Fig. 3 provides for one embodiment of the invention.
The block diagram for the computer system/server that Fig. 4 provides for one embodiment of the invention.
【Embodiment】
In order that the object, technical solutions and advantages of the present invention are clearer, below in conjunction with the accompanying drawings with specific embodiment pair
The present invention is described in detail.
The term used in embodiments of the present invention is the purpose only merely for description specific embodiment, and is not intended to be limiting
The present invention." one kind ", " described " and "the" of singulative used in the embodiment of the present invention and appended claims
It is also intended to including most forms, unless context clearly shows that other implications.
It should be appreciated that term "and/or" used herein is only a kind of incidence relation for describing affiliated partner, represent
There may be three kinds of relations, for example, A and/or B, can be represented：Individualism A, while there is A and B, individualism B these three
Situation.In addition, character "/" herein, it is a kind of relation of "or" to typically represent forward-backward correlation object.
Depending on linguistic context, word as used in this " if " can be construed to " ... when " or " when ...
When " or " in response to determining " or " in response to detection ".Similarly, depending on linguistic context, phrase " if it is determined that " or " if detection
(condition or event of statement) " can be construed to " when it is determined that when " or " in response to determine " or " when the detection (condition of statement
Or event) when " or " in response to detection (condition or event of statement) ".
The method flow diagram that Fig. 1 provides for one embodiment of the invention, as shown in fig. 1, methods described includes：
In 101, the user characteristic data of the relevant information containing customer location is obtained.
In this step, acquired user characteristic data is the data containing customer location relevant information, wherein user
The relevant information of position includes at least one of gps data, Wifi information and IP address.It is i.e. special by acquired user
The geographical position of user can be learnt by levying data.
Alternatively, during one of the present embodiment implements, the customer location included in user characteristic data
Relevant information can be one, for example, only include gps data, or only include Wifi information, then or only comprising IP address.With
Customer location relevant information included in the characteristic of family can be multiple, such as comprising gps data and Wifi information, or
Person includes Wifi information and IP address, or comprising gps data and IP address, then or includes gps data, Wifi information
And IP address.Wherein, gps data is the information for directly reflecting customer location；Wifi information or IP address all include ground
Positional information is managed, after being positioned to Wifi information or IP address, the information of customer location is obtained.
In 102, according to division of the customer location relevant information on geographical position, user characteristic data institute is determined
Corresponding set of tags.
In this step, according to geographical location information contained in user characteristic data, user characteristic data is carried out
Spatial index or polymerization, then according to spatial index or polymerization result, determine the set of tags corresponding to user characteristic data.
Wherein, set of tags is divided previously according to each label on geographical position.Specifically, obtaining each
After label, according to division of each label on geographical position, the close label in geographical position is divided into one group, i.e., each label
All include all labels in a certain geographic range in group.
Alternatively, during one of the present embodiment implements, it can obtain and prestore when obtaining label
Label, can also obtain label by way of real time scan or search.Wherein, label can be POI (point of
Interest, point-of-interest), or AOI (area of interest, area-of-interest).
And when being grouped according to geographical position to each label, divide the close label in geographical position into according to preset rules
One group, so as to complete the packet to label.Longitude and latitude encryption algorithm can be used to be divided, such as using Geohash algorithms,
Encoded by the geographical position to acquired label, then divide all labels with identical coding into one group.May be used also
To be drawn using the spacial analytical method based on GIS (Geographic InformationSystem, GIS-Geographic Information System)
Point, each label in a certain geographic range is polymerize according to different-diameter scope, by each condensed together
Label divides one group into.When being grouped to label, in addition it is also necessary to the number of tags under each set of tags or each set of tags
Geographic range limited, so that it is guaranteed that having suitable number of tags or the suitable geographical model of covering in each set of tags
Enclose, it is to avoid because number of tags is excessive or the excessive training data scale brought of geographic range becomes big unfavorable factor.
The schematic diagram for the labeled packet that Fig. 2 provides for one embodiment of the invention.As shown in Figure 2, label is POI, is used
Geohash algorithms are grouped to acquired label.For example, in figure Geohash values for wtw3jnn set of tags in, have " also
Occupy pet shop ", it is " clean rich to wash during Geohash values in Hu Guan " and " beautiful China photograph " three labels, figure are wtw3jnj set of tags
There are " Building 26, Yongtai road Lane 595 55-57 " and " Building 27, Yongtai road Lane 595 58-59 " two labels.
In this step, using the method similar with dividing set of tags, spatial index or sky are carried out to user characteristic data
Between polymerize, so that it is determined which set of tags is user characteristic data particularly belong to., will be with user for example using Geohash algorithms
Corresponding geographical position is encoded in characteristic, obtains geohash values, then determines the use according to the geohash values
Set of tags belonging to the characteristic of family.
Alternatively, during one of the present embodiment implements, according to the geographical location information of user characteristic data
Identified set of tags can clearly indicate that user characteristic data belongs to only one of which, i.e. acquired geographical location information
Any group label.The set of tags according to determined by the geographical location information of user characteristic data can also have multiple, and for example user is special
The geographical location information for levying data is located at the boundary of packet label, then by the set of tags corresponding to multigroup label in border
It is used as the set of tags corresponding to user characteristic data；Or for another example contain multiple geographical location information in user characteristic data
When, for example user data includes gps data and Wifi information, and the geographical location information obtained according to gps data is A, and according to
The geographical location information that Wifi information is obtained is B, then it is special set of tags corresponding to geographical location information A and B to be all defined as into user
Levy the set of tags corresponding to data.
In 103, the user characteristic data is predicted using the set of tags corresponding classification submodel, obtained
The label of user.
In this step, predict that used classification submodel is previously according to training data training during user characteristic data
Obtain.And train classification submodel used in training data be each submodel included in label and with each label
Associated user characteristic data.
