CN106899668B - Information Push Service processing method in car networking - Google Patents
Information Push Service processing method in car networking Download PDFInfo
- Publication number
- CN106899668B CN106899668B CN201710099060.7A CN201710099060A CN106899668B CN 106899668 B CN106899668 B CN 106899668B CN 201710099060 A CN201710099060 A CN 201710099060A CN 106899668 B CN106899668 B CN 106899668B
- Authority
- CN
- China
- Prior art keywords
- individual
- vehicle
- car networking
- degree
- information
- 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.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
Abstract
The invention proposes Information Push Service processing methods in a kind of novel car networking.In the art, it is first depending on vehicle connection individual attribute feature and car networking is divided into several sub- car networkings, so that the car networking overall situation degree of correlation loss reduction of segmentation front and back;Then, the quantity based on idle server in system, optimization distributes and handles sub- car networking on these servers, thus the workload of balanced idle server;Individual is joined for each vehicle in car networking sub- on server, the highest part of its interest-degree is obtained and is pushed information, and in whole car networking, carries out global adaptation to information is pushed, pushes effect to reach optimal information.The present invention can significantly improve the accuracy of Information Push Service, diversity and efficiency in car networking.
Description
Technical field
The present invention relates to Information Push Service processing techniques in a kind of car networking, belong to intelligent transportation and intersect with information technology
Field.
Background technique
In Baidupedia, alliance, Chinese Internet of Things school-run enterprise is about car networking (IoT:Internet of Vehicles)
Is defined as: car networking is the huge Internet being made of information such as vehicle location, speed and routes.Pass through GPS, RFID, biography
Devices, the vehicles such as sensor, camera image processing can complete the acquisition of itself environment and status information;Pass through internet skill
Art, all vehicles can be by the various information Transmission Convergences of itself to central processing unit;By computer technology, these are a large amount of
The information of vehicle can be analyzed and processed, to calculate the best route of different vehicle, report without delay road conditions and arrange letter
The signal lamp period.
Information Push Service is a key technology of car networking, its main purpose is for each vehicle in car networking
Join individual v, the upper interested information of v of other vehicles connection individual be pushed to v, v is enabled to share the information of other vehicle connection individuals,
And it is not limited to the information oneself possessed.Here information includes diversified forms, can be music, video, text, picture
Deng.
Currently, Information Push Service is mainly based upon the information advancing technique of car networking structure feature in car networking.The skill
Art mainly utilizes the structure feature of the connection individual of vehicle in car networking, by calculate such as mutually cut-off jointly join individual, Jaccard and
The Measure Indexes as similarity such as Adamic/Adar.Individual v is joined for each vehicle in car networking, we are a joining with vehicle
The relevant vehicle connection individual collections of body v are set as R (v), in addition, if vehicle connection individual v is interested in information λ, interest-degree r
(v, λ)=1, otherwise r (v, λ)=0.It is not difficult to find out that v and in R (v) different vehicles connection individual v ' between the degree of correlation and to certain information
Interest-degree be different.The degree of correlation between v and v ' is calculated by Jaccard coefficient method, it may be assumed that
And the Interest Similarity between v and v ' can calculate the interested letter of vehicle connection individual by Jaccard coefficient method
How many common information acquisition in the set of breath, it may be assumed that
Wherein N (v) and N (v ') respectively indicates information aggregate interested to vehicle connection individual v and v '.To which we can be with
Different method for measuring similarity is combined together by different weight factors, is returned by fitting and determines weight factor
Optimal value wt (v, v '), then the calculating as similarity between final vehicle connection individual, that is, indicate are as follows:
fsλ=∑v'∈R(v)r(v',λ)wt(v,v')。
Finally, will be by fsλAppropriate number of information is pushed to vehicle connection individual v by value sequence.
We have found that in existing car networking Information Push Service processing technique Information Push Service accuracy, more
All there is deficiencies for sample and efficiency etc., to influence the quality of Information Push Service in car networking.
Summary of the invention
The purpose of the present invention is to overcome the deficiency in the prior art, proposes Information Push Service processing skill in a kind of car networking
Art, workflow are mainly made of 4 steps:
1, car networking G, the number k for needing pushed information and traversal threshold value t is obtained from input terminal (when pushed information, to need
The vehicle of access joins individual amount), and car networking is divided by d sub- car networking G based on vehicle connection individual attribute feature1,G2,…,
Gd, so that the sub- car networking set { G after segmentation1,G2,…,GdAnd segmentation before car networking G between the global degree of correlation loss most
It is small.
