Driving destination recommendation method and device

A destination and data technology, applied in the field of information processing, can solve the problem that the driving destination recommendation method cannot be recommended

Inactive Publication Date: 2018-07-03
武汉四维图新科技有限公司
12 Cites 5 Cited by

AI-Extracted Technical Summary

Problems solved by technology

[0003] Therefore, the technical problem to be solved by the present invention is that the existing driving destination recommendation met...
View more

Method used

According to the driving history trajectory data of the current user, the driving familiar area of ​​the current user is obtained, by calculating the time similarity and track similarity between the current user and other users in the driving familiar area of ​​the current user, and then obtaining the current user's driving familiarity with the current user. Similar users in the driving familiar area are recommended to the current user according to the POI data of similar users in the similar user set, which improves the current user's familiarity with the route to the POI data.
By classifying the POI data set that the obtained current user seldom goes, obtain the subset of each category that the users of different categories seldom go, classify more POIs that similar users go according to the same category, ...
View more

Abstract

The invention discloses a driving destination recommendation method and device. The driving destination recommendation method comprises the following steps of: obtaining a similar user set according to driving history track data of the current user; obtaining history interest point data of similar users in the similar user set; and determining driving destination interest point data recommended tothe current user according to history interest point data of the current user and the history interest point data of the similar users in the similar user set. According to the method, the problem that existing driving destination recommendation methods cannot carry out recommendation according to destination quality and familiarities, to paths for arriving at destinations, of users is solved.

Application Domain

Special data processing applications

Technology Topic

Point dataComputer network +4

Image

  • Driving destination recommendation method and device
  • Driving destination recommendation method and device
  • Driving destination recommendation method and device

Examples

  • Experimental program(1)

Example Embodiment

[0066] The technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Apparently, the described embodiments are some of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.
[0067] An embodiment of the present invention provides a driving destination recommendation method, such as figure 1 shown, including:
[0068] S11. Obtain a set of similar users according to the current user's driving history trajectory data; wherein, through matching the current user's driving history trajectory data with the navigation electronic map, the user's driving history trajectory record database Da is obtained, and the driving history trajectory record database Da includes but It is not limited to the user number, the number of each road passed, the geographic coordinate information, and the time period of the road passed.
[0069] S12. Obtain the historical point of interest data of similar users in the similar user set; POI (Point of Interest, hereinafter referred to as POI), wherein the POI data of similar users is obtained according to the navigation electronic map of similar users. The existing navigation electronic map includes The information includes road data, background data (including point, line, area), annotation data (including annotation type, visibility level) and index data (including: POI, zip code, intersection), etc. Each POI data includes: Name, type, longitude and latitude information, obtain POI data of similar users within a period of time or POI data of similar users in a fixed area within a period of time to obtain historical POI data of similar users.
[0070] S13, according to the historical point-of-interest data of the current user and the historical point-of-interest data of similar users in the similar user set, determine the point-of-interest data of the driving destination recommended to the current user. Through a large amount of driving history POI data of similar users, the driving history trajectory data of the current user is fused, and then the destinations that similar users frequently go to and the destinations that the user does not often go to or has not been to are recommended to the current user.
[0071] The driving destination recommendation method provided by the embodiment of the present invention obtains a similar user set according to the current user's driving history trajectory data, and then obtains the historical POI data of similar users in the similar user set according to the similar user set, and obtains the historical POI data of similar users in the similar user set according to the current user's history POI data and historical POI data of similar users in the similar user set determine the driving destination POI data recommended to the current user, and similar users obtained through the current user's driving history trajectory improve the current user's historical interest in going to similar users At the same time, what is recommended to the current user is the point-of-interest data of similar users, which improves the accuracy of the recommended results to the current user, and then solves the problem that the existing driving destination recommendation method cannot be based on the advantages and disadvantages of the destination and the user's reach. The familiarity of the destination path for recommendation problems.
[0072] As a specific implementation, such as figure 2 As shown, step S11 specifically includes the following steps:
[0073] S111. Obtain the driving familiar area of ​​the current user according to the driving history track data of the current user;
[0074] S112, acquiring historical driving trajectory data of other users in the familiar driving area;
[0075] S113, according to the current user's driving history trajectory data and other users' driving history trajectory data, calculate the trajectory similarity and time similarity between the current user and other users in the driving familiar area;
[0076] Among them, the trajectory similarity calculation is obtained by the following formula:
[0077]
[0078] Among them, J an is the trajectory similarity between user A and the Nth user, A(u) represents all the road number sets passed by user A, and N(u) represents all the road number sets passed by other user N.
[0079] The temporal similarity calculation is obtained by the following formula:
[0080] T an =1-(|A(P dw )-N(P dw )|+|A(P nw )-N(P nw )|+|A(P di )-N(P di )|
[0081] +|A(P ni )-N(P ni )|)
[0082] Among them, T an is the temporal similarity between user A and user N, A(P dw ), N(P dw ) are the travel frequencies of user A and other users N during daytime and weekdays respectively, A(P nw ), N(P nw ) are respectively the travel frequencies of user A and other users N at night and on weekdays, A(P di ), N(P di ) are the travel frequencies of user A and other users N during daytime and non-working days, respectively, A(P ni ), N(P ni ) are the travel frequencies of user A and other users N at night and non-working days, respectively.
[0083] In the time similarity calculation, the travel time of the user is divided into four dimensions: daytime, night, working day and non-working day, and the user’s travel frequency P dw ,P nw ,P di ,P ni The calculation method is as follows:
[0084]
[0085] S114. Obtain a set of similar users according to the calculation results of trajectory similarity and time similarity.
[0086] According to the current user's driving history trajectory data, the current user's driving familiar area is obtained. By calculating the time similarity and trajectory similarity between the current user and other users in the current user's driving familiar area, and then obtaining the current user's driving familiar area The similar users in the similar user set are recommended to the current user according to the similar user POI data in the similar user set, which improves the current user's familiarity with the route to the POI data.
[0087] Acquiring the driving familiar area in step S111 specifically includes the following steps:
[0088] S1111. Obtain geographical coordinate information in the historical driving track data of the current user; obtain the longitude and latitude information of the geographical coordinates of the roads passed by the current user according to the historical driving POI data of the current user in the navigation electronic map historical records.
[0089] S1112. According to the geographic coordinate information, determine a rectangular area familiar to the user when driving. For example, according to the geographic latitude coordinate information, according to the maximum geographic latitude coordinate information and the minimum geographic latitude coordinate information, the latitude range of the current user's driving track is obtained, and the longitude range of the geographic longitude coordinate information is similarly obtained, and the latitude range and longitude range are respectively used as rectangles The two sides of the area get a rectangular area familiar to the user when driving.
[0090] Step S114 specifically includes:
[0091] S1141. According to the calculation results of trajectory similarity and time similarity, respectively obtain a trajectory similar user set and a time similar user set; that is, obtain a trajectory similarity set H(J a1 , J a2 ,...J an ) and time similarity set K(T a1 ,T a2 ,...,T an );
[0092] S1142. Arranging the similarities of similar users in the trajectory similarity set and the time similarity set respectively;
[0093] S1143, according to the arrangement result, obtain a subset of users with similar trajectories and a subset of users with similar time; arrange the similarities of similar users in the trajectory similarity set from large to small, in order to improve the accuracy of driving destination recommendation results, Select a preset number of similar users with relatively large trajectory similarity to form a user subset with similar trajectory. Similarly, select a preset number of temporally similar users with relatively large temporal similarity in the temporal similarity set to form a temporally similar user subset.
[0094] S1144. Obtain a similar user set according to the trajectory similar user subset and the time similar user subset. For example, if the number of similar users included in the post-trajectory similar user subset is 10, and the number of similar users included in the time similar user subset is 10, then 20 similar similar users are obtained.
[0095] As an optional implementation manner, step S1144 specifically includes:
[0096] Obtain the weight coefficients of the user sub-set with similar trajectory and the user sub-set with similar time respectively, and the weight coefficient is obtained according to the evaluation data of the current user;
[0097] According to the weight coefficient, a similar user set corresponding to the weight coefficient is obtained.
[0098]For example, according to the preset weight coefficients of the user subset with similar trajectories and the subset of users with similar time The number of users is 10, and the number of similar users contained in the time similar user subset is 10, then the number of similar users contained in the obtained similar user set is (a*10+b*10); when using The POIs frequented by similar users in the obtained similar user set are recommended to the current user. After the obtained user evaluation data shows that it cannot meet the user's requirements, the trajectory similar user subset is changed according to the obtained evaluation data of the current user. , the weight coefficients of the time-similar user subsets to obtain a' and b', then the number of similar users contained in the obtained similar user set is (a'*10+b'*10).
[0099] In the familiar area of ​​the user's driving, the similar user set is obtained through the calculation of the trajectory similarity and time similarity, which improves the credibility of the obtained similar user set result, and uses the similar user POI in the similar user set obtained to the current user. Make recommendations to improve the accuracy of recommended driving destinations.
[0100] As a specific implementation manner, step S13 specifically includes the following steps:
[0101] S131. According to the historical point-of-interest data of the current user, obtain a set of first point-of-interest data whose number of times the current user goes to different points of interest is less than a first preset frequency;
[0102] S132. According to the historical point-of-interest data of similar users, obtain a set of second point-of-interest data whose times of similar users going to different points of interest are greater than a second preset frequency;
[0103] S133, according to the set of the first point of interest data and the set of the second point of interest data, determine the point of interest data of the driving destination recommended to the current user.
[0104] For example, let user A's historical POI data collection be A all , to obtain the set A of the first POI data of user A satisfying the first preset frequency by filtering low;
[0105] Let the historical POI data collection of all the similar users G be G all , to obtain a set G of the second POI data of similar users G satisfying the second preset frequency by filtering high :
[0106] Then the POI data set P of the driving destination to be recommended can be calculated by the following formula w :
[0107] P w =A low ∩G high +(G all -G all ∩A all )∩G high
[0108] As an optional implementation manner, step S133 specifically includes the following steps:
[0109] S1331. Classify the data of the set of the first point of interest data to obtain a subset of the first point of interest data of at least one category;
[0110] S1332, according to the category of the subset of the first point of interest data, find a subset of the second point of interest data consistent with the category in the second set of point of interest data;
[0111] S1333. According to the category of the driving destination selected by the current user, determine a subset of the second point-of-interest data that is consistent with the category of the driving destination to be recommended to the current user.
[0112] By classifying the obtained POI data sets that the current users seldom go to, the sub-sets of each category that users of different categories seldom go to are obtained, and the POIs that similar users go to more often are classified according to the same category, then according to The POI category selected by the current user is recommended. For example, if the current user A wants to go shopping, select the POI that the user seldom goes to but similar users often go to from the obtained shopping category and recommend it to the current user, and the current user POI data and Classify the POI data of similar users, and provide POI data according to the user's wishes, which improves the accuracy of POI data.
[0113] Another embodiment of the present invention provides a driving destination recommendation device, such as image 3 shown, including:
[0114] A similar user set acquisition unit 21 is used to obtain a similar user set according to the driving history track data of the current user;
[0115] A historical point of interest data acquisition unit 22, configured to acquire historical point of interest data of similar users in the similar user set;
[0116] A driving destination point of interest data determining unit 23, configured to determine the driving destination point of interest data recommended to the current user according to the historical point of interest data of the current user and the historical point of interest data of similar users in the similar user set .
[0117] Preferably, the similar user set acquisition unit includes:
[0118] A driving familiar area acquisition unit, configured to obtain the current user's driving familiar area according to the current user's driving history trajectory data;
[0119] A driving history trajectory data acquisition unit, configured to acquire the driving history trajectory data of other users in the familiar driving area;
[0120] A similarity calculation unit, configured to calculate the trajectory similarity between the current user and other users in the current user's driving familiar area according to the current user's driving history trajectory data and the other user's driving history trajectory data and time similarity;
[0121] The first similar user set acquisition unit is configured to obtain a similar user set according to the trajectory similarity and time similarity calculation results.
[0122] Preferably, the acquisition unit of the known driving area includes:
[0123] Geographic coordinate information acquisition unit, used to obtain the geographic coordinate information in the current user's driving history track data;
[0124] The area acquisition unit is configured to determine a rectangular area familiar to the user when driving according to the geographical coordinate information.
[0125] Preferably, the first similar user set acquisition unit includes:
[0126] The second similar user set acquisition unit is used to obtain a track similar user set and a time similar user set respectively according to the calculation results of the track similarity and time similarity;
[0127] An arrangement unit is used to arrange the similarities of similar users in the trajectory similarity set and the time similarity set respectively;
[0128] A similar user subset acquisition unit is used to obtain a similar trajectory user subset and a time similar user subset according to the arrangement result;
[0129] The third similar user set obtaining unit is configured to obtain the similar user set according to the track similar user subset and the time similar user subset.
[0130] Preferably, the third similar user set acquisition unit includes:
[0131] a weight coefficient acquisition unit, configured to respectively acquire the weight coefficients of the subset of users with similar trajectories and the subset of users with similar time; the weight coefficients are obtained according to the evaluation data of the current user;
[0132] The fourth similar user set obtaining unit is configured to obtain the similar user set corresponding to the weight coefficient according to the weight coefficient.
[0133] Preferably, the driving destination POI data determination unit includes:
[0134] The first point-of-interest data collection acquisition unit is configured to obtain the first point-of-interest data whose number of times the current user goes to different points of interest is less than a first preset frequency according to the historical point-of-interest data of the current user. collection of
[0135] The second point-of-interest data collection acquisition unit is configured to obtain second point-of-interest data in which the number of times the similar user goes to different points of interest is greater than a second preset frequency according to the historical point-of-interest data of the similar user collection of
[0136] The first driving destination point of interest data determining unit is configured to determine the driving destination point of interest data recommended to the current user according to the set of the first point of interest data and the set of the second point of interest data.
[0137] Preferably, the first driving destination POI data determination unit includes:
[0138] The sub-set acquisition unit of the first point of interest data is configured to classify the data of the set of the first point of interest data to obtain a subset of the first point of interest data of at least one category;
[0139] The sub-set acquiring unit of the second point-of-interest data is configured to, according to the category of the sub-set of the first point-of-interest data, find the second point-of-interest data consistent with the category in the set of the second point-of-interest data. sub-collection of
[0140] The second driving destination POI data determining unit is configured to determine, according to the driving destination category selected by the current user, a subset of second POI data that is consistent with the driving destination category to be recommended to the current user.
[0141] The driving destination recommendation device provided by another embodiment of the present invention obtains a similar user set according to the current user's driving history trajectory data, and then obtains the historical interest point data of similar users in the similar user set according to the similar user set, and according to the current user The historical point of interest data and the historical point of interest data of similar users in the similar user set determine the point of interest data of the driving destination recommended to the current user, which solves the problem that the existing driving destination recommendation method cannot be based on the advantages and disadvantages of the destination and the user's reach. The familiarity of the destination path for recommendation problems.
[0142] Apparently, the above-mentioned embodiments are only examples for clear description, rather than limiting the implementation. For those of ordinary skill in the art, other changes or changes in different forms can be made on the basis of the above description. It is not necessary and impossible to exhaustively list all the implementation manners here. And the obvious changes or changes derived therefrom are still within the scope of protection of the present invention.

PUM

no PUM

Description & Claims & Application Information

We can also present the details of the Description, Claims and Application information to help users get a comprehensive understanding of the technical details of the patent, such as background art, summary of invention, brief description of drawings, description of embodiments, and other original content. On the other hand, users can also determine the specific scope of protection of the technology through the list of claims; as well as understand the changes in the life cycle of the technology with the presentation of the patent timeline. Login to view more.
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products