After the packet to label is completed so that every group of label is corresponded under a submodel, i.e., each submodel respectively
Include whole labels in a certain geographic range.As shown in Figure 2, such as corresponding submodels of set of tags wtw3jnm are son
The corresponding submodel of model A, set of tags wtw3jjv is submodel B, by that analogy.
Label according to included in each submodel, obtains the user characteristic data associated with each label.Then
Using the label in submodel and the user characteristic data associated with each label as the submodel training data.Its
In, it can learn which label corresponding to this feature data be by user characteristic data, and acquired user characteristics
Data can be the characteristic containing customer location relevant information, such as containing in gps data, Wifi information and IP address
At least one.Can also be the characteristic for not containing customer location relevant information, for example user uses the model of mobile phone, product
Board etc..
, can also be to the training data acquired in each submodel before being trained using training data to submodel
Processing is optimized, so that training data used in each submodel is more accurate, and is avoided when training submodel
Introduce unnecessary training data.Optimization processing to sub- model training data includes：To included in submodel label it is excellent
Change and the optimization pair the user characteristic data associated with label.
Wherein, label included in submodel is optimized for：Label at boundaries of packets is respectively divided to facing
Near multiple set of tags.When carrying out the division of set of tags in the geographical position according to label, it is possible in the packet of set of tags
There is label in boundary, then the label at set of tags boundaries of packets is divided to the multiple set of tags closed on, so that
The division of set of tags is more accurate.For example, as shown in Figure 2, label " children's theme park " is located at the set of tags closed on
Wtw3jnm and set of tags wtw3jjv boundary, then label " children's theme park " is respectively divided to set of tags
Wtw3jnm and set of tags wtw3jjv, so that submodel corresponding with set of tags wtw3jnm and set of tags wtw3jjv
A and submodel B include label " children's theme park ".
Pair user characteristic data associated with label is optimized for：Confidence level is less than to the user of default confidence threshold value
Characteristic is filtered out from training data, i.e., the user characteristic data corresponding to each label in each set of tags is carried out
Filtering, so as to avoid submodel from introducing unnecessary training data.Wherein, the confidence level of user characteristic data can be according to geography
At least one of position, the frequency of occurrences or signal intensity etc. are determined.
Alternatively, can be the user that will be far from label physical location during one of the present embodiment implements
Characteristic is filtered, and although some characteristic of such as user is associated with a label, but this feature data of user
Physical location not near the label, then by user's this feature data filtering.Can also be that will only be put in user characteristic data
The higher characteristic of reliability retains, and such as confidence level of Wifi information is the intensity of Wifi signals, then only retains Wifi information
The higher data of middle Wifi signal intensities, the weaker Wifi information of signal intensity is filtered.Can also be will be only a small number of
The characteristic that user possesses is filtered, i.e., filtered the relatively low user characteristic data of the frequency of occurrences.Such as user's hand
The only several users of the characteristic of type number possess, then are filtered the characteristic of the user mobile phone model.
Submodel is trained using the training data after optimization processing, so as to obtain classification submodel so that classification
Submodel can be in training data the user characteristic data associated with label determine corresponding label.In training
When, each submodel is trained using traditional disaggregated model algorithm, and because the training process of each submodel is completely only
It is vertical, therefore multiple submodels are concurrently trained using distributed type assemblies technology, so that needed for considerably reducing training submodel
Time.Submodel of classifying is obtained after the training to submodel is completed, it is possible to using classification submodel to user characteristics number
According to being predicted.
Using the classification submodel determined in step 102 corresponding to set of tags, user characteristic data is predicted, from
And obtain the label of user.I.e. by this step, POI that just can be according to where acquired user characteristic data determines user
Or AOI.
In a step 102, if according to determined by the geographical location information of user characteristic data set of tags only one of which,
To should set of tags classification submodel also only one of which, according to predicting the outcome for the classification submodel, determine user characteristics number
According to corresponding label.If in a step 102, the set of tags according to determined by the geographical location information of user characteristic data has many
When individual, the classification submodel of the multiple set of tags of correspondence also has multiple, then needs the progress that predicts the outcome to multiple classification submodels
Merge, predicted the outcome merging to predict the outcome as final, so that it is determined that the label corresponding to user characteristic data.
Alternatively, during one of the present embodiment implements, can according to user corresponding to disaggregated model it is special
The confidence level for levying data determines to predict the outcome.Wherein, the confidence level of user characteristic data can be according to customer location relevant information
In at least one of geographical position, the frequency of occurrences or signal intensity etc. determine.Such as confidence level is user characteristic data
During the intensity of middle customer location relevant information, if containing gps data and Wifi information, root in acquired user characteristic data
One is obtained according to gps data to predict the outcome, obtaining another according to Wifi information predicts the outcome, but Wifi in user data
The signal of information is weaker, then will be given up by predicting the outcome obtained by Wifi information, by predicting the outcome for being obtained according to gps data
It is defined as the label corresponding to user characteristic data.In another example confidence level is customer location relevant information in user characteristic data
During the frequency of occurrences, if containing IP address and Wifi information in acquired user characteristic data, one is obtained according to IP address
Predict the outcome, obtaining another according to Wifi information predicts the outcome, but the frequency of occurrences of IP address is less than Wifi information
The frequency of occurrences, then will be given up by predicting the outcome obtained by IP address, by the determination that predicts the outcome obtained according to Wifi information
For the label corresponding to user characteristic data.
It can also be training in advance order models, confidence level sequence is carried out to predicting the outcome, so that according to predicting the outcome
Confidence level height determines to predict the outcome.For example, being found when using training data training classification submodel, instructed using gps data
User tag obtained by practicing is the most accurate, next to that using Wifi information, then gps data will be used pre- obtained by being predicted
The confidence level for surveying result is set to highest, and the confidence level of Wifi information is next, then by obtained by confidence level highest gps data
The label being defined as corresponding to user characteristic data that predicts the outcome.
