CN110503487A - One kind getting through method based on the convergent shopping center data in path - Google Patents
One kind getting through method based on the convergent shopping center data in path Download PDFInfo
- Publication number
- CN110503487A CN110503487A CN201910801211.8A CN201910801211A CN110503487A CN 110503487 A CN110503487 A CN 110503487A CN 201910801211 A CN201910801211 A CN 201910801211A CN 110503487 A CN110503487 A CN 110503487A
- Authority
- CN
- China
- Prior art keywords
- path
- label
- personid
- entity
- track
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000002372 labelling Methods 0.000 claims description 5
- 230000003542 behavioural effect Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 239000000463 material Substances 0.000 claims description 3
- 230000000452 restraining effect Effects 0.000 claims description 3
- 230000002123 temporal effect Effects 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims 1
- 238000004458 analytical method Methods 0.000 abstract description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
Abstract
The invention discloses one kind to get through method based on the convergent shopping center data in path, the method for being primarily based on cv obtains the track at the mall of consumer, ID under unique line is created with portrait itself, it is denoted as PersionID, entity ID under mark line simultaneously, then, following information is by priori IDization: being used as metamessage after shops, lift port, parking lot inlet and outlet IDization in advance in kind.When consumer enters shopping center, the analysis that cv will be passed through first will be in the User ID to storage personID, it will create new personID, with consumer at the mall in stroll about, automatically it can will stroll about track and entity information is mapped on corresponding path, and starts convergent iterations component and update path information and label weight, the unique path and corresponding precisely label that the consumer will eventually be generated according to preset threshold value draw a portrait, the progress lean operation of these information, which can be used, in operation personnel and algorithm personnel needs shopping experience to promote consumer.
Description
Technical field
The present invention relates to the digitized application fields of shopping center field of line scape, more particularly to one kind to be based on the convergent purchase in path
Object centre data gets through method.
Background technique
Past 10 years, shopping center was tided in one the comparatively development process of more steady high speed,
Chinese personal consumption and retail business quickly increase at an unprecedented rate, and shopping center is accounted in the retail of Chinese society
Than being up to 40%, from the point of view of fabric, the essence in shopping center is to connect trade company and consumer and existing, reference line
Upper retail, also just gradually toward the conversion of fining direction, the premise of fining, which seeks to more understand, to disappear for the operation in shopping center
Fei Zhe, constructs the clear portrait of consumer to realize precision marketing, this mode be sold on line developed it is very mature, respectively
Big electric business constructs fine consumer representation according to the action trail of consumer on line, and based on this provide personalized recommendation,
A variety of precision marketings such as accurate advertisement realize the equilibrium and maximization of benefit and experience, reference line with better service consumer
On method, shopping center is scene under a similar line, unlike consumer behaviour digitization difficulty under line it is very big,
The prior art is mostly that consumer behaviour is digitized in a manner of Wi-Fi, position etc., and this kind of technology can only generally get PV rank
Information, that is, person-time for coming shopping center can only be got, whom this specific people is, concrete behavior at the mall is then
It can not know.
The information that such methods are realized is too thick spacious, refines the behavior details of consumer on no image of Buddha line like that, finally
It cannot achieve accurate portrait, lean operation under the drawing of data is logical and line on the line of a more crucial step and under line thereupon
It does not know where to begin, the logical drawing of on-line off-line data is in order to which the touching of consumer reaches.
Summary of the invention
For above-mentioned deficiency in the prior art, the present invention provides one kind to be beaten based on the convergent shopping center data in path
Circulation method, the technology that more refines realize the digitlization in shopping center, and the drawing for being dedicated to solving behavior under information and line on line is logical,
Portrait after unified ID and drawing are logical.
In order to achieve the above object of the invention, the present invention use the specific scheme is that
One kind getting through method based on the convergent shopping center data in path, the method for being primarily based on cv obtain consumer
The track in shopping center creates ID under unique line with portrait itself, is denoted as PersionID, while entity ID under mark line, so
Afterwards, following information is by priori IDization: being used as metamessage, these yuan after shops, lift port, parking lot inlet and outlet IDization in advance in kind
Information will draw logical and portrait to provide key parameter to be subsequent.
Further, this method is specifically divided into following steps:
(1) track and entity IDization under consumer's line, estimation ID are recorded as personID, and entity ID is recorded as
EntityID；
(2) behavioural information on line is obtained, ID and time of the act, space (material object that behavior occurs, such as shops) on call wire,
ID is recorded as OnlineID and using the corresponding entity ID of the first step on line；
(3) on-line off-line co-route is created, path is recorded as, a path includes one persionID and one
OnlineID, each ID carry detailed information, the shops ID consumed on time point of the shops ID as where track and appearance, line
Information and consumption time, specifically, each path corresponding personID and onlineID includes weight information, specifically
Such as: " path1 ": " personID1 ": " weight ": 0.4, " EntityID1 ": " 002 ", " onlineID1 ":
{"weight":0.2,"EntityID1":"001"}}}；
(4) implement iteration module, Iterative path path is until restrain, specifically, each path dimension in the iteration module of path
One group of parameter and an onlineID are protected, the combination of personID and each path can form a contextual feature, for
For one personID, the expected revenus that can be obtained on each path is as follows:
Specifically, x indicates the feature of personID, i.e. entity ID and the temporal information of generation included in track, r are indicated
The income of personID and the path, i.e., when co-occurrence occurs for the onlineID in personID and path, income is 1, when some
Time point personID is appeared under some entity ID, does not occur the information of onlineID under the path at this time, then income is
0, when personID is appeared at some time point under some entity ID, the corresponding onlineID of the path occurs elsewhere,
Then income at this time is that -1, theta then indicates the convergence parameter between personID and the path, from the interpretation of mathematics,
There are linear relationships between each personID and each path, assume that the present invention by solving following loss, asks based on this
The hand for obtaining each path holds back parametric optimal solution:
Loss=(Ca-Daθa)2+λ||θa| |, specific analytic solutions are as follows:
Other than obtaining desired value, the present invention also needs a top confidence limit,Based on this, this hair
The convergence path of bright available each personID:
And the corresponding onlineID of the path
OnlineID on logical line as is drawn with the personID, has used following information during calculating and restraining path:
Contextual feature, i.e. entity information and time interval involved in the track personID correspond to onlineID in path and exist
The position when track personID occurs, these information are all to store every time, therefore after having collected certain information,
Parameter can dynamically update, the convergence of realizing route, to find optimal onlineID；
(5) path portrait is realized, it, such as can be by Startbuck's labeling by entity tag are as follows: white collar, coffee, leisure etc.,
It, then can be specifically of the invention by with the corresponding label mapping to consumer of Startbuck if consumer occurred in Startbuck
Respective algorithms are provided for labeling more accurately to describe the consumer, eliminate randomness and real-time update label weight,
Wilson's section-label confidence is calculated for the corresponding each label of each path, label is increased newly and repetition label all can more cenotype
The confidence level answered selects the corresponding label of path eventually by confidence level；
(6) rule tree of establishing label:
{ " domain ": " abstract concept for representing the rule ", " concepts ": the [" expression (entity of the rule
Collection) "], " children ": [" the sub-rule List of the rule "] }, wherein domain is tag set, concepts: the label
Subset, children: similar label；
(7) Bayesian probability that track belongs to some label is calculated:
Pr (S | W) it is the probability that some entity W is label S in the set of track, if peace steps on the probability that shop belongs to sport,
Pr (S) be to fourth arbitrary trajectory in the case where label S probability,
Pr (W | S) it is the probability that some entity appears in label S in track,
Pr (H) is the not prior probability in label S in the case where given arbitrary trajectory,
Pr (W H) is that the probability of label S does not occur in some entity W in track,
Belong to the probability of any label node in track: Wilson_score_interval*p (s | w) Wilson_
Score_interval calculation method is as follows:
The turnout (i.e. the ratio that a track belongs to the co-occurrence entity number of the path) of each label is calculated,
The low confidence interval of each turnout (95%) is calculated,
Confidence level according to the lower limit value of confidence interval as the label,
Principle is explained: the width of confidence interval and the quantity of sample are related, for example, belonging to the reality of A label in some track
Body ID 1,9 are other labels；Another track ownership B label has 3, and 27 are other labels, two labels
Ratio of voting all is 10%, but the confidence level of B is higher than A.
The invention has the benefit that
When consumer enters shopping center, it will the User IDization is arrived storage personID by the analysis of cv first
On, if do not found in historical information, will create new personID, with consumer at the mall in stroll about,
The implemented algorithm of the present invention will can stroll about track automatically and entity information is mapped on corresponding path, and start convergent iterations group
Part updates path information and label weight, and the unique path and phase of the consumer will eventually be generated according to preset threshold value
The accurate label answered is drawn a portrait, and operation personnel and algorithm personnel can be used these information and carry out lean operation to promote consumption
Person needs shopping experience.
Specific embodiment
The present invention is further described below by way of specific embodiment, but the present invention is not limited only to following embodiment.At this
In the range of invention or the contents of the present invention are not being departed from, in spirit and scope, the change that carries out to the present invention is combined or replaced
It changes, will be apparent to the person skilled in the art, and be included within the scope of the present invention.
One kind getting through method based on the convergent shopping center data in path, the method for being primarily based on cv obtain consumer
The track in shopping center creates ID under unique line with portrait itself, is denoted as PersionID, while entity ID under mark line, so
Afterwards, following information is by priori IDization: being used as metamessage, these yuan after shops, lift port, parking lot inlet and outlet IDization in advance in kind
Information will draw logical and portrait to provide key parameter to be subsequent.
This method is specifically divided into following steps:
(1) track and entity IDization under consumer's line, estimation ID are recorded as personID, and entity ID is recorded as
EntityID；
(2) behavioural information on line is obtained, ID and time of the act, space (material object that behavior occurs, such as shops) on call wire,
ID is recorded as OnlineID and using the corresponding entity ID of the first step on line；
(3) on-line off-line co-route is created, path is recorded as, a path includes one persionID and one
OnlineID, each ID carry detailed information, the shops ID consumed on time point of the shops ID as where track and appearance, line
Information and consumption time, specifically, each path corresponding personID and onlineID includes weight information, specifically
Such as: " path1 ": " personID1 ": " weight ": 0.4, " EntityID1 ": " 002 ", " onlineID1 ":
{"weight":0.2,"EntityID1":"001"}}}；
(4) implement iteration module, Iterative path path is until restrain, specifically, each path dimension in the iteration module of path
One group of parameter and an onlineID are protected, the combination of personID and each path can form a contextual feature, for
For one personID, the expected revenus that can be obtained on each path is as follows:Specifically, x indicates the feature of personID, i.e. track
Included in entity ID and the temporal information of generation, r indicates the income of personID and the path, i.e., as personID and
It is 1 that income when co-occurrence, which occurs, for onlineID in path, when some time point personID is appeared under some entity ID, at this time
There is not the information of onlineID under the path, then income is 0, when personID appears in some entity at some time point
When under ID, the corresponding onlineID of the path occurs elsewhere, then income at this time be -1, theta then indicate personID with
Convergence parameter between the path, from the interpretation of mathematics, there are linear relationship between each personID and each path,
Assume that the present invention holds back parametric optimal solution by solving following loss, the hand for acquiring each path based on this:
Loss=(Ca-Daθa)2+λ||θa| |, specific analytic solutions are as follows:
Other than obtaining desired value, the present invention also needs a top confidence limit,Based on this, the present invention can be obtained
To the convergence path of each personID:
And the corresponding onlineID of the path
OnlineID on logical line as is drawn with the personID, has used following information during calculating and restraining path:
Contextual feature, i.e. entity information and time interval involved in the track personID correspond to onlineID in path and exist
The position when track personID occurs, these information are all to store every time, therefore after having collected certain information,
Parameter can dynamically update, the convergence of realizing route, to find optimal onlineID；
(5) path portrait is realized, it, such as can be by Startbuck's labeling by entity tag are as follows: white collar, coffee, leisure etc.,
It, then can be specifically of the invention by with the corresponding label mapping to consumer of Startbuck if consumer occurred in Startbuck
Respective algorithms are provided for labeling more accurately to describe the consumer, eliminate randomness and real-time update label weight,
Wilson's section-label confidence is calculated for the corresponding each label of each path, label is increased newly and repetition label all can more cenotype
The confidence level answered selects the corresponding label of path eventually by confidence level；
(6) rule tree of establishing label:
{ " domain ": " abstract concept for representing the rule ", " concepts ": the [" expression (entity of the rule
Collection) "], " children ": [" the sub-rule List of the rule "] }, wherein domain is tag set, concepts: the label
Subset, children: similar label；
(7) Bayesian probability that track belongs to some label is calculated:
Pr (S | W) it is the probability that some entity W is label S in the set of track, if peace steps on the probability that shop belongs to sport,
Pr (S) be to fourth arbitrary trajectory in the case where label S probability,
Pr (W | S) it is the probability that some entity appears in label S in track,
Pr (H) is the not prior probability in label S in the case where given arbitrary trajectory,
Pr (W H)) it is that the probability of label S does not occur in some entity W in track,
Belong to the probability of any label node in track: Wilson_score_interval*p (s | w) Wilson_
Score_interval calculation method is as follows:
The turnout (i.e. the ratio that a track belongs to the co-occurrence entity number of the path) of each label is calculated,
The low confidence interval of each turnout (95%) is calculated,
Confidence level according to the lower limit value of confidence interval as the label,
Principle is explained: the width of confidence interval and the quantity of sample are related, for example, belonging to the reality of A label in some track
Body ID 1,9 are other labels；Another track ownership B label has 3, and 27 are other labels, two labels
Ratio of voting all is 10%, but the confidence level of B is higher than A.
The above description is only a preferred embodiment of the patent of the present invention, is not intended to limit the invention patent, all at this
Made any modifications, equivalent replacements, and improvements etc., should be included in the invention patent within the spirit and principle of patent of invention
Protection scope within.
Claims (2)
1. one kind gets through method based on the convergent shopping center data in path, which is characterized in that the method for being primarily based on cv obtains
The track at the mall of consumer creates ID under unique line with portrait itself, is denoted as PersionID, while mark line
Lower entity ID, then, following information is by priori IDization: as member after shops, lift port, parking lot inlet and outlet IDization in advance in kind
Information, these metamessages will draw logical and portrait to provide key parameter to be subsequent.
2. a kind of convergent shopping center data in path that are based on according to claim 1 get through method, which is characterized in that should
Method is specifically divided into following steps:
(1) track and entity IDization under consumer's line, estimation ID are recorded as personID, and entity ID is recorded as EntityID；
(2) behavioural information on line is obtained, ID and time of the act, space (material object that behavior occurs, such as shops) on call wire, on line
ID is recorded as OnlineID and using the corresponding entity ID of the first step；
(3) on-line off-line co-route is created, path is recorded as, a path includes one persionID and one
OnlineID, each ID carry detailed information, the shops ID consumed on time point of the shops ID as where track and appearance, line
Information and consumption time, specifically, each path corresponding personID and onlineID includes weight information, specifically
Such as: " path1 ": " personID1 ": " weight ": 0.4, " EntityID1 ": " 002 ", " onlineID1 ":
{"weight":0.2,"EntityID1":"001"}}}；
(4) implement iteration module, Iterative path path is until restrain, specifically, each path maintenance one in the iteration module of path
Group parameter and an onlineID, the combination of personID and each path can form a contextual feature, for one
For personID, the expected revenus that can be obtained on each path is as follows:Specifically, x indicates the feature of personID, i.e., included in track
Entity ID and generation temporal information, r indicates the income of personID and the path, i.e., when in personID and path
It is 1 that income when co-occurrence, which occurs, for onlineID, when some time point personID is appeared under some entity ID, is not occurred at this time
The information of onlineID under the path, then income is 0, when personID is appeared at some time point under some entity ID,
The corresponding onlineID of the path occurs elsewhere, then income at this time be -1, theta then indicate the personID and path it
Between convergence parameter, it is false based on this there are linear relationship between each personID and each path from the interpretation of mathematics
If the present invention holds back parametric optimal solution by solving following loss, the hand for acquiring each path:
Loss=(Ca-Daθa)2+λ||θa| |, specific analytic solutions are as follows:
Other than obtaining desired value, the present invention also needs a top confidence limit,
It is of the invention based on this
The convergence path of available each personID:
And the corresponding onlineID of the path is and this
PersonID draws onlineID on logical line, and following information has been used during calculating and restraining path: context is special
It levies, i.e. entity information and time interval involved in the track personID, corresponds to onlineID in the track personID in path
Position when appearance, these information are all to store every time, therefore after having collected certain information, parameter can dynamic
It updates, the convergence of realizing route, to find optimal onlineID；
(5) path portrait is realized, it, such as can be by Startbuck's labeling are as follows: white collar, coffee, leisure etc. such as disappear by entity tag
Expense person occurred in Startbuck, then can be by with the corresponding label mapping to consumer of Startbuck, and specifically the present invention is mark
Labelization provide respective algorithms more accurately to describe the consumer, eliminate randomness and real-time update label weight, are every
The corresponding each label of a path calculates Wilson's section-label confidence, newly-increased label and repeats label and can all update accordingly
Confidence level selects the corresponding label of path eventually by confidence level；
(6) rule tree of establishing label:
" domain ": " abstract concept for representing the rule ", " concepts ": [" expression (entity set) of the rule "], "
The sub-rule List " of children ": the [" rule] }, wherein domain is tag set, concepts: the subset of the label,
Children: similar label；
(7) Bayesian probability that track belongs to some label is calculated:
Pr (S | W) it is the probability that some entity W is label S in the set of track, if peace steps on the probability that shop belongs to sport,
Pr (S) be to fourth arbitrary trajectory in the case where label S probability,
Pr (W | S) it is the probability that some entity appears in label S in track,
Pr (H) is the not prior probability in label S in the case where given arbitrary trajectory,
Pr (W H) is that the probability of label S does not occur in some entity W in track,
Belong to the probability of any label node in track: Wilson_score_interval*p (s | w) Wilson_score_
Interval calculation method is as follows:
The turnout (i.e. the ratio that a track belongs to the co-occurrence entity number of the path) of each label is calculated,
The low confidence interval of each turnout (95%) is calculated,
Confidence level according to the lower limit value of confidence interval as the label,
Principle is explained: the width of confidence interval and the quantity of sample are related, for example, belonging to the entity ID of A label in some track
1,9 are other labels；Another track ownership B label has 3, and 27 are other labels, the ballot of two labels
Ratio is all 10%, but the confidence level of B is higher than A.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910801211.8A CN110503487A (en) | 2019-08-28 | 2019-08-28 | One kind getting through method based on the convergent shopping center data in path |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910801211.8A CN110503487A (en) | 2019-08-28 | 2019-08-28 | One kind getting through method based on the convergent shopping center data in path |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110503487A true CN110503487A (en) | 2019-11-26 |
Family
ID=68588584
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910801211.8A Pending CN110503487A (en) | 2019-08-28 | 2019-08-28 | One kind getting through method based on the convergent shopping center data in path |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110503487A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017127571A1 (en) * | 2016-01-19 | 2017-07-27 | Magic Leap, Inc. | Augmented reality systems and methods utilizing reflections |
| CN107767168A (en) * | 2017-09-19 | 2018-03-06 | 神策网络科技（北京）有限公司 | User behavior data processing method and processing device, electronic equipment and storage medium |
| CN109359180A (en) * | 2018-09-20 | 2019-02-19 | 腾讯科技（深圳）有限公司 | User's portrait generation method, device, electronic equipment and computer-readable medium |
-
2019
- 2019-08-28 CN CN201910801211.8A patent/CN110503487A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017127571A1 (en) * | 2016-01-19 | 2017-07-27 | Magic Leap, Inc. | Augmented reality systems and methods utilizing reflections |
| CN107767168A (en) * | 2017-09-19 | 2018-03-06 | 神策网络科技（北京）有限公司 | User behavior data processing method and processing device, electronic equipment and storage medium |
| CN109359180A (en) * | 2018-09-20 | 2019-02-19 | 腾讯科技（深圳）有限公司 | User's portrait generation method, device, electronic equipment and computer-readable medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu | Combining GIS and the Huff Model to Analyze Suitable Locations for a New Asian Supermarket in the Minneapolis and St. Paul, Minnesota USA | |
| CN104991924B (en) | Method and apparatus for the address for determining new supply centre | |
| Tanahashi et al. | Design considerations for optimizing storyline visualizations | |
| Hu et al. | A framework for extracting urban functional regions based on multiprototype word embeddings using points-of-interest data | |
| CN104699818A (en) | Multi-source heterogeneous multi-attribute POI (point of interest) integration method | |
| Gao et al. | Modelling and application of fuzzy adaptive minimum spanning tree in tourism agglomeration area division | |
| Archambault | Structural differences between two graphs through hierarchies | |
| CN111831899B (en) | Navigation interest point recommendation method, device, server and readable storage medium | |
| CN111932318B (en) | Region division method and device, electronic equipment and computer readable storage medium | |
| CN108256590A (en) | A kind of similar traveler recognition methods based on compound first path | |
| CN109903095A (en) | Data processing method, device, electronic equipment and computer readable storage medium | |
| CN110443288A (en) | A kind of track similarity calculation method based on sequence study | |
| CN103455555B (en) | Recommendation method and recommendation apparatus based on mobile terminal similarity | |
| CN109903138A (en) | A kind of individual commodity recommendation method | |
| CN108959643A (en) | Generate method, apparatus, server and the storage medium of label | |
| Wang et al. | Group-based personalized location recommendation on social networks | |
| CN109345946A (en) | Map data displaying method, device, electronic equipment and computer storage medium | |
| CN109282826A (en) | The method and its system of map multi-point path planning are carried out based on polymerization methods | |
| Nguyen et al. | DBSTexC: Density-based spatio-textual clustering on twitter | |
| Zhu et al. | Probabilistic routing using multimodal data | |
| CN107133835A (en) | A kind of method and device for analyzing commercial quality | |
| CN110503487A (en) | One kind getting through method based on the convergent shopping center data in path | |
| CN109272351A (en) | Passenger flow moving-wire and passenger flow hot-zone determine method and device | |
| Pratiwi et al. | Digital Capability and Communication Skill for Empowering Self-Efficacy in Tourism Industry | |
| CN107291860A (en) | Seed user determines method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191126 |