2, according to the idle server quantity s in system, by d obtained in step 1 sub- car networking G1,G2,…,GdPoint
It is fitted in these idle servers, this s platform idle server starts the sub- car networking above parallel processing respectively later.
3, for car networking G on serverxEach vehicle in (1≤x≤d) joins individual v, obtains all not on v
The set of information compositionThen, forIn each Pre-Evaluation information λ, calculate v to the interest-degree of λ.WhenIn
The interest-degree of all Pre-Evaluation information calculates finish after, return the highest k information λ of interest-degree1,λ2,…,λk。
4, individual v is joined for each vehicle in car networking G, obtains the vehicle directly related with v and join individual collections M, obtains simultaneously
Vehicle connection individual collections R directly related with v in sub- car networking is taken, specified quantity l and v phase are then chosen in set M-R
The vehicle of pass joins individual, joins individual for selected vehicle, obtains the highest k information of its interest-degree.When all selected vehicles join
The acquisition of information of individual finishes, and is incorporated in part pushed information obtained in step 3, obtains the highest k letter of final interest-degree
It ceases and sends vehicle connection individual v to.
The invention has the following advantages that
1, the present invention can effectively improve the diversity of car networking information push by the extreme saturation to vehicle connection individual.
2, the present invention can be mentioned significantly by the seamless combination to vehicle connection individual local message push and global information push
The coverage rate of high car networking information push.
3, the present invention is based on the car networking segmentations of minimum global degree of correlation loss, and then parallelization handles the side of sub- car networking
Formula can significantly reduce the time overhead of car networking information push.
Detailed description of the invention
Work flow diagram Fig. 1 of the invention
Fig. 2 calculates vehicle connection individual v to the process flow diagram of Pre-Evaluation information inf interest-degree
Specific embodiment
The invention proposes Information Push Service processing techniques in a kind of car networking, are first depending on vehicle connection individual attribute feature
Car networking is divided into several sub- car networkings, so that the car networking overall situation degree of correlation loss reduction of segmentation front and back;Then, it is based on
The quantity of idle server in system, optimization distributes and handles sub- car networking on these servers, thus balanced leisure service
The workload of device;Individual is joined for each vehicle in car networking sub- on server, the highest part of its interest-degree is obtained and is pushed away
It delivers letters breath, and in whole car networking, carries out global adaptation to information is pushed, push effect to reach optimal information.
The present invention can significantly improve the accuracy of Information Push Service, diversity and efficiency in car networking.Workflow such as Fig. 1 institute
Show:
Step 1, from input terminal obtain car networking G, need pushed information number k and traversal threshold value t (when pushed information,
The vehicle connection individual amount for needing to access), and car networking is divided by d sub- car networking G based on vehicle connection individual attribute feature1,
G2,…,Gd, so that the sub- car networking set { G after segmentation1,G2,…,GdAnd segmentation before car networking G between the global degree of correlation damage
It loses minimum.
Step 2, according to the idle server quantity s in system, by d obtained in step 1 sub- car networking G1,G2,…,
GdIt is assigned in these idle servers, this s platform idle server starts the sub- car networking above parallel processing respectively later.
Step 3, for car networking G on serverxEach vehicle in (1≤x≤d) joins individual v, obtains not on v
The set of all information compositionsThen, forIn each Pre-Evaluation information λ, calculate v to the interest-degree of λ.WhenIn all Pre-Evaluation information interest-degree calculate and finish after, return to the highest k information λ of interest-degree1,λ2,…,λk。
Step 4 joins individual v for each vehicle in car networking G, obtains the vehicle directly related with v and joins individual collections M, together
When obtain the vehicle directly related with v in sub- car networking and join individual collections R, then chosen in set M-R specified quantity l and
The relevant vehicle connection individual of v, joins individual for selected vehicle, obtains the highest k information of its interest-degree.When all selected vehicles
The acquisition of information of connection individual finishes, and the comprehensive local pushed information obtained in step 3 obtains final interest-degree highest k
Information simultaneously sends vehicle connection individual v to.
Each step is described in detail below:
One '
In step 1 of the invention, it includes three parts that vehicle, which joins individual attribute feature:
1) the statistics attributive character of vehicle connection individual, the statistics attributive character including car owner: age, gender, native place, duty
Industry, previous graduate college, work unit, obtains the driving license time etc. at income;The statistics attributive character of vehicle: the brand of vehicle, model, vehicle
Age, mileage number, number, maintenance frequency violating the regulations etc..