From the foregoing it can be that user characteristic data can be predicted under multiple classification submodels simultaneously, then will
Predicting the outcome obtained by multiple classification submodels merge obtain it is final predict the outcome, and due to each classification submodel
Scale it is all smaller, therefore predetermined speed of Forecasting Methodology provided by the present invention is better than predetermined speed of conventional model.
Structure drawing of device provided in an embodiment of the present invention is described in detail below.As shown in Figure 3, described device includes：
Training unit 31, acquiring unit 32, determining unit 33 and predicting unit 34.
Training unit 31, for obtaining submodel of classifying previously according to training data training.
The used training data when training submodel of training unit 31 includes each label and closed with each label
The user characteristic data of connection.Therefore training unit 31 using training data train submodel before, it is necessary to obtain each label with
And the user characteristic data associated with label.
After training unit 31 obtains each label, according to the division on each label geographical position, set of tags is obtained.Specifically
Ground, training unit 31 is after each label is obtained, according to division of each label on geographical position, by the close label in geographical position
It is divided into all labels included in one group, i.e., each set of tags in a certain geographic range.
Alternatively, during one of the present embodiment implements, training unit 31 can obtain pre- when obtaining label
The label first stored, can also obtain label by way of real time scan or search.Wherein, label can be POI (point
Of interest, point-of-interest), or AOI (area ofinterest, area-of-interest).
And training unit 31 is according to geographical position to each label when being grouped, according to preset rules by geographical position phase
Near label divides one group into, so as to complete the packet to label.Training unit 31 can use longitude and latitude encryption algorithm to be drawn
Point, such as using Geohash algorithms, encoded by the geographical position to acquired label, then there will be identical coding
All labels divide one group into.Training unit 31 can also use based on GIS (Geographic Information System,
GIS-Geographic Information System) spacial analytical method divided, using different polymerization diameters, by a certain geographic range
Each label is polymerize, and divides each label condensed together into one group.
Training unit 31 to each label when being grouped, in addition it is also necessary to the number of tags under each set of tags or each
The geographic range of set of tags covering is limited, so that it is guaranteed that having suitable number of tags or covering to close in each set of tags
Suitable geographic range, it is to avoid because number of tags is excessive or the excessive training data scale brought of geographic range becomes greatly unfavorable
Factor.
Training unit 31 is after the packet to label is completed so that every group of label corresponds to a submodel respectively, i.e., each
All include whole labels in a certain geographic range under submodel.Label according to included in each submodel, training is single
Member 31 obtains the user characteristic data associated with each label.Then training unit 31 by the label in submodel and with it is every
The associated user characteristic data of individual label as the submodel training data.Wherein, it can be obtained by user characteristic data
Know which label corresponding to this feature data be, and acquired user characteristic data can be related containing customer location
The characteristic of information, such as containing at least one in gps data, Wifi information and IP address.Can also be not contain
The characteristic of customer location relevant information, such as user uses the model of mobile phone, brand.
Training unit 31, can also be to acquired in each submodel before being trained using training data to submodel
Training data optimize processing so that training data used in each submodel is more accurate, and avoid in instruction
Unnecessary training data is introduced when practicing submodel.The optimization processing of the sub- model training data of 31 pairs of training unit includes：Antithetical phrase
The optimization of label and the optimization pair the user characteristic data associated with label included in model.
Wherein, training unit 31 is optimized for label included in submodel：By the label at boundaries of packets point
The multiple set of tags closed on are not divided to.When carrying out the division of set of tags in the geographical position according to label, it is possible in mark
There is label at the boundaries of packets of label group, then the label at set of tags boundaries of packets is divided to the multiple labels closed on
Group, so that the division of set of tags is more accurate.For example, as shown in Figure 2, label " children's theme park " is located at what is closed on
Set of tags wtw3jnm and set of tags wtw3jjv boundary, then label " children's theme park " is respectively divided to set of tags
Wtw3jnm and set of tags wtw3jjv, so that submodel corresponding with set of tags wtw3jnm and set of tags wtw3jjv
A and submodel B include label " children's theme park ".
Training unit 31 pairs of user characteristic datas associated with label are optimized for：Confidence level is less than default confidence level
The user characteristic data of threshold value is filtered out from training data, i.e., special to the user corresponding to each label in each set of tags
Levy data to be filtered, so as to avoid submodel from introducing unnecessary training data.Wherein, the confidence level of user characteristic data can
To be determined according at least one of geographical position, the frequency of occurrences or signal intensity etc..
Alternatively, during one of the present embodiment implements, training unit 31 can will be far from label actual bit
The user characteristic data put is filtered, and although some characteristic of such as user is associated with a label, but user should
The physical location of characteristic is not near the label, then by user's this feature data filtering.Training unit 31 only can also will
The characteristic that confidence level is higher in user characteristic data retains, and such as confidence level of Wifi information is the intensity of Wifi signals,
Then only retain the higher data of Wifi signal intensities in Wifi information, the weaker Wifi information of signal intensity is filtered.Instruction
The characteristic that only a few users possess can also be filtered by practicing unit 31, such as characteristic of user mobile phone model
Only several users possess, then are filtered the characteristic of the user mobile phone model.
Training unit 31 is trained using the training data after optimization processing to submodel, so as to obtain classification submodule
Type so that the user characteristic data associated with label that classification submodel can be in training data determines corresponding
Label.Training unit 31 is in training, and each submodel is trained using traditional disaggregated model algorithm, and due to every height
The training process of model is completely independent, therefore concurrently trains multiple submodels using distributed type assemblies technology, so that greatly
Reduce the time needed for training submodel.Submodel of classifying is obtained after the training to submodel is completed, it is possible to use and divide
Class submodel is predicted to user characteristic data.
Acquiring unit 32, the user characteristic data for obtaining the relevant information containing customer location.
User characteristic data acquired in acquiring unit 32 is the data containing customer location relevant information, wherein user position
The relevant information put includes at least one of gps data, Wifi information and IP address.I.e. by acquired in acquiring unit 32
User characteristic data can learn the geographical position of user.
Alternatively, during one of the present embodiment implements, in the user characteristic data that acquiring unit 32 is obtained
Comprising customer location relevant information can be one, for example only include gps data, or only include Wifi information, then or
Person only includes IP address.The customer location relevant information included in user characteristic data that acquiring unit 32 is obtained can be many
It is individual, such as comprising gps data and Wifi information, either comprising Wifi information and IP address or comprising gps data and
IP address, then or include gps data, Wifi information and IP address.Wherein, gps data is directly reflection customer location
Information；Wifi information or IP address all include geographical location information, after being positioned to Wifi information or IP address, obtain
Obtain the information of customer location.
Determining unit 33, for the division according to the customer location relevant information on geographical position, determines that user is special
Levy the set of tags corresponding to data.
Determining unit 33 is carried out empty according to geographical location information contained in user characteristic data to user characteristic data
Between index or polymerize, then according to spatial index or polymerization result, determine the set of tags corresponding to user characteristic data.
Determining unit 33 carries out space using the method similar with the division of training unit 31 set of tags to user characteristic data
Index or spatial clustering, so that it is determined which set of tags is user characteristic data particularly belong to.For example using Geohash algorithms,
It will be encoded with geographical position corresponding in user characteristic data, geohash values obtained, then according to the geohash values
Determine the set of tags belonging to the user characteristic data.
Alternatively, during one of the present embodiment implements, determining unit 33 is according to the ground of user characteristic data
Managing set of tags determined by positional information can clearly indicate that user is special with only one of which, i.e. acquired geographical location information
Levy which group label is data belong to.The set of tags according to determined by the geographical location information of user characteristic data of determining unit 33 also may be used
Multiple to have, the geographical location information of such as user characteristic data is located at the boundary of packet label, then by many of border
Set of tags corresponding to group label is used as the set of tags corresponding to user characteristic data；Or for another example in user characteristic data
During containing multiple geographical location information, for example user data includes gps data and Wifi information, the ground obtained according to gps data
Reason positional information is A, and the geographical location information obtained according to Wifi information is B, then by corresponding to geographical location information A and B
Set of tags is all defined as the set of tags corresponding to user characteristic data.
Predicting unit 34, it is pre- for being carried out using the corresponding classification submodel of the set of tags to the user characteristic data
Survey, obtain the label of user.
Using set of tags determined by determining unit 33, it is right with set of tags institute that predicting unit 34 is determined by training unit 31
The classification submodel answered, then classification submodel is predicted to user characteristic data determined by the utilization of predicting unit 34, from
And obtain the label of user.I.e. by predicting unit 34, user place just can be determined according to acquired user characteristic data
POI or AOI.
If it is determined that the set of tags only one of which according to determined by the geographical location information of user characteristic data of unit 33, then right
Should set of tags classification submodel also only one of which, predicting unit 34 predicts the outcome according to the classification submodel, it is determined that with
Label corresponding to the characteristic of family.If it is determined that the label according to determined by the geographical location information of user characteristic data of unit 33
When group has multiple, the classification submodel of the multiple set of tags of correspondence also has multiple, then predicting unit 34 is needed to multiple classification submodules
Predicting the outcome for type is merged, and is predicted the outcome merging to predict the outcome as final, so that it is determined that user characteristic data institute
Corresponding label.
Alternatively, during one of the present embodiment implements, predicting unit 34 can be right according to disaggregated model institute
The confidence level of user characteristic data is answered to determine to predict the outcome.Wherein, the confidence level of user characteristic data can be according to customer location
At least one of geographical position, the frequency of occurrences or signal intensity in relevant information etc. are determined.Such as confidence level is user
In characteristic during the intensity of customer location relevant information, if in acquired user characteristic data containing gps data and
Wifi information, obtains one according to gps data and predicts the outcome, obtaining another according to Wifi information predicts the outcome, but uses
The signal of Wifi information is weaker in user data, then will be given up by predicting the outcome obtained by Wifi information, will be obtained according to gps data
To the label that is defined as corresponding to user characteristic data of predicting the outcome.In another example confidence level is user position in user characteristic data
When putting the frequency of occurrences of relevant information, if containing IP address and Wifi information in acquired user characteristic data, according to IP
Address obtains one and predicted the outcome, and obtaining another according to Wifi information predicts the outcome, but the frequency of occurrences of IP address will
Less than the frequency of occurrences of Wifi information, then it will be given up by predicting the outcome obtained by IP address, by what is obtained according to Wifi information
Predict the outcome and be defined as label corresponding to user characteristic data.
Predicting unit 34 can carry out confidence level sequence, so that according to pre- with training in advance order models to predicting the outcome
The confidence level height for surveying result determines to predict the outcome.For example, being found when using training data training classification submodel, use
User tag obtained by gps data training is the most accurate, next to that using Wifi information, then gps data will be used to carry out pre-
The confidence level predicted the outcome obtained by surveying is set to highest, and the confidence level of Wifi information is next, then by confidence level highest GPS
Predicting the outcome obtained by data is defined as the label corresponding to user characteristic data.
From the foregoing it can be that predicting unit 34 can be carried out user characteristic data under multiple classification submodels simultaneously
Prediction, then predicting the outcome obtained by multiple classification submodels is merged obtain it is final predict the outcome, and due to every
The scale of individual classification submodel is all smaller, therefore predetermined speed of Forecasting Methodology provided by the present invention is better than conventional model
Predetermined speed.
Fig. 4 shows the frame suitable for being used for the exemplary computer system/server 012 for realizing embodiment of the present invention
Figure.The computer system/server 012 that Fig. 4 is shown is only an example, to the function of the embodiment of the present invention and should not be used
Range band carrys out any limitation.
As shown in figure 4, computer system/server 012 is showed in the form of universal computing device.Computer system/clothes
The component of business device 012 can include but is not limited to：One or more processor or processing unit 016, system storage
028, the bus 018 of connection different system component (including system storage 028 and processing unit 016).
Bus 018 represents the one or more in a few class bus structures, including memory bus or Memory Controller,
Peripheral bus, graphics acceleration port, processor or the local bus using any bus structures in a variety of bus structures.Lift
For example, these architectures include but is not limited to industry standard architecture (ISA) bus, MCA (MAC)
Bus, enhanced isa bus, VESA's (VESA) local bus and periphery component interconnection (PCI) bus.
Computer system/server 012 typically comprises various computing systems computer-readable recording medium.These media can be appointed
The usable medium what can be accessed by computer system/server 012, including volatibility and non-volatile media, movably
With immovable medium.
System storage 028 can include the computer system readable media of form of volatile memory, for example, deposit at random
Access to memory (RAM) 030 and/or cache memory 032.Computer system/server 012 may further include other
Removable/nonremovable, volatile/non-volatile computer system storage medium.Only as an example, storage system 034 can
For reading and writing immovable, non-volatile magnetic media (Fig. 4 is not shown, is commonly referred to as " hard disk drive ").Although in Fig. 4
It is not shown, it can provide for the disc driver to may move non-volatile magnetic disk (such as " floppy disk ") read-write, and pair can
The CD drive of mobile anonvolatile optical disk (such as CD-ROM, DVD-ROM or other optical mediums) read-write.In these situations
Under, each driver can be connected by one or more data media interfaces with bus 018.Memory 028 can include
At least one program product, the program product has one group of (for example, at least one) program module, and these program modules are configured
To perform the function of various embodiments of the present invention.
Program/utility 040 with one group of (at least one) program module 042, can be stored in such as memory
In 028, such program module 042 includes --- but being not limited to --- operating system, one or more application program, other
The realization of network environment is potentially included in each or certain combination in program module and routine data, these examples.Journey
Sequence module 042 generally performs function and/or method in embodiment described in the invention.
Computer system/server 012 can also with one or more external equipments 014 (for example keyboard, sensing equipment,
Display 024 etc.) communication, in the present invention, computer system/server 012 is communicated with outside radar equipment, can also be with
One or more enables a user to the equipment communication interacted with the computer system/server 012, and/or with causing the meter
Any equipment (such as network interface card, modulation that calculation machine systems/servers 012 can be communicated with one or more of the other computing device
Demodulator etc.) communication.This communication can be carried out by input/output (I/O) interface 022.Also, computer system/clothes
Business device 012 can also pass through network adapter 020 and one or more network (such as LAN (LAN), wide area network (WAN)
And/or public network, such as internet) communication.As illustrated, network adapter 020 by bus 018 and computer system/
Other modules communication of server 012.It should be understood that although not shown in the drawings, computer system/server 012 can be combined
Using other hardware and/or software module, include but is not limited to：Microcode, device driver, redundant processing unit, outside magnetic
Dish driving array, RAID system, tape drive and data backup storage system etc..
Processing unit 016 is stored in program in system storage 028 by operation, thus perform various function application with
And data processing, a kind of method of processing data is for example realized, can be included：
Obtain the user characteristic data of the relevant information containing customer location；
According to division of the customer location relevant information on geographical position, the mark corresponding to user characteristic data is determined
Label group；
The user characteristic data is predicted using the set of tags corresponding classification submodel, the mark of user is obtained
Label；
Wherein, each set of tags corresponds to a classification submodel respectively.
Above-mentioned computer program can be arranged in computer-readable storage medium, i.e., the computer-readable storage medium is encoded with
Computer program, the program by one or more computers when being performed so that one or more computers are performed in the present invention
State the method flow shown in embodiment and/or device operation.For example, by the method stream of said one or multiple computing devices
Journey, can include：
Obtain the user characteristic data of the relevant information containing customer location；
According to division of the customer location relevant information on geographical position, the mark corresponding to user characteristic data is determined
Label group；
The user characteristic data is predicted using the set of tags corresponding classification submodel, the mark of user is obtained
Label；
Wherein, each set of tags corresponds to a classification submodel respectively.
Over time, the development of technology, medium implication is more and more extensive, and the route of transmission of computer program is no longer limited by
Tangible medium, directly can also be downloaded from network etc..Any combination of one or more computer-readable media can be used.
Computer-readable medium can be computer-readable signal media or computer-readable recording medium.Computer-readable storage medium
Matter for example may be-but not limited to-system, device or the device of electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor, or
Combination more than person is any.The more specifically example (non exhaustive list) of computer-readable recording medium includes：With one
Or the electrical connections of multiple wires, portable computer diskette, hard disk, random access memory (RAM), read-only storage (ROM),
Erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disc read-only storage (CD-ROM), light
Memory device, magnetic memory device or above-mentioned any appropriate combination.In this document, computer-readable recording medium can
Be it is any include or storage program tangible medium, the program can be commanded execution system, device or device use or
Person is in connection.
Computer-readable signal media can be included in a base band or as the data-signal of carrier wave part propagation,
Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including --- but
It is not limited to --- electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be
Any computer-readable medium beyond computer-readable recording medium, the computer-readable medium can send, propagate or
Transmit for being used or program in connection by instruction execution system, device or device.
The program code included on computer-readable medium can be transmitted with any appropriate medium, including --- but do not limit
In --- wireless, electric wire, optical cable, RF etc., or above-mentioned any appropriate combination.
It can be write with one or more programming languages or its combination for performing the computer that the present invention is operated
Program code, described program design language includes object oriented program language-such as Java, Smalltalk, C++,
Also including conventional procedural programming language-such as " C " language or similar programming language.Program code can be with
Fully perform, partly perform on the user computer on the user computer, as independent software kit execution, a portion
Divide part execution or the execution completely on remote computer or server on the remote computer on the user computer.
Be related in the situation of remote computer, remote computer can be by the network of any kind --- including LAN (LAN) or
Wide area network (WAN) is connected to subscriber computer, or, it may be connected to outer computer (is for example provided using Internet service
Business comes by Internet connection).
Using technical scheme provided by the present invention, by obtaining with after the set of tags corresponding to user characteristic data, making
User characteristic data is predicted with the classification submodel corresponding to set of tags, so as to improve to user characteristic data
Predetermined speed and the degree of accuracy.
In several embodiments provided by the present invention, it should be understood that disclosed system, apparatus and method can be with
Realize by another way.For example, device embodiment described above is only schematical, for example, the unit
Divide, only a kind of division of logic function there can be other dividing mode when actually realizing.
The unit illustrated as separating component can be or may not be it is physically separate, it is aobvious as unit
The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On NE.Some or all of unit therein can be selected to realize the mesh of this embodiment scheme according to the actual needs
's.
In addition, each functional unit in each embodiment of the invention can be integrated in a processing unit, can also
That unit is individually physically present, can also two or more units it is integrated in a unit.Above-mentioned integrated list
Member can both be realized in the form of hardware, it would however also be possible to employ hardware adds the form of SFU software functional unit to realize.
The above-mentioned integrated unit realized in the form of SFU software functional unit, can be stored in an embodied on computer readable and deposit
In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are to cause a computer
Equipment (can be personal computer, server, or network equipment etc.) or processor (processor) perform the present invention each
The part steps of embodiment methods described.And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (Read-
Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disc or CD etc. it is various
Can be with the medium of store program codes.
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
God is with principle, and any modification, equivalent substitution and improvements done etc. should be included within the scope of protection of the invention.
Claims (22)
1. a kind of method of processing data, it is characterised in that methods described includes：
Obtain the user characteristic data of the relevant information containing customer location；
According to division of the customer location relevant information on geographical position, the label corresponding to user characteristic data is determined
Group；
The user characteristic data is predicted using the set of tags corresponding classification submodel, the label of user is obtained；
Wherein, each set of tags corresponds to a classification submodel respectively.
2. according to the method described in claim 1, it is characterised in that the classification submodel is advance using following training method
Obtain：
Obtain label and the user characteristic data associated with label；
According to division of each label on geographical position, the label is grouped；
Label that each set of tags is included and the user characteristic data associated with label are instructed respectively as training data
Practice the corresponding classification submodel of each set of tags.
3. method according to claim 2, it is characterised in that in the division according to each label on geographical position, to institute
When stating label and being grouped, the label in boundaries of packets is divided to the multiple set of tags closed on.
4. method according to claim 2, it is characterised in that the label for being included each set of tags and with
When the associated user characteristic data of label is as training data, further comprise：
The user characteristic data that confidence level is less than default confidence threshold value is filtered out from training data.
5. according to the method described in claim 1, it is characterised in that the label includes point-of-interest, or area-of-interest.
6. according to the method described in claim 1, it is characterised in that the customer location relevant information includes gps data, Wifi
At least one of information and IP address.
7. according to the method described in claim 1, it is characterised in that it is described according to the customer location relevant information in geographical position
The division put, determines that the set of tags corresponding to user characteristic data includes：
The geographical location information included according to the user characteristic data, to the user characteristic data carry out spatial index or
Polymerization；
According to spatial index or polymerization result, the corresponding set of tags of the user characteristic data is determined.
8. according to the method described in claim 1, it is characterised in that the classification submodel using corresponding to the set of tags
The user characteristic data is predicted, obtaining the label of user includes：
If the classification submodel only one of which, the user characteristic data is sent to the classification submodel, according to this point
Predicting the outcome for class submodel, obtains the label of user；Or,
If the classification submodel has multiple, the characteristic of each classification submodel of correspondence in user characteristic data is sent
Into corresponding classification submodel, according to the prediction amalgamation result of multiple classification submodels, the label of user is obtained.
9. method according to claim 8, it is characterised in that the prediction amalgamation result according to multiple classification submodels
Obtaining the label of user includes：
The confidence level of user characteristic data determines to predict the outcome according to corresponding to each classification submodel, obtains the label of user；Or
Person,
Confidence level according to being predicted the outcome obtained by each classification submodel determines to predict the outcome, and obtains the label of user.
10. the method according to claim 4 or 9, it is characterised in that the confidence level of the user characteristic data is according to geographical
At least one of position, the frequency of occurrences or signal intensity etc. are determined.
11. a kind of device of processing data, it is characterised in that described device includes：
Acquiring unit, the user characteristic data for obtaining the relevant information containing customer location；
Determining unit, for the division according to the customer location relevant information on geographical position, determines user characteristic data
Corresponding set of tags；
Predicting unit, for being predicted using the corresponding classification submodel of the set of tags to the user characteristic data, is obtained
To the label of user；
Wherein, each set of tags corresponds to a classification submodel respectively.
12. device according to claim 11, it is characterised in that described device also includes training unit, for using such as
Lower training method training obtains submodel of classifying：
Obtain label and the user characteristic data associated with label；
According to division of each label on geographical position, the label is grouped；
Label that each set of tags is included and the user characteristic data associated with label are instructed respectively as training data
Practice the corresponding classification submodel of each set of tags.
13. device according to claim 12, it is characterised in that the training unit for according to each label in geography
When division on position is grouped to the label, the label in boundaries of packets is divided to the multiple set of tags closed on.
14. device according to claim 12, it is characterised in that the training unit is for each set of tags to be wrapped
It is specific to perform when the label that contains and the user characteristic data associated with label are as training data：
The user characteristic data that confidence level is less than default confidence threshold value is filtered out from training data.
15. device according to claim 11, it is characterised in that the label includes point-of-interest, or region of interest
Domain.
16. device according to claim 11, it is characterised in that the customer location relevant information include gps data,
At least one of Wifi information and IP address.
17. device according to claim 11, it is characterised in that the determining unit is for according to the customer location
Division of the relevant information on geographical position, it is specific to perform when determining the set of tags corresponding to user characteristic data：
The geographical location information included according to the user characteristic data, to the user characteristic data carry out spatial index or
Polymerization；
According to spatial index or polymerization result, the corresponding set of tags of the user characteristic data is determined.
18. device according to claim 11, it is characterised in that the predicting unit is for utilizing the set of tags institute
Corresponding classification submodel is predicted to the user characteristic data, specific to perform when obtaining the label of user：
If the classification submodel only one of which, the user characteristic data is sent to the classification submodel, according to this point
Predicting the outcome for class submodel, obtains the label of user；Or,
If the classification submodel has multiple, the characteristic of each classification submodel of correspondence in user characteristic data is sent
Into corresponding classification submodel, according to the prediction amalgamation result of multiple classification submodels, the label of user is obtained.
19. device according to claim 18, it is characterised in that the predicting unit is for according to multiple classification submodules
It is specific to perform when the prediction amalgamation result of type obtains the label of user：
The confidence level of user characteristic data determines to predict the outcome according to corresponding to each classification submodel, obtains the label of user；Or
Person,
Confidence level according to being predicted the outcome obtained by each classification submodel determines to predict the outcome, and obtains the label of user.
20. the device according to claim 14 or 19, it is characterised in that the confidence level of the user characteristic data is according to ground
At least one of position, the frequency of occurrences or signal intensity etc. is managed to determine.
21. a kind of equipment, it is characterised in that the equipment includes：
One or more processors；
Storage device, for storing one or more programs,
When one or more of programs are by one or more of computing devices so that one or more of processors are real
The existing method as described in any in claim 1-10.
22. a kind of storage medium for including computer executable instructions, the computer executable instructions are by computer disposal
For performing the method as described in any in claim 1-10 when device is performed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710260035.2A CN107194412A (en) | 2017-04-20 | 2017-04-20 | A kind of method of processing data, device, equipment and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710260035.2A CN107194412A (en) | 2017-04-20 | 2017-04-20 | A kind of method of processing data, device, equipment and computer-readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107194412A true CN107194412A (en) | 2017-09-22 |
Family
ID=59871340
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710260035.2A Pending CN107194412A (en) | 2017-04-20 | 2017-04-20 | A kind of method of processing data, device, equipment and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107194412A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108021986A (en) * | 2017-10-27 | 2018-05-11 | 平安科技（深圳）有限公司 | Electronic device, multi-model sample training method and computer-readable recording medium |
| CN108922079A (en) * | 2018-06-05 | 2018-11-30 | 阿里巴巴集团控股有限公司 | A kind of data processing method, article detection method, device |
| CN109063736A (en) * | 2018-06-29 | 2018-12-21 | 考拉征信服务有限公司 | Data classification method, device, electronic equipment and computer readable storage medium |
| CN109102159A (en) * | 2018-07-18 | 2018-12-28 | 平安科技（深圳）有限公司 | Passenger's rating model generation method, device, computer equipment and storage medium |
| CN109199604A (en) * | 2018-08-31 | 2019-01-15 | 浙江大学宁波理工学院 | A kind of pedicle screw based on multiple features most preferably enters independent positioning method |
| CN110059771A (en) * | 2019-05-10 | 2019-07-26 | 合肥工业大学 | A kind of interactive vehicle data classification method in the case where sequence is supported |
| CN110490625A (en) * | 2018-05-11 | 2019-11-22 | 北京京东尚科信息技术有限公司 | User preference determines method and device, electronic equipment, storage medium |
| CN110598122A (en) * | 2018-06-13 | 2019-12-20 | 百度在线网络技术（北京）有限公司 | Social group mining method, device, equipment and storage medium |
| CN111629319A (en) * | 2019-02-28 | 2020-09-04 | 中国移动通信有限公司研究院 | Position prediction method and device |
| CN112699206A (en) * | 2021-03-23 | 2021-04-23 | 上海迹寻科技有限公司 | User position and residence analysis method and device |
| CN114065947A (en) * | 2021-11-15 | 2022-02-18 | 深圳大学 | Data access speculation method and device, storage medium and electronic equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102047249A (en) * | 2008-05-27 | 2011-05-04 | 高通股份有限公司 | Method and apparatus for aggregating and presenting data associated with geographic locations |
| CN103093237A (en) * | 2013-01-15 | 2013-05-08 | 中国科学院自动化研究所 | Face detecting method based on structural model |
| CN103455545A (en) * | 2012-06-04 | 2013-12-18 | 国际商业机器公司 | Location estimation of social network users |
| CN103945533A (en) * | 2014-05-15 | 2014-07-23 | 济南嘉科电子技术有限公司 | Big data based wireless real-time position positioning method |
| CN104463633A (en) * | 2014-12-19 | 2015-03-25 | 成都品果科技有限公司 | User segmentation method based on geographic position and interest point information |
| CN106131959A (en) * | 2016-08-11 | 2016-11-16 | 电子科技大学 | A kind of dual-positioning method divided based on Wi Fi signal space |
| CN106462627A (en) * | 2014-12-29 | 2017-02-22 | 谷歌公司 | Analyzing semantic places and related data from a plurality of location data reports |
| CN106507476A (en) * | 2016-11-22 | 2017-03-15 | 上海斐讯数据通信技术有限公司 | A kind of WiFi localization methods and server and location model construction method |
-
2017
- 2017-04-20 CN CN201710260035.2A patent/CN107194412A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102047249A (en) * | 2008-05-27 | 2011-05-04 | 高通股份有限公司 | Method and apparatus for aggregating and presenting data associated with geographic locations |
| CN103455545A (en) * | 2012-06-04 | 2013-12-18 | 国际商业机器公司 | Location estimation of social network users |
| CN103093237A (en) * | 2013-01-15 | 2013-05-08 | 中国科学院自动化研究所 | Face detecting method based on structural model |
| CN103945533A (en) * | 2014-05-15 | 2014-07-23 | 济南嘉科电子技术有限公司 | Big data based wireless real-time position positioning method |
| CN104463633A (en) * | 2014-12-19 | 2015-03-25 | 成都品果科技有限公司 | User segmentation method based on geographic position and interest point information |
| CN106462627A (en) * | 2014-12-29 | 2017-02-22 | 谷歌公司 | Analyzing semantic places and related data from a plurality of location data reports |
| CN106131959A (en) * | 2016-08-11 | 2016-11-16 | 电子科技大学 | A kind of dual-positioning method divided based on Wi Fi signal space |
| CN106507476A (en) * | 2016-11-22 | 2017-03-15 | 上海斐讯数据通信技术有限公司 | A kind of WiFi localization methods and server and location model construction method |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108021986A (en) * | 2017-10-27 | 2018-05-11 | 平安科技（深圳）有限公司 | Electronic device, multi-model sample training method and computer-readable recording medium |
| CN110490625A (en) * | 2018-05-11 | 2019-11-22 | 北京京东尚科信息技术有限公司 | User preference determines method and device, electronic equipment, storage medium |
| CN108922079A (en) * | 2018-06-05 | 2018-11-30 | 阿里巴巴集团控股有限公司 | A kind of data processing method, article detection method, device |
| CN110598122B (en) * | 2018-06-13 | 2022-04-01 | 百度在线网络技术（北京）有限公司 | Social group mining method, device, equipment and storage medium |
| CN110598122A (en) * | 2018-06-13 | 2019-12-20 | 百度在线网络技术（北京）有限公司 | Social group mining method, device, equipment and storage medium |
| CN109063736A (en) * | 2018-06-29 | 2018-12-21 | 考拉征信服务有限公司 | Data classification method, device, electronic equipment and computer readable storage medium |
| CN109102159A (en) * | 2018-07-18 | 2018-12-28 | 平安科技（深圳）有限公司 | Passenger's rating model generation method, device, computer equipment and storage medium |
| CN109199604B (en) * | 2018-08-31 | 2020-12-01 | 浙江大学宁波理工学院 | Multi-feature-based pedicle screw optimal entry point positioning method |
| CN109199604A (en) * | 2018-08-31 | 2019-01-15 | 浙江大学宁波理工学院 | A kind of pedicle screw based on multiple features most preferably enters independent positioning method |
| CN111629319A (en) * | 2019-02-28 | 2020-09-04 | 中国移动通信有限公司研究院 | Position prediction method and device |
| CN111629319B (en) * | 2019-02-28 | 2022-05-31 | 中国移动通信有限公司研究院 | Position prediction method and device |
| CN110059771B (en) * | 2019-05-10 | 2021-01-15 | 合肥工业大学 | Interactive vehicle data classification method under ordering support |
| CN110059771A (en) * | 2019-05-10 | 2019-07-26 | 合肥工业大学 | A kind of interactive vehicle data classification method in the case where sequence is supported |
| CN112699206A (en) * | 2021-03-23 | 2021-04-23 | 上海迹寻科技有限公司 | User position and residence analysis method and device |
| CN112699206B (en) * | 2021-03-23 | 2021-07-13 | 上海迹寻科技有限公司 | User position and residence analysis method and device |
| CN114065947A (en) * | 2021-11-15 | 2022-02-18 | 深圳大学 | Data access speculation method and device, storage medium and electronic equipment |
| CN114065947B (en) * | 2021-11-15 | 2022-07-22 | 深圳大学 | Data access speculation method and device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107194412A (en) | A kind of method of processing data, device, equipment and computer-readable storage medium | |
| CN107067003A (en) | Extracting method, device, equipment and the computer-readable storage medium of region of interest border | |
| CN105091890B (en) | Method and apparatus for recommended candidate path | |
| CN107133645B (en) | Method, equipment and storage medium for predicting order cancelling behavior of passenger | |
| CN110378483A (en) | The system and method for training machine learning model being deployed on analog platform | |
| CN104737565A (en) | Method relating to predicting the future state of a mobile device user | |
| CN106875066A (en) | With the Forecasting Methodology of car travel behaviour, device, server and storage medium | |
| CN106971279A (en) | Estimate method, device, equipment and the storage medium of driver's order behavior | |
| CN107038157A (en) | Identification error detection method, device and storage medium based on artificial intelligence | |
| CN109710709A (en) | Interest point data processing method, device, electronic equipment and storage medium | |
| CN107133697A (en) | Estimate method, device, equipment and the storage medium of driver's order wish | |
| US11288533B2 (en) | Systems and methods for identifying a service qualification of a unit of a community | |
| CN106934026A (en) | Trip services mining method, device, equipment and storage medium | |
| CN110097121A (en) | A kind of classification method of driving trace, device, electronic equipment and storage medium | |
| CN111044045A (en) | Navigation method and device based on neural network and terminal equipment | |
| CN106767835A (en) | Localization method and device | |
| CN111522838A (en) | Address similarity calculation method and related device | |
| EP3443482A1 (en) | Classifying entities in digital maps using discrete non-trace positioning data | |
| CN110471999A (en) | Trajectory processing method, apparatus, equipment and medium | |
| US20190171745A1 (en) | Open ended question identification for investigations | |
| US20200286243A1 (en) | Iterative approach for weakly-supervised action localization | |
| CN109635868A (en) | Determination method, apparatus, electronic equipment and the storage medium of barrier classification | |
| US11220269B2 (en) | Fording depth estimation | |
| JP2021047184A (en) | Method, apparatus, device and computer readable storage medium to be used in navigation | |
| US20180357190A1 (en) | Memory sharing for application offload from host processor to integrated sensor hub |
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 |