CN106327236A - Method and device for determining user action track - Google Patents
Method and device for determining user action track Download PDFInfo
- Publication number
- CN106327236A CN106327236A CN201510406207.3A CN201510406207A CN106327236A CN 106327236 A CN106327236 A CN 106327236A CN 201510406207 A CN201510406207 A CN 201510406207A CN 106327236 A CN106327236 A CN 106327236A
- Authority
- CN
- China
- Prior art keywords
- sequence
- fragment
- place
- frequent
- 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.)
- Granted
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses a method and a device for determining a user action track, which relate to the technical field of communication and aim at improving the determined user action track accuracy and further improving the efficiency for navigation and location recommendation for the user. The method comprises steps: R site sequences are determined according to a track sequence formed by N position data of the user; according to the R site sequences, M target sequence fragments are determined; and a line formed when sites in any of the M target sequence fragments are serially connected according to a time order serves as the action track of the user. The technical scheme provided by the embodiment of the invention can be used in a process of determining the user action track.
Description
Technical field
The present invention relates to communication technical field, particularly relate to a kind of method determining user's movement track and
Device.
Background technology
The movement track of user all has important potential value in multiple fields；Such as, use is being learnt
After the movement track of family, can be user's navigation, realize place recommendation more accurately for user, it is also possible to
Advertisement addressing etc. is optimized for trade company.Along with terminal unit (such as, the intelligence with automatic positioning function
Mobile phone, panel computer etc.) universal, the position data of user can be able to record at any time, how root
Determine that according to the position data of user the movement track of user has great importance.
At present, learning that user, after the position data of different time points, finds out respectively from different time
Put the place that the positional distance of corresponding position data representative is nearest, by these places according to time order and function
The route that sequential series is formed is as the movement track of user.
The movement track determined by said method, due to user to place be not necessarily distance position
Putting the place that data are nearest, the accuracy of the movement track of the user determined by said method is the highest,
Utilize this action track to carry out navigating for user, place efficiency when recommending the highest.
Summary of the invention
Embodiments of the invention provide a kind of method and device determining user's movement track, are used for improving
The accuracy of the movement track of the user determined, and then rise to that user carries out navigating, place is when recommending
Efficiency.
For reaching above-mentioned purpose, embodiments of the invention adopt the following technical scheme that
First aspect, it is provided that a kind of method determining user's movement track, including:
R place sequence is determined according to the track sets being made up of N number of position data of user；Its
In, a place sequence is made up of N number of place, and the n-th place in described N number of place is institute
State in the place set that the nth position data in N number of position data are corresponding；One position
Place collection corresponding to data is combined into: the mould in all places in the preset range centered by this position data
Stick with paste collection；1≤n≤N, R >=1, n, N and R are integer；
M target sequence fragment is determined according to described R place sequence；Wherein, when a sequence
Frequent degree in the fragment multiple places sequence in described R place sequence more than preset frequent degree,
And the probability sum of the plurality of place sequence more than predetermined probabilities time, this sequence fragment is target sequence
Fragment, sequence fragment frequent degree in a place sequence refers to that this sequence fragment is in this place
The number of times occurred in sequence；The probability of one place sequence refers to the whole places in this place sequence
The product of probability, the probability in a place refers to this place general in the place belonging to this place is gathered
Rate；M >=1, M is integer；
By the place in any one the target sequence fragment in described M target sequence fragment according to
The route that time order and function order series winding is formed is as the movement track of described user.
In conjunction with first aspect, in the implementation that the first is possible, described according to described R place
Sequence determines M target sequence fragment, including:
The frequent fragment Candidate Set of the frequent fragment a length of x+1 of generation according to a length of x, wherein,
As x=0, the frequent fragment Candidate Set of a length of 1 is: form all of described R place sequence
The set of different places；X >=0, x is integer；
Each place sequence in described R place sequence scans the frequent of described a length of x+1
Each sequence fragment in fragment Candidate Set, obtains in the frequent fragment Candidate Set of described a length of x+1
Each sequence fragment each place sequence in described R place sequence in frequent degree；
When a sequence fragment in the frequent fragment Candidate Set of described a length of x+1 is in multiple places sequence
Frequent degree in row is more than presetting the probability sum of frequent degree and the plurality of place sequence more than presetting
During probability, determine the frequent fragment that this sequence fragment is an a length of x+1；
Determine that M frequent fragment in whole described frequent fragments is M target sequence fragment.
In conjunction with the first possible implementation of first aspect, in the implementation that the second is possible,
Described M the frequent fragment determined in whole described frequent fragments is M target sequence fragment,
Including:
Determine that M Guan Bi frequent fragment in whole described frequent fragments is M target sequence sheet
Section；Wherein, the Guan Bi frequent fragment in described whole described frequent fragment refers to: for described entirely
The frequent fragment of the sub-piece of any one frequent fragment in the described frequent fragment in portion.
In conjunction with the implementation that the first possible implementation of first aspect or the second are possible,
In the third possible implementation, in a described place sequence, scan described a length of x+1's
During a sequence fragment in frequent fragment Candidate Set, by this sequence fragment of automat record
Frequent degree in this place sequence.
In conjunction with first aspect, first aspect the first possible implementation to the third possible reality
Existing mode any one, in the 4th kind of possible implementation, described track sets is that described user exists
Track sets in preset time period.
In conjunction with first aspect, first aspect the first possible implementation to the 4th kind of possible reality
Existing mode any one, in the 5th kind of possible implementation, described method also includes:
The target sequence fragment of multiple users is clustered, obtains k cluster cluster, k >=1, k
For integer；
Each user is expressed as the user vector with k dimension, a corresponding cluster of dimension
Cluster, the value in a dimension is the target sequence fragment of user in the cluster cluster that this dimension is corresponding
Quantity；
Set up gauss hybrid models, and mix according to Gauss described in the user vector matching of the plurality of user
The parameter of matched moulds type, described gauss hybrid models is made up of multiple Gauss models, a Gauss model pair
Answering a user community, multiple user community corresponding to the plurality of Gauss model are by the plurality of user
Constitute.