2) interest of vehicle connection individual likes attributive character, including information interested to car owner, pushed and gives other car owners
Information, information format interested to subject categories, car owner interested to car owner etc..
3) the physical location attributive character of vehicle connection individual: including the city where vehicle, zone of action, postcode, average vehicle
Speed, running time, even more accurate geographical location attributive character etc..
In addition, the car networking G in step 1 is indicated with graph structure, i.e. G=(V, E, W), wherein V is all vehicle connection in G
The set of body composition, E are the set that correlativity forms between vehicle connection is individual, and W is the set of degree of correlation composition between vehicle connection individual, E
In each correlativity there is a degree of correlation in W to be corresponding to it.Individual v is joined for any two vehicle in car networking G1
And v2If they have correlativity, i.e. (v1,v2) ∈ E, then their degree of correlation W (v1,v2) calculate as follows,
If without correlativity, the degree of correlation 0:
Step 1-1, z attributive character a for joining individual relatedness computation for vehicle in car networking G is obtained1,a2,…,az, into
And obtain corresponding with G | V | × z vehicle joins individual-attributive character matrix:
Wherein vi(aj) indicate that vehicle joins individual viIn j-th of attributive character ajOn value, 1≤i≤| V |, 1≤j≤z.
Step 1-2, individual v is joined for vehicle each in G, calculates its attributive character mean valueAnd by matrix
VA calculates covariance matrix corresponding with it:
Step 1-3, vehicle is joined into individual v1Z attributive character value be organized into one-dimensional vector and its transposition respectively,Vehicle is joined into individual v2Z attributive character value point
It is not organized into one-dimensional vector and its transposition,
Step 1-4, v is calculated1And v2Between the degree of correlation:
After the relatedness computation between vehicles all in G connection individual finishes, in order to which the overall situation for dividing front and back car networking is related
Loss reduction is spent, we are based on following steps and G is divided into d sub- car networkings, and in the specific implementation, d takes in 10,15 and 20
One number:
(1, using k-means (k- mean value) clustering algorithm G is divided into d sub- car networking class IG={ G1,G2,…,Gd,
And calculate the global degree of correlation of IG:
(2, following operation is repeated to sub- car networking class IGIt is secondary:
(2.1, copy IG a copy IC={ G1,G2,…,Gd};
(2.2, randomly select from ICTo sub- car networking, and to each pair of sub- car networking G being selectedτAnd GεIt executes such as
Lower operation obtains new sub- car networking IC '={ G1’,G2’,…,Gd' }:
(2.2.1, obtain GτIn cause and G since car networking is dividedεBetween the degree of correlation lose nτA vehicle connection individualG is obtained simultaneouslyεIn cause and G since car networking is dividedτBetween the degree of correlation lose nεA vehicle connection individual
(2.2.2, in VτMiddle selection and GεIt is highest that middle vehicle joins the individual degree of correlationA vehicle connection individual, is denoted asMeanwhile in VεMiddle selection and GτIt is highest that middle vehicle joins the individual degree of correlationA vehicle connection individual, note
For
(2.2.3, by GτMiddle vehicle connection individualWith GεMiddle vehicle connection individual
It exchanges, obtains two new sub- car networking Gφ' and Gζ';
(2.3, more global degree of correlation gc (IC) and gc (IC '), if gc (IC) < gc (IC '), IG is updated to
IC';
(3, return to final d sub- car networking IG={ G1,G2,…,Gd}。
Two,
In step 2 of the invention, in order to keep the workload of idle server balanced, we take the mode of being implemented as follows
By d sub- car networking G1,G2,…,GdIt is assigned in s platform idle server: for every sub- car networking Gx(1≤x≤d) is obtained
Its vehicle joins individual amount nx, and calculate the distribution average value upper boundThen s is divided into this d sub- car networkings
Group, so that being grouped optimal measurementMinimum, wherein mθFor all sub- car networkings in θ platform idle server
Vehicle joins individual amount summation.
Three,
In step 3 of the invention, vehicle connection individual v is calculated to process flow such as Fig. 2 of each Pre-Evaluation information λ interest-degree
It is shown.
During processing, we obtain 3 inputs, i.e. vehicle connection individual v, Pre-Evaluation information λ and traversal threshold value first
t;And initialize 4 intermediate variables, i.e. orderly information listIt traverses vehicle and joins individual ordered listTraverse number b=0 with
And vehicle joins individual median v '=v.Then, it in the range of traversing threshold value t allows, obtains currently traversed vehicle and joins individual v '
Directly related vehicle joins individual collections R;And then it obtains the maximum vehicle of push ability value in R and joins individual pm, wherein pmPush ability
Value cap (pm) it is represented by v ' and pmBetween the degree of correlation, i.e. cap (pm)=W (v ', pm);And by pmIt is inserted into traversal vehicle connection individual
In ordered list H.Then, for pmOn the information aggregate S that is stored, obtain in S with the maximum information i of the λ interest degree of correlationm,
Middle imWith the interest degree of correlation inco (i of λm, λ) and it may be expressed as:
Wherein e=2.72 is the bottom of natural logrithm,To possess i in the sub- car networking where v simultaneouslymWith the vehicle connection of λ
Set composed by body,And inp,λRespectively known vehicle connection individual p is to imWith the interest-degree of λ,He is possessed for p
The interest-degree average value of all information;And by imIt is inserted into orderly information list I.After reaching traversal threshold value t, we are last
Based on ordered list I and H, vehicle connection individual v is calculated to the interest-degree in of λv,λ:
Four,
In step 4 of the invention, when choosing specified quantity l vehicle connection individuals relevant to v in set M-R, we
Use functionImplement l, i.e.,And it is chosen in set M-R according to the degree of correlation size with v
L vehicle joins individual v before choosing1,v2,…,vl.Then, individual v is joined for each vehicley(1≤y≤l), based on the interest in step 3
Calculation method is spent, the highest information λ of interest-degree is obtained1 (y),λ2 (y),…,λk (y).Finally, we merge step 3 and step 4 generates
(l+1) k information obtain UI={ λ1,λ2,…,λk,λ1 (1),λ2 (1),…,λk (1),…,λ1 (v),λ2 (v),…,λk (v), from
The highest k information m_ λ of interest-degree is obtained in UI1,m_λ2,…,m_λkAnd send vehicle connection individual v to.
Claims (1)
1. Information Push Service processing method in a kind of car networking, which comprises the steps of:
Step 1 obtains car networking G, the number k for needing pushed information and traversal threshold value t from input terminal, and traversal threshold value t is push
When information, the vehicle connection individual amount for needing to access, and car networking is divided by d sub- car networkings based on vehicle connection individual attribute feature
G1,G2,…,Gd, so that the sub- car networking set { G after segmentation1,G2,…,GdRelated to the overall situation between the car networking G before segmentation
Spend loss reduction;
Step 2, according to the idle server quantity s in system, by d obtained in step 1 sub- car networking G1,G2,…,GdPoint
It is fitted in these idle servers, this s platform idle server starts the sub- car networking above parallel processing respectively later;
Step 3, for car networking G on serverxIn each vehicle join individual v, wherein 1≤x≤d, institute of the acquisition not on v
The set being made of informationThen, forIn each Pre-Evaluation information λ, calculate v to the interest-degree of λ;When
In all Pre-Evaluation information interest-degree calculate and finish after, return to the highest k information λ of interest-degree1,λ2,…,λk;
Step 4 joins individual v for each vehicle in car networking G, obtains the vehicle directly related with v and joins individual collections M, obtains simultaneously
Vehicle connection individual collections R directly related with v in sub- car networking is taken, specified quantity l and v phase are then chosen in set M-R
The vehicle of pass joins individual, joins individual for selected vehicle, obtains the highest k information of its interest-degree;When all selected vehicles join
The acquisition of information of individual finishes, the comprehensive local pushed information obtained in step 3, obtains the highest k letter of final interest-degree
It ceases and sends vehicle connection individual v to;
In the step 1, car networking G is indicated with graph structure, i.e. G=(V, E, W), and wherein V is all vehicle connection individual compositions in G
Set, E is the set that correlativity forms between vehicle connection individual, and W is the set that the degree of correlation forms between vehicle connection individual, every in E
A correlativity has a degree of correlation in W to be corresponding to it;Individual v is joined for any two vehicle in car networking G1And v2, such as
Fruit they have correlativity, i.e. (v1,v2) ∈ E, then their degree of correlation W (v1,v2) calculate as follows, if do not had
There is correlativity, then the degree of correlation is 0:
Step 1-1, z attributive character a for joining individual relatedness computation for vehicle in car networking G is obtained1,a2,…,az, and then obtain
Take corresponding with G | V | × z vehicle joins individual-attributive character matrix:
Wherein vi(aj) indicate that vehicle joins individual viIn j-th of attributive character ajOn value, 1≤i≤| V |, 1≤j≤z;
Step 1-2, individual v is joined for vehicle each in G, calculates its attributive character mean valueAnd it is counted by matrix V A
Calculate covariance matrix corresponding with it:
Step 1-3, vehicle is joined into individual v1Z attributive character value be organized into one-dimensional vector and its transposition respectively,Vehicle is joined into individual v2Z attributive character value point
It is not organized into one-dimensional vector and its transposition,
Step 1-4, v is calculated1And v2Between the degree of correlation:
After the relatedness computation between vehicles all in G connection individual finishes, in order to divide the global degree of correlation damage of front and back car networking
Minimum is lost, G is divided into d sub- car networkings based on following steps, d takes a number in 10,15 and 20:
(1, using k-means (k- mean value) clustering algorithm G is divided into d sub- car networking IG={ G1,G2,…,Gd, and calculate
The global degree of correlation of IG:
(2, following operation is repeated to sub- car networking IGIt is secondary:
(2.1, copy IG a copy IC={ G1,G2,…,Gd};
(2.2, randomly select from ICTo sub- car networking, and to each pair of sub- car networking G being selectedτAnd GεExecute following behaviour
Make, obtains new sub- car networking IC '={ G1’,G2’,…,Gd' }:
(2.2.1, obtain GτIn cause and G since car networking is dividedεBetween the degree of correlation lose nτA vehicle connection individualG is obtained simultaneouslyεIn cause and G since car networking is dividedτBetween the degree of correlation lose nεA vehicle connection individual
(2.2.2, in VτMiddle selection and GεIt is highest that middle vehicle joins the individual degree of correlationA vehicle connection individual, is denoted asMeanwhile in VεMiddle selection and GτIt is highest that middle vehicle joins the individual degree of correlationA vehicle connection individual, note
For
(2.2.3, by GτMiddle vehicle connection individualWith GεMiddle vehicle connection individualMutually
It changes, obtains two new sub- car networking Gφ' and Gζ';
(2.3, more global degree of correlation gc (IC) and gc (IC '), if gc (IC) < gc (IC '), is updated to IC ' for IG;
(3, return to final d sub- car networking IG={ G1,G2,…,Gd};
In the step 2, in order to keep the workload of idle server balanced, take the mode of being implemented as follows by d sub- car networkings
G1,G2,…,GdIt is assigned in s platform idle server: for every sub- car networking Gx, wherein 1≤x≤d obtains its vehicle connection
Individual amount nx, and calculate the distribution average value upper boundThen s group is divided into this d sub- car networkings, makes score
The optimal measurement of groupMinimum, wherein m θ is that the vehicle of all sub- car networkings in θ platform idle server joins individual
Quantity summation;
In the step 3, vehicle connection individual v is calculated to the process flow of each Pre-Evaluation information λ interest-degree:
3 inputs, i.e. vehicle connection individual v, Pre-Evaluation information λ and traversal threshold value t are obtained first;And 4 intermediate variables are initialized,
That is orderly information listIt traverses vehicle and joins individual ordered listIt traverses number b=0 and vehicle joins individual median v '=v;
Then, it in the range of traversing threshold value t allows, obtains the vehicle that currently traversed vehicle connection individual v ' is directly related and joins individual collections
R;And then it obtains the maximum vehicle of push ability value in R and joins individual pm, wherein pmPush ability value cap (pm) be represented by v ' with
pmBetween the degree of correlation, i.e. cap (pm)=W (v ', pm);And by pmTraversal vehicle is inserted into join in individual ordered list H;Then, for
pmOn the information aggregate S that is stored, obtain in S with the maximum information i of the λ interest degree of correlationm, wherein imWith the interest degree of correlation of λ
inco(im, λ) and it may be expressed as:
Wherein e=2.72 is the bottom of natural logrithm,To possess i in the sub- car networking where v simultaneouslymJoin individual institute with the vehicle of λ
The set of composition,And inp,λRespectively known vehicle connection individual p is to imWith the interest-degree of λ,All letters are possessed to him for p
The interest-degree average value of breath;And by imIt is inserted into orderly information list I;After reaching traversal threshold value t, we have finally been based on
Sequence table I and H calculate vehicle connection individual v to the interest-degree in of λv,λ:
In the step 4, when choosing specified quantity l vehicle connection individuals relevant to v in set M-R, function is usedCome
Implement l, i.e.,And l vehicle connection individual before being selected in set M-R according to the degree of correlation size with v
v1,v2,…,vl;Then, individual v is joined for each vehicley, wherein 1≤y≤l, based on the interest-degree calculation method in step 3,
Obtain the highest information λ of interest-degree1 (y),λ2 (y),…,λk (y);Finally, merging (l+1) k letter that step 3 and step 4 generate
Breath obtains UI={ λ1,λ2,…,λk,λ1 (1),λ2 (1),…,λk (1),…,λ1 (v),λ2 (v),…,λk (v), interest-degree is obtained from UI
Highest k information m_ λ1,m_λ2,…,m_λkAnd send vehicle connection individual v to.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710099060.7A CN106899668B (en) | 2017-02-23 | 2017-02-23 | Information Push Service processing method in car networking |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710099060.7A CN106899668B (en) | 2017-02-23 | 2017-02-23 | Information Push Service processing method in car networking |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106899668A CN106899668A (en) | 2017-06-27 |
CN106899668B true CN106899668B (en) | 2019-12-03 |
Family
ID=59185096
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710099060.7A Active CN106899668B (en) | 2017-02-23 | 2017-02-23 | Information Push Service processing method in car networking |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106899668B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110446184B (en) * | 2019-07-29 | 2021-06-08 | 华南理工大学 | Multi-mode switching Internet of vehicles routing method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103078930A (en) * | 2012-12-31 | 2013-05-01 | 广东工业大学 | Information distribution system based on vehicle internet |
CN103886073A (en) * | 2014-03-24 | 2014-06-25 | 河南理工大学 | Coal mine information recommendation system based on collaborative filtering |
CN105631531A (en) * | 2015-11-26 | 2016-06-01 | 东莞酷派软件技术有限公司 | Driving friend recommendation method, driving friend recommendation device and server |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104901817B (en) * | 2014-03-07 | 2018-07-10 | 腾讯科技(北京)有限公司 | Target information method for pushing and device |
-
2017
- 2017-02-23 CN CN201710099060.7A patent/CN106899668B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103078930A (en) * | 2012-12-31 | 2013-05-01 | 广东工业大学 | Information distribution system based on vehicle internet |
CN103886073A (en) * | 2014-03-24 | 2014-06-25 | 河南理工大学 | Coal mine information recommendation system based on collaborative filtering |
CN105631531A (en) * | 2015-11-26 | 2016-06-01 | 东莞酷派软件技术有限公司 | Driving friend recommendation method, driving friend recommendation device and server |
Also Published As
Publication number | Publication date |
---|---|
CN106899668A (en) | 2017-06-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101976294B1 (en) | Driving route matching method and apparatus and storage medium | |
Cobo et al. | A bibliometric analysis of the intelligent transportation systems research based on science mapping | |
Chen et al. | Probabilistic modeling of traffic lanes from GPS traces | |
Wu et al. | Probabilistic robust route recovery with spatio-temporal dynamics | |
Tan et al. | Grid-based data management in distributed simulation | |
CN105897584B (en) | Paths planning method and controller | |
CN103810299A (en) | Image retrieval method on basis of multi-feature fusion | |
US20160162793A1 (en) | Method and apparatus for decision tree based search result ranking | |
Yu et al. | Online clustering for trajectory data stream of moving objects | |
CN105913668B (en) | Method is surveyed in a kind of orientation deck car test based on huge traffic data statistics | |
CN109992786A (en) | A kind of semantic sensitive RDF knowledge mapping approximate enquiring method | |
CN112734219B (en) | Vehicle transportation running behavior analysis method and system | |
CN106021456A (en) | Point-of-interest recommendation method fusing text and geographic information in local synergistic arrangement | |
Ayani et al. | Optimizing cell-size in grid-based DDM | |
CN108229999B (en) | Method and device for evaluating competitive products | |
CN106919719A (en) | A kind of information completion method towards big data | |
CN113159357A (en) | Data processing method and device, electronic equipment and computer readable storage medium | |
Fiosina | Explainable Federated Learning for Taxi Travel Time Prediction. | |
CN106899668B (en) | Information Push Service processing method in car networking | |
CN107316306B (en) | A kind of diameter radar image fast partition method based on Markov model | |
CN111079940B (en) | Decision tree model establishing method and using method for real-time fake-licensed car analysis | |
Wang et al. | Abnormal trajectory detection based on geospatial consistent modeling | |
CN109740750B (en) | Data collection method and device | |
Fiosina | Interpretable privacy-preserving collaborative deep learning for taxi trip duration forecasting | |
CN107391728B (en) | Data mining method and data mining device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |