CN110363304A - Location model construction method, device, computer equipment and storage medium - Google Patents
Location model construction method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110363304A CN110363304A CN201910526952.XA CN201910526952A CN110363304A CN 110363304 A CN110363304 A CN 110363304A CN 201910526952 A CN201910526952 A CN 201910526952A CN 110363304 A CN110363304 A CN 110363304A
- Authority
- CN
- China
- Prior art keywords
- location
- model
- incidence relation
- anchor point
- user
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
This application involves a kind of location model construction method, device, computer equipment and storage mediums.It is related to artificial intelligence field.The described method includes: the incidence relation is under set scene by generating in human-computer interaction process from the incidence relation obtained between user trajectory figure and anchor point in third-party platform；The training sample with positioning mark is generated according to the incidence relation, the user trajectory figure is training sample, is positioning mark with the associated anchor point of the user trajectory figure；Obtain machine learning model predetermined；The training sample is input in the machine learning model and carries out model training, generates location model；The location model is updated into location-based platform.Handmarking's difficulty can be overcome using this method.
Description
Technical field
This application involves field of computer technology, set more particularly to a kind of location model construction method, device, computer
Standby and storage medium.
Background technique
Before constructing supervised learning model, usually we need to do following Data Preparation: from huge sample
Some representative sample datas are selected in set and are marked.In order to make model training accuracy rate with higher, increase
Large sample size is the most common way, and therefore, carrying out artificial mark to great amount of samples will be a huge engineering.Sample is artificial
Mark consumes a large amount of manpower and material resources, becomes the building maximum weakness of supervised learning, the application, which is intended to explore one kind, to be overcome
The location model construction method of handmarking's difficulty.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of location model that can overcome handmarking's difficulty
Construction method, device, computer equipment and storage medium.
A kind of location model construction method, which comprises
The user trajectory figure for positioning mark will be needed to be sent to third-party platform；
From the incidence relation obtained between user trajectory figure and anchor point in third-party platform, the incidence relation is to set
Determine under scene by generating in human-computer interaction process；
The training sample with positioning mark is generated according to the incidence relation, the user trajectory figure is training sample,
It is positioning mark with the associated anchor point of the user trajectory figure；
Obtain machine learning model predetermined；
The training sample is input in the machine learning model and carries out model training, generates location model；
The location model is updated into location-based platform.
In one embodiment, the method also includes:
Receive the real-time position information of user terminal uploads；
User trajectory figure is generated according to the real-time position information；
The user trajectory figure is input in the location model, the location information of the location model output is obtained；
Location based service is provided to the user terminal according to the location information.
In one embodiment, pass is associated between user trajectory figure and anchor point from obtaining in third-party platform described
After system, further includes:
According to the associated attribute tags of the anchor point, classify to the incidence relation；
Multiple groups training sample is generated according to the different types of incidence relation；
Multiple location models are obtained based on the training of training sample described in multiple groups, each location model is set for identification
The anchor point of attribute.
In one embodiment, the third-party platform is gaming platform；It is described that user's rail is obtained from third-party platform
Incidence relation between mark figure and anchor point, further includes:
The gaming platform generates game example, and monitors the game example, to grasp from the human-computer interaction of game player
The positioning markup information of the user trajectory figure is obtained in work；
The positioning markup information is obtained from the gaming platform, and the user is generated according to the positioning markup information
Incidence relation between trajectory diagram and anchor point.
In one embodiment, described to be associated with pass between user trajectory figure and anchor point from obtaining in third-party platform
System, further includes:
Generate handmarking's task；
Handmarking's task is published at least one third-party platform in the form of the H5 page or web page interlinkage；
Receive the label result of the third-party platform feedback.
A kind of location model construction device, described device include:
Incidence relation obtains module, for the user trajectory figure for needing to position mark to be sent to third-party platform, from the
The incidence relation between user trajectory figure and anchor point is obtained in tripartite's platform, the incidence relation is passed through under set scene
It is generated in human-computer interaction process；
Training sample generation module, it is described for generating the training sample with positioning mark according to the incidence relation
User trajectory figure is training sample, is positioning mark with the associated anchor point of the user trajectory figure；
Model obtains module, for obtaining machine learning model predetermined；
Model training module carries out model training for the training sample to be input in the machine learning model,
Generate location model；
Model modification module, for updating the location model into location-based platform.
In one embodiment, described device further include: location based service module, for receiving user terminal uploads
Real-time position information；User trajectory figure is generated according to the real-time position information；The user trajectory figure is input to described
In location model, the location information of the location model output is obtained；It is provided according to the location information to the user terminal
Location based service.
In one embodiment, described device further includes categorization module, for according to the associated attribute mark of the anchor point
Label, classify to the incidence relation；Multiple groups training sample is generated according to the different types of incidence relation；Based on multiple groups
The training sample training obtains multiple location models, and each location model sets the positioning of attribute for identification
Point.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
The step of device realizes method described above when executing the computer program.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
The step of method described above is realized when row.
Above-mentioned location model construction method, device, computer equipment and storage medium, by setting field in third-party platform
Human-computer interaction under scape obtains the incidence relation between user trajectory figure and anchor point, using user trajectory figure as training sample,
With the associated anchor point of user trajectory figure as positioning mark, the training sample with positioning mark is automatically generated, without consumption
Special manpower is manually marked, and is overcome handmarking's bring low efficiency, the disadvantages of time-consuming, is reduced and be modeled as
This, shortens the modeling period of supervised learning, i.e., overcomes handmarking's difficulty when constructing location model.
Detailed description of the invention
Fig. 1 is the application scenario diagram of location model construction method in one embodiment；
Fig. 2 is the flow diagram of location model construction method in one embodiment；
Fig. 3 is to provide a user the signal of process involved in location based service based on location model in one embodiment
Figure；
Fig. 4 is the user trajectory figure of user in another embodiment；
Fig. 5 is the structural block diagram of location model construction device in one embodiment；
Fig. 6 is the internal structure chart of computer equipment in one embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
Location model construction method provided by the present application, can be applied in application environment as shown in Figure 1.Application environment
In include user terminal 102, LBS server 104 (Location Based Service, location based service device) and third
Fang Pingtai 106, wherein third-party platform 106 includes third-party server 106A and the user interacted with third-party server 106A
Terminal 106B.User terminal 102 is communicated by network with LBS server 104.LBS server 104 and third-party platform 106 can
To be communicated by network.LBS server 104 is between the user trajectory figure and anchor point that third-party platform obtains magnanimity
Incidence relation, and using these incidence relations as training sample training location model, LBS server 104 be based on location model to
User terminal 102 provides location based service.
User terminal in above-mentioned terminal 102 and third-party platform 106 can be, but not limited to be various personal computers,
Laptop, smart phone, tablet computer and portable wearable device, in LBS server 104 and third-party platform 106
Server-side can be realized with the server cluster of independent server either multiple servers composition.
In one embodiment, as shown in Fig. 2, providing a kind of location model construction method, it is applied to Fig. 1 in this way
In LBS server for be illustrated, comprising the following steps:
Step 202, the user trajectory figure for positioning mark will be needed to be sent to third-party platform.
Step 204, from the incidence relation obtained between user trajectory figure and anchor point in third-party platform, incidence relation is
By generating in human-computer interaction process under set scene.
Third-party platform can be gaming platform or with human-computer interaction modules such as game module, entertaining task modules
Other application platform.Third-party platform issues user trajectory figure, obtains terminal side user in user's rail by human-computer interaction module
The anchor point selected on mark figure, and then establish the incidence relation between user trajectory figure and anchor point.Namely third-party platform
The incidence relation between the user trajectory figure and anchor point of big data quantity is generated by the human-computer interaction process under set scene.
Anchor point is the location point with certain particular community determined from the numerous anchor points of user trajectory figure, is such as positioned
Point can be house address, work unit etc..
Step 206, the training sample with positioning mark being generated according to incidence relation, user trajectory figure is training sample,
It is positioning mark with the associated anchor point of user trajectory figure.
Anchor point mark has been carried out to user trajectory figure by human-computer interaction module in third-party platform, has generated user
Incidence relation between trajectory diagram and anchor point.LBS server, will using the user trajectory figure in incidence relation as training sample
Positioning with the associated anchor point of user trajectory figure as training sample marks.User trajectory figure is as the input to training pattern
Object has carried out monitor model as output (supervisory signals of model) desired by model with the associated anchor point of user trajectory figure
Training.
Further, training sample is generated according to a part of incidence relation, is generated and is tested according to another part incidence relation
The generating mode of sample, training sample and test sample is identical.
Step 208, machine learning model predetermined is obtained.
Machine learning model in this step is Supervised machine learning model, preassigns machine learning model, when need
When generating location model, the Supervised machine learning model pre-defined is directly acquired.Preassigned machine learning mould
Type can be support vector machines (Support Vector Machines), linear regression (linear regression), logic
Return (logistic regression), naive Bayesian (naive Bayes) linear discriminant analysis (linear
Discriminant analysis), decision tree (decision trees) etc..
Step 210, training sample is input in machine learning model and carries out model training, generate location model.
After determining machine learning model, the training sample with positioning mark is input in machine learning model, is determined
Model parameter, it is location model that model parameter, which is input to the training pattern obtained after machine learning model,.To location model
User trajectory figure is inputted, location model will predict the anchor point with certain particular community in the user trajectory figure.
Step 212, location model is updated into location-based platform.
There is location based service module, by location model deployment/update to local positions of generation in LBS server
Service module in or LBS server send other for location model and provide in the platform of location based service, such as navigate
Platform provides a user various location based services so that platform is based on location model.Such as based on model output based on residence
Positioning, provide the navigation routine gone home to user terminal.
The location model construction method of the present embodiment is set without being manually labeled to sample by third-party platform
Determine the human-computer interaction under scene and obtain the markup information of sample, this markup information collection mode is without consuming special manpower object
Power reduces modeling cost, shortens the modeling period of supervised learning.
In one embodiment, it after generating location model, can be disposed in the platform arbitrarily with location based service
The location model, and service is provided to user terminal based on the location model.LBS server in Fig. 1 is to be deployed with to be based on
The server of the service of position, as shown in figure 3, illustrating how to mention based on location model to user terminal by taking LBS server as an example
For location based service, specifically comprise the following steps:
Step 302, LBS server receives the location information of user terminal real-time report, and raw according to real-time position information
At the user trajectory figure of the user terminal.
Real-time position information in LBS server acquisition user terminal is interior or one month nearly at nearly one week, and according to every
It real-time position information generates a user trajectory figure, and then obtains a plurality of user trajectory figure in one week or one month.
In one embodiment, LBS server calculates the concentration degree between a plurality of user trajectory figure, and is gone according to concentration degree
It removes and the lower user trajectory figure of other trajectory diagram concentration degrees, wherein the concentration degree between user trajectory figure is according to user
The one or more that one or more in trajectory diagram represents in location point and another user trajectory figure represents between location point
Distance depending on, represent that the distance between location point is bigger, and concentration degree is lower.
As shown in figure 4, four user trajectory figures in figure, the wherein user trajectory figure 4. collection with user trajectory figure 1. 2. 3.
Moderate is lower, therefore, when carrying out setting attribute positioning point prediction, first according to concentration degree, gets rid of user trajectory figure 4., only will
The user trajectory figure being 1. 2. 3. made of user trajectory figure is input in location model.
Step 304, user trajectory figure is input in location model, obtains the location information of location model output.
Analysis prediction is carried out using user trajectory figure of the location model built in advance to input, obtains the user trajectory
Scheme the corresponding anchor point with setting attributive character, such as shelter, company, the public place of entertainment place often gone.
Step 306, location based service is provided to user terminal according to location information.
The anchor point with setting attribute that LBS server predicts location model is sent to user terminal, monitoring users
Confirmation operation the anchor point is added in user information if user determines anchor point.LBS server can be according to user
Associated anchor point, provides a user location based service.It can according to the location based service that anchor point provides a user
To be to push the best navigation routine gone home to user in quitting time point, the navigation for going to company is pushed to user in the work hours
Route.Customer consumption level is either determined according to the home address and company position of user, is provided a user and customer consumption
The product information being on close level, such as financial product.
In the present embodiment, the recommendation of personalized information of the movement track based on user is realized, compare traditional collection
The various consumer records of user, various web page browsing records, the real-time position information for collecting user are simpler quick.
In one embodiment, location model is generated according to the incidence relation between user trajectory figure and anchor point, also wrapped
Include: the incidence relation between the user trajectory figure and anchor point that third-party platform obtains, the anchor point in incidence relation have
Attribute tags, wherein attribute tags include shelter label, work unit's label, school's label and public place of entertainment label etc..LBS clothes
Device be engaged according to the attribute tags of anchor point, incidence relation is classified, the anchor point with same attribute tags is divided into one
Class.It is the incidence relation of shelter, anchor point is the incidence relation of work unit, positioning that incidence relation after division, which includes anchor point,
Point is the incidence relation of school, and anchor point is the incidence relation etc. of public place of entertainment.
After classification, LBS server generates multiple groups training sample according to different types of incidence relation, such as generates corresponding shelter
Training sample, the training sample of corresponding work unit, the training sample of corresponding school and the training sample of corresponding public place of entertainment
This etc..Then location model is generated further according to every group of group training sample, the positioning of each location model respective attributes for identification
Point.The location model that such as training sample based on shelter the obtains shelter position in user trajectory figure for identification, is based on work
The location model that the training sample of unit the obtains work unit in user trajectory figure for identification.
In one embodiment, the method for position point prediction being carried out by the location model of classifying type are as follows: obtain user's rail
Mark figure determines anchor point attribute to be positioned, searches location model corresponding with anchor point attribute, user trajectory figure is input to
In the location model of lookup, the positioning result of model is obtained.If positioning properties to be positioned be it is multiple, search multiple positioning moulds
User trajectory figure is separately input into multiple location models by type, obtains multiple positioning results to get multiple and different attributes are arrived
Anchor point.
In the present embodiment, classified by the attribute of anchor point to incidence relation, is based on sorted incidence relation number
According to the location model that the training sample of generation constructs, the anchor point for setting type can be identified more accurately, improved
The accuracy of model orientation.
In one embodiment, the third-party platform in step 202 is gaming platform, and the construction method of location model is specific
Are as follows: when LBS server provides location based service to user terminal, if desired user trajectory figure is marked, LBS service
Device will need the user trajectory figure for positioning mark to be sent to gaming platform, and user trajectory figure is generated game example by gaming platform,
And issue the game example.Gaming platform monitors the human-machine operation of the game player based on the game example, and from human-computer interaction
The markup information of user trajectory figure in game example is obtained in operation, markup information is sent to LBS server by gaming platform,
LBS server provides a user location based service or gaming platform for user trajectory figure and positioning according to markup information
The incidence relation of point is sent to LBS server, and LBS server provides a user location based service according to incidence relation.
In another embodiment, when LBS server provides location based service to user terminal, if there is handmarking
Task (is not limited only to the handmarking to user trajectory figure, can also be other label tasks, such as label target face),
Handmarking's task is sent to third-party platform by LBS server, and third-party platform obtains label knot by man-machine interactive operation
Fruit, third-party platform will mark result to be sent to LBS server, and LBS server is provided a user according to label result based on position
The service set.
While providing location based service, LBS server can constantly obtain user's rail from third-party platform
Incidence relation between mark figure and anchor point, when incidence relation data reach setting quantitative value, LBS server is based on association and closes
System generates location model and directly inputs user trajectory figure when LBS server needs to be labeled user trajectory figure again
It can be obtained markup information into location model.
In another embodiment, it is if desired artificial to mark when LBS server provides location based service to user terminal
It clocking, handmarking here is suitable for the handmarking of various data, it is not limited only to the handmarking to user trajectory figure,
LBS server generates handmarking's task, and handmarking's task is published at least one third party in the form of small routine
Platform, small routine obtain label as a result, and label result is fed back into LBS server, LBS server according to label result to
Family provides location based service, or generates training sample according to label result, generates automatic identification mark based on training sample
The artificial intelligence model of note provides location based service to user terminal based on artificial intelligence model.
LBS server is when providing location based service, if be related to handmarking's process to trajectory diagram, tradition
Way be to generate handmarking's task, by location based service could be executed after handmarking under line, many times, manually
The workload of label task is all very big.In the present embodiment, handmarking's task is pushed to gaming platform, gaming platform is based on
Handmarking's task generates game example, need to only introduce trajectory diagram in generating game example, and adds specific label
Method.Since server is according to trajectory diagram and specific labeling method, game is automatically generated under existing scene of game
Numerous game players are marked in example, substantially increase the efficiency of label, so that LBS server can be based on these marks
Information provides corresponding location based service.
It should be understood that although each step in the flow chart of Fig. 2-3 is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, these steps can execute in other order.Moreover, at least one in Fig. 2-3
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, the execution sequence in these sub-steps or stage is also not necessarily successively
It carries out, but can be at least part of the sub-step or stage of other steps or other steps in turn or alternately
It executes.
In one embodiment, as shown in figure 5, providing a kind of location model construction device, which includes:
Incidence relation obtains module 502, for from the pass obtained between user trajectory figure and anchor point in third-party platform
Connection relationship, the incidence relation are under set scene by generating in human-computer interaction process.
Training sample generation module 504, for generating the training sample with positioning mark according to the incidence relation,
Described in user trajectory figure be the training sample, be that the training sample is determined with the associated anchor point of user trajectory figure
Position mark.
Model obtains module 506, for obtaining machine learning model predetermined.
Model training module 508 carries out model training for the training sample to be input to the machine learning model,
Generate location model.
Model modification module 510, for updating the location model into location based service.
In one embodiment, location model construction device further include: location based service device module is used for receiving
The real-time position information that family terminal uploads；User trajectory figure is generated according to the real-time position information；By the user trajectory figure
It is input in the location model, obtains the location information of the location model output；According to the location information to the use
Family terminal provides location based service.
In one embodiment, location model construction device further includes categorization module, for being associated with according to the anchor point
Attribute tags, classify to the incidence relation；Multiple groups training sample is generated according to the different types of incidence relation；
Multiple location models are obtained based on the training of training sample described in multiple groups, each location model sets determining for attribute for identification
Site.
In one embodiment, the third-party platform is gaming platform；Incidence relation obtain module 502, be also used to by
The user trajectory figure for needing to position mark is sent to the gaming platform；Wherein, the gaming platform generates game example,
And the game example is monitored, the positioning mark letter of the user trajectory figure is obtained from the man-machine interactive operation of game player
Breath；The positioning markup information is obtained from the gaming platform, and the user trajectory is generated according to the positioning markup information
Incidence relation between figure and anchor point.
In one embodiment, incidence relation obtains module 502, is also used to generate handmarking's task；By handmarking
Task is published at least one platform in the form of small routine；Receive the label result of the platform feedback.
Specific about location model construction device limits the limit that may refer to above for location model construction method
Fixed, details are not described herein.Modules in above-mentioned location model construction device can fully or partially through software, hardware and its
Combination is to realize.Above-mentioned each module can be embedded in the form of hardware or independently of in the processor in computer equipment, can also be with
It is stored in the memory in computer equipment in a software form, in order to which processor calls the above modules of execution corresponding
Operation.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 6.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment is for storing machine learning model data.The network interface of the computer equipment is used for and external terminal
It is communicated by network connection.To realize a kind of location model construction method when the computer program is executed by processor.
It will be understood by those skilled in the art that structure shown in Fig. 6, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, which is stored with
Computer program, the processor perform the steps of the acquisition user trajectory figure from third-party platform when executing computer program
Incidence relation between anchor point, the incidence relation are by generating in the human-computer interaction process under set scene；Root
The training sample with positioning mark is generated according to the incidence relation, wherein the user trajectory figure is the training sample, with
The associated anchor point of user trajectory figure is that the positioning of the training sample marks；Obtain machine learning model；By the instruction
Practice sample and be input to the machine learning model progress model training, generates location model；The location model is updated to base
In the service of position.
In one embodiment, the reality for obtaining user terminal is also performed the steps of when processor executes computer program
When location information；User trajectory figure is generated according to the real-time position information；The user trajectory figure is input to the positioning
In model, the location information of the location model output is obtained；It is based on according to the location information to user terminal offer
The service of position.
In one embodiment, it also performs the steps of when processor executes computer program and is closed according to the anchor point
The attribute tags of connection classify to the incidence relation；Multiple groups training sample is generated according to the different types of incidence relation
This；The training of training sample described in multiple groups obtains multiple location models, and each location model sets determining for attribute for identification
Site.
In one embodiment, the third-party platform is gaming platform；Processor is described from third-party platform in execution
When the middle incidence relation step obtained between user trajectory figure and anchor point, also executing the following steps: will need to position mark
The user trajectory figure is sent to the gaming platform；Wherein, the gaming platform generates game example, and monitors the game
Example obtains the markup information of the user trajectory figure from the man-machine interactive operation of game player；From the gaming platform
Obtain the markup information of the user trajectory figure.
In one embodiment, processor execute described from third-party platform and obtain user trajectory figure and anchor point it
Between incidence relation step when, also execute the following steps: generate handmarking's task；By handmarking's task with the shape of small routine
Formula is published at least one platform；Receive the label result of the platform feedback.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program is performed the steps of when being executed by processor from the pass obtained between user trajectory figure and anchor point in third-party platform
Connection relationship, the incidence relation are by generating in the human-computer interaction process under set scene；It is raw according to the incidence relation
At the training sample with positioning mark, wherein the user trajectory figure is the training sample, closed with the user trajectory figure
The anchor point of connection is that the positioning of the training sample marks；Obtain machine learning model；The training sample is input to described
Machine learning model carries out model training, generates location model；The location model is updated into location based service.
In one embodiment, it is also performed the steps of when computer program is executed by processor and obtains user terminal
Real-time position information；User trajectory figure is generated according to the real-time position information；It is described fixed that the user trajectory figure is input to
In bit model, the location information of the location model output is obtained；Base is provided to the user terminal according to the location information
Service in position.
In one embodiment, it also performs the steps of when computer program is executed by processor according to the anchor point
Associated attribute tags classify to the incidence relation；Multiple groups training is generated according to the different types of incidence relation
Sample；The training of training sample described in multiple groups obtains multiple location models, and each location model sets attribute for identification
Anchor point.
In one embodiment, the third-party platform is gaming platform；It is described that user's rail is obtained from third-party platform
When incidence relation between mark figure and anchor point is executed by processor, also performing the steps of will need to position described in mark
User trajectory figure is sent to the gaming platform；Wherein, the gaming platform generates game example, and it is real to monitor the game
Example, obtains the markup information of the user trajectory figure from the man-machine interactive operation of game player；It is obtained from the gaming platform
Take the markup information of the user trajectory figure.
In one embodiment, described from the incidence relation obtained in third-party platform between user trajectory figure and anchor point
When being executed by processor, also performs the steps of and generate handmarking's task；Handmarking's task is sent out in the form of small routine
Cloth is at least one platform；Receive the label result of the platform feedback.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of location model construction method, which comprises
The user trajectory figure for positioning mark will be needed to be sent to third-party platform；
From the incidence relation obtained between user trajectory figure and anchor point in third-party platform, the incidence relation is in setting field
By generating in human-computer interaction process under scape；
The training sample with positioning mark is generated according to the incidence relation, the user trajectory figure is training sample, with institute
The associated anchor point of user trajectory figure is stated as positioning mark；
Obtain machine learning model predetermined；
The training sample is input in the machine learning model and carries out model training, generates location model；
The location model is updated into location-based platform.
2. the method according to claim 1, wherein the method also includes:
Receive the real-time position information of user terminal uploads；
User trajectory figure is generated according to the real-time position information；
The user trajectory figure is input in the location model, the location information of the location model output is obtained；
Location based service is provided to the user terminal according to the location information.
3. the method according to claim 1, wherein it is described from third-party platform obtain user trajectory figure with
After incidence relation between anchor point, further includes:
According to the associated attribute tags of the anchor point, classify to the incidence relation；
Multiple groups training sample is generated according to the different types of incidence relation；
Multiple location models are obtained based on the training of training sample described in multiple groups, each location model sets attribute for identification
The anchor point.
4. according to claim 1 to method described in 3 any one, which is characterized in that the third-party platform is gaming platform；
The incidence relation from third-party platform between acquisition user trajectory figure and anchor point, further includes:
The gaming platform generates game example, and monitors the game example, from the man-machine interactive operation of game player
Obtain the positioning markup information of the user trajectory figure；
The positioning markup information is obtained from the gaming platform, and the user trajectory is generated according to the positioning markup information
Incidence relation between figure and anchor point.
5. method according to claim 1 to 3, which is characterized in that described to obtain user from third-party platform
Incidence relation between trajectory diagram and anchor point, further includes:
Generate handmarking's task；
Handmarking's task is published at least one third-party platform in the form of the H5 page or web page interlinkage；
Receive the label result of the third-party platform feedback.
6. a kind of location model construction device, which is characterized in that described device includes:
Incidence relation obtains module, for the user trajectory figure for needing to position mark to be sent to third-party platform, from third party
The incidence relation between user trajectory figure and anchor point is obtained in platform, the incidence relation is under set scene by man-machine
It is generated in interactive process；
Training sample generation module, for generating the training sample with positioning mark, the user according to the incidence relation
Trajectory diagram is training sample, is positioning mark with the associated anchor point of the user trajectory figure；
Model obtains module, for obtaining machine learning model predetermined；
Model training module carries out model training for the training sample to be input in the machine learning model, generates
Location model；
Model modification module, for updating the location model into location-based platform.
7. device according to claim 6, which is characterized in that described device further include: location based service module is used
In the real-time position information for receiving user terminal uploads；User trajectory figure is generated according to the real-time position information；By the use
Family trajectory diagram is input in the location model, obtains the location information of the location model output；According to the location information
Location based service is provided to the user terminal.
8. device according to claim 6, which is characterized in that described device further includes categorization module, for according to
The associated attribute tags of anchor point, classify to the incidence relation；It is generated according to the different types of incidence relation more
Group training sample；Multiple location models are obtained based on the training of training sample described in multiple groups, each location model is for identification
Set the anchor point of attribute.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 5 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 5 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910526952.XA CN110363304A (en) | 2019-06-18 | 2019-06-18 | Location model construction method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910526952.XA CN110363304A (en) | 2019-06-18 | 2019-06-18 | Location model construction method, device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110363304A true CN110363304A (en) | 2019-10-22 |
Family
ID=68216304
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910526952.XA Pending CN110363304A (en) | 2019-06-18 | 2019-06-18 | Location model construction method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110363304A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018077745A1 (en) * | 2016-10-24 | 2018-05-03 | Bayerische Motoren Werke Aktiengesellschaft | Method for analysing object labels in images on the basis of models |
| CN108074414A (en) * | 2017-12-19 | 2018-05-25 | 广州小鹏汽车科技有限公司 | It is a kind of that path traffic information based reminding method and system are often walked based on user behavior |
| CN108108455A (en) * | 2017-12-28 | 2018-06-01 | 广东欧珀移动通信有限公司 | Method for pushing, device, storage medium and the electronic equipment of destination |
| CN109739926A (en) * | 2019-01-09 | 2019-05-10 | 南京航空航天大学 | A kind of mobile object destination prediction technique based on convolutional neural networks |
-
2019
- 2019-06-18 CN CN201910526952.XA patent/CN110363304A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018077745A1 (en) * | 2016-10-24 | 2018-05-03 | Bayerische Motoren Werke Aktiengesellschaft | Method for analysing object labels in images on the basis of models |
| CN108074414A (en) * | 2017-12-19 | 2018-05-25 | 广州小鹏汽车科技有限公司 | It is a kind of that path traffic information based reminding method and system are often walked based on user behavior |
| CN108108455A (en) * | 2017-12-28 | 2018-06-01 | 广东欧珀移动通信有限公司 | Method for pushing, device, storage medium and the electronic equipment of destination |
| CN109739926A (en) * | 2019-01-09 | 2019-05-10 | 南京航空航天大学 | A kind of mobile object destination prediction technique based on convolutional neural networks |
Non-Patent Citations (4)
| Title |
|---|
| 孙欢: "众包标注的学习算法研究", 《中国优秀博硕士学位论文全文数据库（硕士）信息科技辑》 * |
| 孙欢: "众包标注的学习算法研究", 《中国优秀博硕士学位论文全文数据库（硕士）信息科技辑》, no. 03, 15 March 2016 (2016-03-15), pages 140 - 143 * |
| 柯扬斌: "基于深度学习和回归型分类模型的出行目的地预测", 《中国优秀博硕士学位论文全文数据库（硕士）基础科学辑》 * |
| 柯扬斌: "基于深度学习和回归型分类模型的出行目的地预测", 《中国优秀博硕士学位论文全文数据库（硕士）基础科学辑》, no. 12, 15 January 2019 (2019-01-15), pages 008 - 76 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10958748B2 (en) | Resource push method and apparatus | |
| CN112632385B (en) | Course recommendation method, course recommendation device, computer equipment and medium | |
| Metan et al. | Real time selection of scheduling rules and knowledge extraction via dynamically controlled data mining | |
| CN110366734A (en) | Optimization neural network framework | |
| CN109684554A (en) | The determination method and news push method of the potential user of news | |
| CN109726105A (en) | Test data building method, device, equipment and storage medium | |
| CN112256537B (en) | Model running state display method and device, computer equipment and storage medium | |
| CN111914178A (en) | Information recommendation method and device based on artificial intelligence, electronic equipment and storage medium | |
| CN109033277A (en) | Class brain system, method, equipment and storage medium based on machine learning | |
| CN108062377A (en) | The foundation of label picture collection, definite method, apparatus, equipment and the medium of label | |
| Nagy et al. | Ten questions concerning reinforcement learning for building energy management | |
| CN106874204A (en) | Automatic test method for customizing and custom-built system | |
| CN110471585A (en) | Function of application icon methods of exhibiting, device and computer equipment | |
| CN114546365B (en) | Flow visualization modeling method, server, computer system and medium | |
| CN110363304A (en) | Location model construction method, device, computer equipment and storage medium | |
| US20230229691A1 (en) | Methods and systems for prediction of a description and icon for a given field name | |
| CN116578774A (en) | Method, device, computer equipment and storage medium for pre-estimated sorting | |
| CN109872188A (en) | Assets suggest generation method, device, computer equipment and storage medium | |
| CN116204709A (en) | Data processing method and related device | |
| US20230195742A1 (en) | Time series prediction method for graph structure data | |
| CN110414690A (en) | The method and device of prediction is executed using machine learning model | |
| CN115883172A (en) | Anomaly monitoring method and device, computer equipment and storage medium | |
| Getz et al. | Simulation and analysis of animal movement paths using numerus model builder | |
| EP4172824A1 (en) | Automated machine learning: a unified, customizable, and extensible system | |
| Thabet et al. | Towards intelligent serious games: deep knowledge tracing with hybrid prediction models |
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 |