Second aspect, it is provided that a kind of device determining user's movement track, including:
First determines unit, for determining according to the track sets being made up of N number of position data of user
R place sequence；Wherein, a place sequence is made up of N number of place, in described N number of place
N-th place is in the place set that the nth position data in described N number of position data are corresponding
One；Place collection corresponding to one position data is combined into: the preset range centered by this position data
The fuzzy set in interior all places；1≤n≤N, R >=1, n, N and R are integer；
Second determines unit, for determining M target sequence fragment according to described R place sequence；
Wherein, when the frequent degree in the sequence fragment multiple places sequence in described R place sequence
When being more than predetermined probabilities more than the probability sum presetting frequent degree and the plurality of place sequence, this sequence
Column-slice section is target sequence fragment, and sequence fragment frequent degree in a place sequence refers to this
The number of times that sequence fragment occurs in this place sequence；The probability of one place sequence refers to this place sequence
The product of the probability in the whole places in row, the probability in a place refers in the place belonging to this place
The probability in this place in set；M >=1, M is integer；
Performance element, for by any one the target sequence sheet in described M target sequence fragment
The route that place in Duan is formed according to time order and function order series winding is as the movement track of described user.
In conjunction with second aspect, in the implementation that the first is possible, described second determines that unit includes:
Signal generating unit, generates the frequent fragment of a length of x+1 for the frequent fragment according to a length of x
Candidate Set, wherein, as x=0, the frequent fragment Candidate Set of a length of 1 is: form described R
The set of all different places of place sequence；X >=0, x is integer；
Scanning element, scans described length in each place sequence in described R place sequence
Degree is each sequence fragment in the frequent fragment Candidate Set of x+1, obtains the frequency of described a length of x+1
The each sequence fragment in numerous fragment Candidate Set each place sequence in described R place sequence
In frequent degree；
First determines subelement, for as in the frequent fragment Candidate Set of described a length of x+1
Sequence fragment frequent degree in the sequence of multiple places is more than presetting frequent degree and the plurality of place sequence
When the probability sum of row is more than predetermined probabilities, determine that this sequence fragment is the frequent of an a length of x+1
Fragment；
Second determines subelement, for determining M frequent fragment in whole described frequent fragments
For M target sequence fragment.
In conjunction with the first possible implementation of second aspect, in the implementation that the second is possible,
Described second determine subelement specifically for:
Determine that M Guan Bi frequent fragment in whole described frequent fragments is M target sequence sheet
Section；Wherein, the Guan Bi frequent fragment in described whole described frequent fragment refers to: for described entirely
The frequent fragment of the sub-piece of any one frequent fragment in the described frequent fragment in portion.
In conjunction with the implementation that the first possible implementation of second aspect or the second are possible,
In the third possible implementation, described scanning element scans described in a described place sequence
During a sequence fragment in the frequent fragment Candidate Set of a length of x+1, remembered by automat
Record this sequence fragment frequent degree in this place sequence.
In conjunction with second aspect, second aspect the first possible implementation to the third possible reality
Existing mode any one, in the 4th kind of possible implementation, described track sets is that described user exists
Track sets in preset time period.
In conjunction with second aspect, second aspect the first possible implementation to the 4th kind of possible reality
Existing mode any one, in the 5th kind of possible implementation, described device also includes:
Cluster cell, for the target sequence fragment of multiple users being clustered, obtains k cluster
Cluster, k >=1, k is integer；
Represent unit, for each user being expressed as the user vector with k dimension, a dimension
The corresponding cluster cluster of degree, the value in a dimension is user in the cluster cluster that this dimension is corresponding
The quantity of target sequence fragment；
Computing unit, is used for setting up gauss hybrid models, and according to the user vector of the plurality of user
The parameter of gauss hybrid models described in matching, described gauss hybrid models is made up of multiple Gauss models,
One corresponding user community of Gauss model, multiple user community that the plurality of Gauss model is corresponding
It is made up of the plurality of user.
The method and device that the embodiment of the present invention provides, can be according to by the historical position data structure of user
The movement track become determines the place sequence that user is corresponding, when a sequence fragment is in multiple places sequence
In the biggest and the plurality of place sequence the probability sum of frequent degree the biggest time, illustrate user through should
The probability in the place in sequence fragment is the biggest.If the value of default frequent degree and predetermined probabilities is carried out rationally
When arranging, compared with prior art, the method provided according to embodiments of the present invention determines target sequence
After fragment, the movement track of the user determined according to this target sequence fragment is relatively accurate, according to this action
Track be user carry out navigating, place efficiency when recommending higher.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below by right
In embodiment or description of the prior art, the required accompanying drawing used is briefly described, it should be apparent that,
Accompanying drawing in describing below is only some embodiments of the present invention, for those of ordinary skill in the art
From the point of view of, on the premise of not paying creative work, it is also possible to obtain the attached of other according to these accompanying drawings
Figure.
The flow chart of a kind of method determining user's movement track that Fig. 1 provides for the embodiment of the present invention；
The schematic diagram of the track sets that Fig. 2 provides for the embodiment of the present invention；
The structural representation of a kind of device determining user's movement track that Fig. 3 provides for the embodiment of the present invention
Figure；
Fig. 4 provide for the embodiment of the present invention another determine that the structure of device of user's movement track is shown
It is intended to；
Fig. 5 provide for the embodiment of the present invention another determine that the structure of device of user's movement track is shown
It is intended to.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is entered
Row clearly and completely describes, it is clear that described embodiment is only a part of embodiment of the present invention,
Rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not having
Have and make the every other embodiment obtained under creative work premise, broadly fall into present invention protection
Scope.
The executive agent of the embodiment of the present invention can be server, work station and common PC (Personal
Computer, personal computer) etc., server is specifically as follows HP DL server catalyst Catalyst, work
Z820 series of tasks station it is specifically as follows as station.The position data of the user in the embodiment of the present invention is
Referring to the historical position data of user, the position data of user can be a certain moment user position
Latitude and longitude value, should in the case of, the track sets of user can be according to user's row within a period of time
Dynamic track obtain according to the tactic multiple latitude and longitude value of time order and function.The position data of user is also
Can represent by other means, this is not specifically limited by the embodiment of the present invention.
Embodiments provide a kind of method determining user's movement track, as it is shown in figure 1, bag
Include:
101, R place sequence is determined according to the track sets being made up of N number of position data of user；
Wherein, a place sequence is made up of N number of place, and the n-th place in described N number of place is
In the place set that nth position data in described N number of position data are corresponding one；One position
The place collection putting data corresponding is combined into: all places in the preset range centered by this position data
Fuzzy set；1≤n≤N, R >=1, n, N and R are integer.
Wherein, the place in the place set that position data is corresponding can have one or more, and one
Individual position is any one position data in N number of position data.
Place set in place be specifically as follows square, bar, luxury stores, Italian restaurant,
The places such as McDonald, library, public transport stop board, parking lot and museum, can also be other certainly
Place, this is not limited by the embodiment of the present invention.
Concrete, when the number difference in the place in N number of place set that N number of position data is corresponding
For r1、r2、…rNTime, the place sequence determined according to the track sets being made up of this N number of position data
Row can be R, R=r1·r2·…·rN。
Exemplary, as in figure 2 it is shown, a track sets is made up of the 4 of user position datas,
These 4 position datas are respectively N1、N2、N3And N4, this track represented by track sets reality
May be curvilinear path, for simplicity in Fig. 2, this track is depicted as straight line.With N1、N2、N3
And N4Broken circle for the center of circle represents the scope preset centered by this position data, then N1Corresponding
Place set be the fuzzy set that is made up of place A, B and C, N2Corresponding place set is for by field
The fuzzy set that institute A, B and C are constituted, N3Corresponding place set is be made up of place A and D
Fuzzy set, N4Corresponding place set is the fuzzy set being made up of place B.
Then may determine that 18 place sequences according to this track sets, be respectively as follows: 1. [A, A, A,
B]；2. [A, A, D, B]；3. [A, B, A, B]；4. [A, B, D, B]；5. [A, C, A,
B]；6. [A, C, D, B]；7. [B, A, A, B]；8. [B, A, D, B]；9. [B, B, A,
B]；10. [B, B, D, B]；11. [B, C, A, B]；12. [B, C, D, B]；13. [C, A,
A, B]；14. [C, A, D, B]；15. [C, B, A, B]；16. [C, B, D, B]；17. [C,
C, A, B]；18. [C, C, D, B].
102, M target sequence fragment is determined according to described R place sequence；Wherein, when one
Frequent degree in the sequence fragment multiple places sequence in described R place sequence is more than presetting frequency
When the probability sum of numerous degree and the plurality of place sequence is more than predetermined probabilities, this sequence fragment is mesh
Mark sequence fragment, sequence fragment frequent degree in a place sequence refers to that this sequence fragment exists
The number of times occurred in this place sequence；It is whole that the probability of one place sequence refers in this place sequence
The product of the probability in place, the probability in a place refer to belonging to this place place gather in this
Probability；M >=1, M is integer.
Wherein, target sequence fragment can be made up of a place, it is also possible to is made up of multiple places.
Concrete, when a place sequence is [C, B, A, B], sequence fragment [AB], sequence
Fragment [CA], sequence fragment [BB] and sequence fragment [CB] etc. are the sub-piece of this place sequence, i.e.
The sub-piece of one place sequence be made up of at least one place in this place sequence and this at least
One place arranges according to its sequencing occurred in affiliated place sequence.
Based on the example described in Fig. 2, sequence fragment [A] occurs in that 3 times in the 1st place sequence,
Then the sequence fragment [A] frequent degree in the 1st place sequence is 3.Sequence fragment [AB] is at the 3rd
Occur in that in the sequence of place 2 times, then the sequence fragment [AB] frequent degree in the 3rd place sequence is
2.It should be noted that when place sequence is [A, C, B, A, E, B, D, B], sequence
The fragment [AB] frequent degree in this place sequence is also 2, when i.e. determining the frequent degree of sequence fragment,
Several places in sequence fragment need not to be occurred in the sequence of place continuously, as long as these several
Several places of sequencing and this sequencing when occurring in the sequence of place consistent.
Concrete, in the place set that position data is corresponding, the probability in each place is user
The probability in this place is gone to from the position that this position data represents.Due to the nearest apart from this position data
Place, the probability that user goes to this place from the position that this position data represents is the biggest, therefore, at this
Between the position that in the set of place, the probability in each place and this place and this position data represent away from
From being inversely proportional to.
In the place set that position data is corresponding, the probability in a place can pass through Rayleigh
(Rayleigh) distribution calculates, and is specifically as follows:Wherein, d is
This position data represent position and this place between distance, f (d) be this place set in this
Probability, the general value of σ is 1.
Exemplary, based on the example shown in Fig. 2, N2Corresponding place set is for by place A, B
The fuzzy set constituted with C, wherein, in this place is gathered, the probability of place B is f (d2), wherein,
As in figure 2 it is shown, d2For N2And the distance between the B of place.
As a example by the 4th place sequence, the probability of the 4th place sequence is P=
f(d1)·f(d2)·f(d3)·f(d4), wherein, d1For N1And the distance between A, d2For N2With B it
Between distance, d3For N3And the distance between D, d4For N4And the distance between B.
It addition, presetting in the range of the circumference D centered by this position data when a position data
Rice scope time, this position data corresponding place set in, the probability in a place can also lead to
Cross in the following manner to calculate:Wherein, the d position that to be this place represent with this position data it
Between distance.
Exemplary, as a example by the 4th place sequence, the probability of the 4th place sequence is
103, by the place in any one the target sequence fragment in described M target sequence fragment
The route formed according to time order and function order series winding is as the movement track of described user.
Generally, the movement track of user every day can have certain rule (such as working clan,
Fixing movement track is typically had) on weekdays.The method that the embodiment of the present invention provides is permissible
The place sequence that user is corresponding is determined according to the movement track being made up of the historical position data of user, when
One sequence fragment frequent degree in the sequence of multiple places is the biggest and the probability of the plurality of place sequence
When sum is the biggest, illustrate that the probability in user place in this sequence fragment is the biggest.If by default frequency
When the value of numerous degree and predetermined probabilities is reasonably arranged, compared with prior art, real according to the present invention
After the method that executing example provides determines target sequence fragment, the user's determined according to this target sequence fragment
Movement track is relatively accurate, according to this action track be user carry out navigating, place recommend time efficiency relatively
High.
Optionally, step 102 when implementing can be:
1021, the frequent fragment according to a length of x generates the frequent fragment Candidate Set of a length of x+1,
Wherein, as x=0, the frequent fragment Candidate Set of a length of 1 is: form described R place sequence
The set of all different places；X >=0, x is integer.
1022, each place sequence in described R place sequence scans described a length of x+1
Frequent fragment Candidate Set in each sequence fragment, obtain described a length of x+1 frequent fragment wait
Frequent in each sequence fragment in selected works each place sequence in described R place sequence
Degree.
1023, when a sequence fragment in the frequent fragment Candidate Set of described a length of x+1 is multiple
Frequent degree in the sequence of place is big more than the probability sum presetting frequent degree and the plurality of place sequence
When predetermined probabilities, determine the frequent fragment that this sequence fragment is an a length of x+1.
1024, determine that M frequent fragment in whole described frequent fragments is M target sequence
Fragment.
Exemplary, based on the example described in Fig. 2, the frequent fragment Candidate Set of a length of 1 is composition
The set in all places of above-mentioned 18 place sequences, is i.e. made up of place [A], [B], [C] and [D]
Set.
Concrete, can use Apriori algorithm that the frequent fragment of a length of x is generated a length of x+1
Frequent fragment Candidate Set.It addition, the frequent fragment of a length of x+1 can also obtain in the following manner
To: the frequent fragment of a length of 1 is sequentially inserted in the place in the frequent fragment of a length of x,
In order to simplicity describes, hereinafter the embodiment of the present invention is described as a example by this situation.
Exemplary, when frequent fragment when a length of 1 is [A] and [B], the most a length of 2 frequent
Fragment Candidate Set is the set being made up of sequence fragment [AA], [AB], [BA] and [BB].
Exemplary, the frequent fragment when a length of 2 is [AB] and [AC], the frequent sheet of a length of 1
When section is [A], [D], [A] is inserted in [AB] sequence fragment obtained is [AAB] and [ABA]；
[D] is inserted in [AB] sequence fragment obtained is [DAB], [ADB] and [ABD]；In like manner, will
[A] is inserted in [AC], and [D] is inserted in [AC] sequence fragment that can obtain other, the most a length of
The frequent fragment Candidate Set of 3 serve as reasons [AAB], [ABA], [DAB], [ADB], [ABD], [AAC],
The set that [ACA], [DAC], [ADC] and [ACD] forms.
Optionally, in a described place sequence, scan the frequent fragment candidate of described a length of x+1
During the sequence fragment concentrated, by this sequence fragment of automat record in this place sequence
In frequent degree.
In the case of Gai, first, each sequence fragment in frequent fragment Candidate Set be arranged one automatically
Machine.As a example by above-mentioned place sequence 15, as the sequence fragment [CB] that sweep length is 2 and [CA],
Now [CB] corresponding first automat, [CA] corresponding second automat, then scan first place C
Time, two automats simultaneously enter original state C.When scanning second place B, first certainly
Motivation enters second state B, owing to B is the final of the sequence fragment [CB] that the first automat is corresponding
State, therefore, the frequent degree of [CB] adds 1.When scanning the 3rd place A, second automat
Enter second state A, owing to A is the end-state of the sequence fragment [CA] that the second automat is corresponding,
Therefore, the frequent degree of [CA] adds 1.Afterwards, two automats continue to scan on, until the field scanned
Do C time, the first automat and the second automat reenter original state C, repeat mistake above
Journey, until scanning to last position of place sequence, determines that sequence fragment is in this place sequence
Frequent degree.
Optionally, step 1024 when implementing can be: determines whole described frequent fragments
In M Guan Bi frequent fragment be M target sequence fragment；Wherein, described whole described frequency
Guan Bi frequent fragment in numerous fragment refers to: any one in described whole described frequent fragment
The frequent fragment of the sub-piece of individual frequent fragment.It should be noted that a frequent fragment is not self
Sub-piece.
Generally, can determine that substantial amounts of target sequence sheet according to a track sets of user
Section, this optional method, it is possible to reduce the number of target sequence fragment.
Exemplary, if in whole frequent fragments, have 6 frequent fragments, wherein, length
Be 1 frequent fragment be [A], [B], [C]；The frequent fragment of a length of 2 is [AB], [AC]；Long
Degree be the frequent fragment of 3 be [ABC].Then it is [ABC] due to [A], [B], [C], [AB], [AC]
Sub-piece, frequent fragment [ABC] for Guan Bi frequent fragment, then may determine that Guan Bi frequent fragment
[ABC] is target sequence fragment.
Above embodiment described the target sequence utilizing a track sets of a user to determine user
The method of column-slice section, when user has multiple track sets, it is possible to use said method is according to multiple rails
Mark sequence determines the target sequence fragment of user.
Optionally, described track sets is described user track sets in preset time period.
Concrete, preset time period can be within 1 day, can also to be a morning or afternoon, the most also
Can be other preset time period, this be not specifically limited by the embodiment of the present invention.
When implementing, owing to user can carry out different activities in the different time periods, in order to carry
The accuracy of the movement track of the user that height determines, it is possible to use the track sequence of user every day in working day
In row determine the target sequence fragment of user or utilize day off, the track sets of user every day determines use
The target sequence fragment at family.The target sequence fragment determined in the case of utilizing this determines the action rail of user
During mark, different recommendations can be provided the user from day off according to the movement track of user on weekdays
Content, improves the efficiency recommended.
Optionally, said method can also comprise the following steps:
(1) the target sequence fragment of multiple users is clustered, obtain k cluster cluster, k >=
1, k is integer.
(2) each user is expressed as the user vector with k dimension, a dimension correspondence one
Individual cluster cluster, the value in a dimension is the target sequence of user in the cluster cluster that this dimension is corresponding
The quantity of column-slice section.
(3) gauss hybrid models is set up, and according to described in the user vector matching of the plurality of user
The parameter of gauss hybrid models, described gauss hybrid models is made up of multiple Gauss models, a Gauss
The corresponding user community of model, multiple user community corresponding to the plurality of Gauss model are by described many
Individual user is constituted.
In this optional method, the target sequence fragment of user can be true according to a track sets
Fixed target sequence fragment, it is also possible to for the target sequence fragment determined according to multiple track sets.
Exemplary, K-means (K average), K-medoids (K median) etc. can be passed through
The target sequence fragment of multiple users is clustered by algorithm, certainly can also be real by other algorithm
Existing, this is not limited by the embodiment of the present invention.
Assume that above-mentioned multiple user is divided into J user community, respectively C1,...,CJ, GMM
(Gaussian Mixture Model, gauss hybrid models) is made up of J Gauss model, and one high
The corresponding user community of this model.One Gauss model can be by user corresponding to this Gauss model
User vector structure in group.Assume I user vector (generally, the I having I user
Value bigger), then the probability density function of the GMM of J Gauss model linear addition composition is:1≤j≤J, 1≤i≤I, i, I, j and J are integer.
Wherein, P (Cj) it is the probability that occurs in I user of jth user community, P (Ui|Cj) it is
The probability that i-th user occurs in jth user community, P (Ui| Θ) it is parameter i-th when being Θ
The probability that user occurs.
Θ includes 2J parameter: be respectively μ1,...,μJAnd Σ1,...,ΣJ, wherein, μjAnd ΣjUse for jth
The parameter that family group is corresponding.Hereinafter μj gRefer to that g (g >=0, g is integer) takes turns calculated
The μ that jth user community is correspondingj, Σj gRefer to that g takes turns calculated jth user community corresponding
Σj, the process of the parameter of matching GMM (i.e. calculates μ1,...,μJAnd Σ1,...,ΣJProcess) including:
(1) utilize according to μj (g)And Σj (g)Calculated P (Ui|Cj) and P (Cj), according toCalculate μj (g+1), utilize μj (g+1), according to CalculateAccording to μj (g+1)WithCalculate P (Ui|Cj) and
P(Cj).Wherein, μ is calculatedj (g+1)WithTime P (the C that usesj|Ui) can be by according to μj (g)And Σj (g)Meter
P (the U obtainedi|Cj) and P (Cj) be calculated.
(2) g=g+1 is made.
Repeat step (1) and (2), until P (Ui|Cj) convergence, by P (Ui|Cj) convergence time μjWith
ΣjIt is defined as the parameter that jth user community is corresponding.
As g=0, initialize μjAnd Σj:
According to μj (0)And Σj (0)P (U can be calculatedi|Cj) and P (Cj), wherein, UiRefer to the user of i-th user
Vector.
The parameter that user vector according to each user is corresponding with each user community may determine that each
User belongs to the probability of each user community, when a user belongs to the maximum probability of a user community
Time, determine that the user community belonging to this user is this user community.So may determine that according to the method
Go out the user community belonging to above-mentioned multiple user.
This optional method, by substantial amounts of target sequence fragment corresponding for user through cluster, is reduced to
User vector, and establish GMM model according to the user vector of multiple users, matching GMM
The parameter of model so that may determine that this user is in each user community according to the user vector of user
Probability, by the user community being defined as belonging to user of maximum probability.This optional method, can be right
The user being in same user community carries out identical recommendation, improves the efficiency recommended.
Embodiments provide a kind of device 30 determining user's movement track, for performing such as figure
The method of the determination user's movement track shown in 1, as it is shown on figure 3, this device 30 includes:
First determines unit 301, for according to the track sets being made up of N number of position data of user
Determine R place sequence；Wherein, a place sequence is made up of N number of place, described N number of field
The n-th place in is the place collection that the nth position data in described N number of position data are corresponding
In conjunction one；Place collection corresponding to one position data is combined into: pre-centered by this position data
If the fuzzy set in all places in scope；1≤n≤N, R >=1, n, N and R are integer；
Second determines unit 302, for determining M target sequence sheet according to described R place sequence
Section；Wherein, when the frequency in the sequence fragment multiple places sequence in described R place sequence
When numerous degree is more than predetermined probabilities more than the probability sum presetting frequent degree and the plurality of place sequence,
This sequence fragment is target sequence fragment, and sequence fragment frequent degree in a place sequence is
Refer to the number of times that this sequence fragment occurs in this place sequence；The probability of one place sequence refers to this
The product of the probability in the whole places in institute's sequence, the probability in a place refers to belonging to this place
The probability in this place in the set of place；M >=1, M is integer；
Performance element 303, for by any one target sequence in described M target sequence fragment
The route that place in fragment is formed according to time order and function order series winding is as the action rail of described user
Mark.
Optionally, as shown in Figure 4, described second determines that unit 302 may include that
Signal generating unit 3021, generates the frequent of a length of x+1 for the frequent fragment according to a length of x
Fragment Candidate Set, wherein, as x=0, the frequent fragment Candidate Set of a length of 1 is: composition is described
The set of all different places of R place sequence；X >=0, x is integer；
Scanning element 3022, scans in each place sequence in described R place sequence
Each sequence fragment in the frequent fragment Candidate Set of described a length of x+1, obtains described a length of
The each sequence fragment in the frequent fragment Candidate Set of x+1 each field in described R place sequence
Frequent degree in institute's sequence；
First determines subelement 3023, for when in the frequent fragment Candidate Set of described a length of x+1
One sequence fragment frequent degree in the sequence of multiple places is more than presetting frequent degree and the plurality of field
When the probability sum of institute's sequence is more than predetermined probabilities, determine that this sequence fragment is an a length of x+1's
Frequent fragment；
Second determines subelement 3024, for determining M in whole described frequent fragments frequently
Fragment is M target sequence fragment.
Optionally, described second determine subelement 3024 specifically for:
Determine that M Guan Bi frequent fragment in whole described frequent fragments is M target sequence sheet
Section；Wherein, the Guan Bi frequent fragment in described whole described frequent fragment refers to: for described entirely
The frequent fragment of the sub-piece of any one frequent fragment in the described frequent fragment in portion.
Optionally, described scanning element 3022 scans described a length of in a described place sequence
During a sequence fragment in the frequent fragment Candidate Set of x+1, by this sequence of automat record
Column-slice section frequent degree in this place sequence.
Optionally, described track sets is described user track sets in preset time period.
Optionally, as shown in Figure 4, described device 30 can also include:
Cluster cell 304, for the target sequence fragment of multiple users being clustered, obtains k
Cluster cluster, k >=1, k is integer；
Represent unit 305, for each user being expressed as the user vector with k dimension, one
The corresponding cluster cluster of individual dimension, the value in a dimension is in the cluster cluster that this dimension is corresponding
The quantity of the target sequence fragment of user；
Computing unit 306, is used for setting up gauss hybrid models, and according to the user of the plurality of user
The parameter of gauss hybrid models described in vector matching, described gauss hybrid models is by multiple Gauss model structures
Become, a corresponding user community of Gauss model, multiple users that the plurality of Gauss model is corresponding
Group is made up of the plurality of user.
Generally, the movement track of user every day can have certain rule (such as working clan,
Fixing movement track is typically had) on weekdays.The device that the embodiment of the present invention provides, permissible
The place sequence that user is corresponding is determined according to the movement track being made up of the historical position data of user, when
One sequence fragment frequent degree in the sequence of multiple places is the biggest and the probability of the plurality of place sequence
When sum is the biggest, illustrate that the probability in user place in this sequence fragment is the biggest.If by default frequency
When the value of numerous degree and predetermined probabilities is reasonably arranged, compared with prior art, real according to the present invention
After the method that executing example provides determines target sequence fragment, the user's determined according to this target sequence fragment
Movement track is relatively accurate, according to this action track be user carry out navigating, place recommend time efficiency relatively
High.
Hardware realize on, the unit in said apparatus 30 can be embedded in the form of hardware or
In processor independent of device 30, it is also possible to be stored in a software form in the memorizer of device 30,
So that processor calls performs the operation that above unit is corresponding, this processor can be centre
Reason unit (CPU), microprocessor, single-chip microcomputer etc..
As it is shown in figure 5, the embodiment of the present invention provides another kind to determine the device 50 of user's movement track,
For the method performing determination user's movement track as shown in Figure 1, this device 50 includes: storage
Device 501, processor 502 and bus system 503.
Wherein, it is to be coupled by bus system 503 between memorizer 501 and processor 502
, wherein memorizer 501 may comprise random access memory, it is also possible to also includes non-volatile depositing
Reservoir, for example, at least one disk memory.Bus system 503, can be isa bus, PCI
Bus or eisa bus etc..This bus system 503 can be divided into address bus, data/address bus, control
Bus processed etc..For ease of representing, Fig. 5 only represents with a thick line, it is not intended that only one
Bus or a type of bus.
Memorizer 501 for one group of code of storage, this code be used for controlling processor 502 perform with
Lower action:
R place sequence is determined according to the track sets being made up of N number of position data of user；Its
In, a place sequence is made up of N number of place, and the n-th place in described N number of place is institute
State in the place set that the nth position data in N number of position data are corresponding；One position
Place collection corresponding to data is combined into: the mould in all places in the preset range centered by this position data
Stick with paste collection；1≤n≤N, R >=1, n, N and R are integer；
M target sequence fragment is determined according to described R place sequence；Wherein, when a sequence
Frequent degree in the fragment multiple places sequence in described R place sequence more than preset frequent degree,
And the probability sum of the plurality of place sequence more than predetermined probabilities time, this sequence fragment is target sequence
Fragment, sequence fragment frequent degree in a place sequence refers to that this sequence fragment is in this place
The number of times occurred in sequence；The probability of one place sequence refers to the whole places in this place sequence
The product of probability, the probability in a place refers to this place general in the place belonging to this place is gathered
Rate；M >=1, M is integer；
By the place in any one the target sequence fragment in described M target sequence fragment according to
The route that time order and function order series winding is formed is as the movement track of described user.
Optionally, described processor 502 specifically for:
The frequent fragment Candidate Set of the frequent fragment a length of x+1 of generation according to a length of x, wherein,
As x=0, the frequent fragment Candidate Set of a length of 1 is: form all of described R place sequence
The set of different places；X >=0, x is integer；
Each place sequence in described R place sequence scans the frequent of described a length of x+1
Each sequence fragment in fragment Candidate Set, obtains in the frequent fragment Candidate Set of described a length of x+1
Each sequence fragment each place sequence in described R place sequence in frequent degree；
When a sequence fragment in the frequent fragment Candidate Set of described a length of x+1 is in multiple places sequence
Frequent degree in row is more than presetting the probability sum of frequent degree and the plurality of place sequence more than presetting
During probability, determine the frequent fragment that this sequence fragment is an a length of x+1；
Determine that M frequent fragment in whole described frequent fragments is M target sequence fragment.
Optionally, described processor 502 specifically for:
Determine that M Guan Bi frequent fragment in whole described frequent fragments is M target sequence sheet
Section；Wherein, the Guan Bi frequent fragment in described whole described frequent fragment refers to: for described entirely
The frequent fragment of the sub-piece of any one frequent fragment in the described frequent fragment in portion.
Optionally, described processor 502 scans described a length of x+1 in a described place sequence
Frequent fragment Candidate Set in a sequence fragment during, by this tract of automat record
Section frequent degree in this place sequence.
Optionally, described track sets is described user track sets in preset time period.
Optionally, described processor 502 is additionally operable to:
The target sequence fragment of multiple users is clustered, obtains k cluster cluster, k >=1, k
For integer；
Each user is expressed as the user vector with k dimension, a corresponding cluster of dimension
Cluster, the value in a dimension is the target sequence fragment of user in the cluster cluster that this dimension is corresponding
Quantity；
Set up gauss hybrid models, and mix according to Gauss described in the user vector matching of the plurality of user
The parameter of matched moulds type, described gauss hybrid models is made up of multiple Gauss models, a Gauss model pair
Answering a user community, multiple user community corresponding to the plurality of Gauss model are by the plurality of user
Constitute.
Generally, the movement track of user every day can have certain rule (such as working clan,
Fixing movement track is typically had) on weekdays.The device that the embodiment of the present invention provides, permissible
The place sequence that user is corresponding is determined according to the movement track being made up of the historical position data of user, when
One sequence fragment frequent degree in the sequence of multiple places is the biggest and the probability of the plurality of place sequence
When sum is the biggest, illustrate that the probability in user place in this sequence fragment is the biggest.If by default frequency
When the value of numerous degree and predetermined probabilities is reasonably arranged, compared with prior art, real according to the present invention
After the method that executing example provides determines target sequence fragment, the user's determined according to this target sequence fragment
Movement track is relatively accurate, according to this action track be user carry out navigating, place recommend time efficiency relatively
High.
In several embodiments provided herein, it should be understood that disclosed apparatus and method,
Can realize by another way.Such as, device embodiment described above is only schematically
, such as, the division of described module, it is only a kind of logic function and divides, actual permissible when realizing
Have other dividing mode, the most multiple modules or assembly can in conjunction with or be desirably integrated into another
System, or some features can ignore, or do not perform.
The described module illustrated as separating component can be or may not be physically separate,
The parts shown as module can be or may not be physical module, i.e. may be located at a ground
Side, or can also be distributed on multiple NE.Can select therein according to the actual needs
Some or all of unit realizes the purpose of the present embodiment scheme.
It addition, each functional module in each embodiment of the present invention can be integrated in a processing module
In, it is also possible to two or more modules are integrated in a module.Above-mentioned integrated module both may be used
To use the form of hardware to realize, it would however also be possible to employ hardware adds the form of software function module and realizes.
The above-mentioned integrated module realized with the form of software function module, can be stored in a calculating
In machine read/write memory medium.Above-mentioned software function module is stored in a storage medium, if including
Dry instruction is with so that a computer equipment (can be personal computer, server, or network
Equipment etc.) perform the part steps of method described in each embodiment of the present invention.And aforesaid storage medium
Including: USB flash disk, portable hard drive, read only memory (Read-Only Memory is called for short ROM),
Random access memory (Random Access Memory is called for short RAM), magnetic disc or CD
Etc. the various media that can store program code.
Last it is noted that above example is only in order to illustrate technical scheme, rather than right
It limits；Although the present invention being described in detail with reference to previous embodiment, this area common
Skilled artisans appreciate that the technical scheme described in foregoing embodiments still can be repaiied by it
Change, or wherein portion of techniques feature is carried out equivalent；And these amendments or replacement, not
The essence making appropriate technical solution departs from the spirit and scope of various embodiments of the present invention technical scheme.
Claims (12)
1. the method determining user's movement track, it is characterised in that including:
R place sequence is determined according to the track sets being made up of N number of position data of user；Wherein,
One place sequence is made up of N number of place, and the n-th place in described N number of place is described N number of
In the place set that nth position data in position data are corresponding one；One position data correspondence
Place collection be combined into: the fuzzy set in all places in the preset range centered by this position data；1≤n
≤ N, R >=1, n, N and R are integer；
M target sequence fragment is determined according to described R place sequence；Wherein, when a tract
Frequent degree in the section multiple places sequence in described R place sequence is more than presetting frequent degree and institute
When stating the probability sum of multiple places sequence more than predetermined probabilities, this sequence fragment is target sequence fragment,
One sequence fragment frequent degree in a place sequence refers to that this sequence fragment is in this place sequence
The number of times occurred；The probability of one place sequence refers to taking advantage of of the probability in the whole places in this place sequence
Long-pending, the probability in a place refers to the probability in this place in the place belonging to this place is gathered；M >=1,
M is integer；
By the place in any one the target sequence fragment in described M target sequence fragment according to time
Between the route that formed of sequencing series winding as the movement track of described user.
Method the most according to claim 1, it is characterised in that described according to described R place
Sequence determines M target sequence fragment, including:
The frequent fragment Candidate Set of the frequent fragment a length of x+1 of generation according to a length of x, wherein, when
During x=0, the frequent fragment Candidate Set of a length of 1 is: form all differences of described R place sequence
The set in place；X >=0, x is integer；
Each place sequence in described R place sequence scans the frequent of described a length of x+1
Each sequence fragment in fragment Candidate Set, obtains in the frequent fragment Candidate Set of described a length of x+1
Frequent degree in each sequence fragment each place sequence in described R place sequence；
When a sequence fragment in the frequent fragment Candidate Set of described a length of x+1 is in multiple places sequence
Frequent degree in row is general more than presetting more than the probability sum presetting frequent degree and the plurality of place sequence
During rate, determine the frequent fragment that this sequence fragment is an a length of x+1；
Determine that M frequent fragment in whole described frequent fragments is M target sequence fragment.
Method the most according to claim 2, it is characterised in that described determine whole described frequencies
M frequent fragment in numerous fragment is M target sequence fragment, including:
Determine that M Guan Bi frequent fragment in whole described frequent fragments is M target sequence fragment；
Wherein, the Guan Bi frequent fragment in described whole described frequent fragment refers to: not for described whole institute
State the frequent fragment of the sub-piece of any one frequent fragment in frequent fragment.
The most according to the method in claim 2 or 3, it is characterised in that a described place sequence
During row scan a sequence fragment in the frequent fragment Candidate Set of described a length of x+1, logical
Cross this sequence fragment of automat record frequent degree in this place sequence.
5. according to the method described in any one of claim 1-4, it is characterised in that described track sets
For described user track sets in preset time period.
6. according to the method described in any one of claim 1-5, it is characterised in that described method is also wrapped
Include:
The target sequence fragment of multiple users being clustered, obtain k cluster cluster, k >=1, k is
Integer；
Each user is expressed as the user vector with k dimension, a corresponding cluster group of dimension
Bunch, the value in a dimension is the number of the target sequence fragment of user in the cluster cluster that this dimension is corresponding
Amount；
Set up gauss hybrid models, and according to Gaussian Mixture described in the user vector matching of the plurality of user
The parameter of model, described gauss hybrid models is made up of multiple Gauss models, a Gauss model correspondence one
Individual user community, multiple user community that the plurality of Gauss model is corresponding are made up of the plurality of user.
7. the device determining user's movement track, it is characterised in that including:
First determines unit, for determining R according to the track sets being made up of N number of position data of user
Individual place sequence；Wherein, a place sequence is made up of N number of place, n-th in described N number of place
Individual place is in the place set that the nth position data in described N number of position data are corresponding；
Place collection corresponding to one position data is combined into: all fields in the preset range centered by this position data
Fuzzy set；1≤n≤N, R >=1, n, N and R are integer；
Second determines unit, for determining M target sequence fragment according to described R place sequence；
Wherein, big when the frequent degree in the sequence fragment multiple places sequence in described R place sequence
When the probability sum of default frequent degree and the plurality of place sequence is more than predetermined probabilities, this tract
Section is target sequence fragment, and sequence fragment frequent degree in a place sequence refers to this tract
The number of times that section occurs in this place sequence；It is complete that the probability of one place sequence refers in this place sequence
The product of the probability in place, portion, the probability in a place refer to belonging to this place place gather in this
Probability；M >=1, M is integer；
Performance element, for by any one the target sequence fragment in described M target sequence fragment
In the route that formed according to time order and function order series winding of place as the movement track of described user.
Device the most according to claim 7, it is characterised in that described second determines that unit includes:
Signal generating unit, the frequent fragment generating a length of x+1 for the frequent fragment according to a length of x is waited
Selected works, wherein, as x=0, the frequent fragment Candidate Set of a length of 1 is: form described R field
The set of all different places of institute's sequence；X >=0, x is integer；
Scanning element, scans described length in each place sequence in described R place sequence
For each sequence fragment in the frequent fragment Candidate Set of x+1, obtain the frequent sheet of described a length of x+1
Frequency in each sequence fragment in section Candidate Set each place sequence in described R place sequence
Numerous degree；
First determines subelement, for when a sequence in the frequent fragment Candidate Set of described a length of x+1
Column-slice section frequent degree in the sequence of multiple places is more than presetting frequent degree and the plurality of place sequence
When probability sum is more than predetermined probabilities, determine the frequent fragment that this sequence fragment is an a length of x+1；
Second determines subelement, for determining that M frequent fragment in whole described frequent fragments is
M target sequence fragment.
Device the most according to claim 8, it is characterised in that described second determines that subelement has
Body is used for:
Determine that M Guan Bi frequent fragment in whole described frequent fragments is M target sequence fragment；
Wherein, the Guan Bi frequent fragment in described whole described frequent fragment refers to: not for described whole institute
State the frequent fragment of the sub-piece of any one frequent fragment in frequent fragment.
Device the most according to claim 8 or claim 9, it is characterised in that described scanning element is one
Individual described place sequence scans a tract in the frequent fragment Candidate Set of described a length of x+1
During Duan, by this sequence fragment of automat record frequent degree in this place sequence.
11. according to the device described in any one of claim 7-10, it is characterised in that described track sequence
It is classified as described user track sets in preset time period.
12. according to the device described in any one of claim 7-11, it is characterised in that described device is also
Including:
Cluster cell, for the target sequence fragment of multiple users being clustered, obtains k cluster group
Bunch, k >=1, k is integer；
Represent unit, for each user being expressed as the user vector with k dimension, a dimension
A corresponding cluster cluster, the value in a dimension is the mesh of user in the cluster cluster that this dimension is corresponding
The quantity of mark sequence fragment；
Computing unit, is used for setting up gauss hybrid models, and intends according to the user vector of the plurality of user
Closing the parameter of described gauss hybrid models, described gauss hybrid models is made up of multiple Gauss models, one
The corresponding user community of Gauss model, multiple user community corresponding to the plurality of Gauss model are by described
Multiple users are constituted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510406207.3A CN106327236B (en) | 2015-07-10 | 2015-07-10 | Method and device for determining action track of user |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510406207.3A CN106327236B (en) | 2015-07-10 | 2015-07-10 | Method and device for determining action track of user |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106327236A true CN106327236A (en) | 2017-01-11 |
| CN106327236B CN106327236B (en) | 2020-04-14 |
Family
ID=57725345
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510406207.3A Active CN106327236B (en) | 2015-07-10 | 2015-07-10 | Method and device for determining action track of user |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106327236B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108647735A (en) * | 2018-05-15 | 2018-10-12 | 广州杰赛科技股份有限公司 | User's trip law analytical method, device, equipment and storage medium |
| CN108921876A (en) * | 2018-07-10 | 2018-11-30 | 北京旷视科技有限公司 | Method for processing video frequency, device and system and storage medium |
| WO2020057275A1 (en) * | 2018-09-17 | 2020-03-26 | 北京京东尚科信息技术有限公司 | Trajectory determination method and apparatus, and time recommendation method, apparatus and system |
| CN113286333A (en) * | 2020-02-19 | 2021-08-20 | 华为技术有限公司 | Network selection method and device |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104239556A (en) * | 2014-09-25 | 2014-12-24 | 西安理工大学 | Density clustering-based self-adaptive trajectory prediction method |
-
2015
- 2015-07-10 CN CN201510406207.3A patent/CN106327236B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104239556A (en) * | 2014-09-25 | 2014-12-24 | 西安理工大学 | Density clustering-based self-adaptive trajectory prediction method |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108647735A (en) * | 2018-05-15 | 2018-10-12 | 广州杰赛科技股份有限公司 | User's trip law analytical method, device, equipment and storage medium |
| CN108647735B (en) * | 2018-05-15 | 2021-01-12 | 广州杰赛科技股份有限公司 | User travel rule analysis method, device, equipment and storage medium |
| CN108921876A (en) * | 2018-07-10 | 2018-11-30 | 北京旷视科技有限公司 | Method for processing video frequency, device and system and storage medium |
| WO2020057275A1 (en) * | 2018-09-17 | 2020-03-26 | 北京京东尚科信息技术有限公司 | Trajectory determination method and apparatus, and time recommendation method, apparatus and system |
| CN113286333A (en) * | 2020-02-19 | 2021-08-20 | 华为技术有限公司 | Network selection method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106327236B (en) | 2020-04-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102102992B (en) | Multistage network division-based preliminary screening method for matched roads and map matching system | |
| EP3822880A1 (en) | Load prediction method and apparatus based on neural network | |
| CN101210825B (en) | Map information generating systems | |
| CN111854779B (en) | Route planning method and device, electronic equipment and readable storage medium | |
| CN105528955B (en) | Generate the method and its device of road network | |
| CN106327236A (en) | Method and device for determining user action track | |
| CN104270714A (en) | Method and device for determining action trails of user | |
| CN104320789B (en) | Game theory-based vehicle networking RSU optimal configuration method | |
| CN108073727A (en) | The data processing method and device of place search | |
| CN104270818A (en) | Wireless positioning method for vehicles in park | |
| CN106595683A (en) | Navigation method and device | |
| CN111814079A (en) | Crowd sensing Internet of vehicles user screening method combining clustering and CMAB | |
| CN111832869A (en) | Vehicle scheduling method and device, electronic equipment and storage medium | |
| CN106980924A (en) | Electric car charging and conversion electric resource recommendation method and device based on requirement forecasting | |
| CN108510050A (en) | It is a kind of based on shuffling the feature selection approach to leapfrog | |
| CN113706220A (en) | User portrait determination, user demand prediction method, and data processing system | |
| CN111415035A (en) | Method and device for pre-estimating building function house type ratio | |
| CN103971136A (en) | Large-scale data-oriented parallel structured support vector machine classification method | |
| CN107704942A (en) | Distribution path determines method, apparatus and equipment | |
| CN105096589B (en) | Method of selecting representative node in traffic road, system and client | |
| CN108985488A (en) | The method predicted to individual trip purpose | |
| CN115620252A (en) | Trajectory rectification method and device, computer equipment and storage medium | |
| CN113358127A (en) | Road weight determination method, model generation method, route planning method and route planning device | |
| CN105643944A (en) | 3D Printer stability control method and control system | |
| CN107124727A (en) | A kind of PCI optimization method and devices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |