CN106446157B - Travel destination recommendation method and device - Google Patents
Travel destination recommendation method and device Download PDFInfo
- Publication number
- CN106446157B CN106446157B CN201610844639.7A CN201610844639A CN106446157B CN 106446157 B CN106446157 B CN 106446157B CN 201610844639 A CN201610844639 A CN 201610844639A CN 106446157 B CN106446157 B CN 106446157B
- Authority
- CN
- China
- Prior art keywords
- travel destination
- weight
- user
- candidate
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Abstract
The invention discloses a method and a device for recommending a travel destination, wherein the method comprises the following steps: determining a regular travel destination of the user according to the user track data; determining a recent interest travel destination of the user according to the user retrieval data; determining candidate travel destinations according to the regular travel destinations and the recent interest travel destinations; and respectively determining the confidence score of each candidate travel destination, sequencing the candidate travel destinations according to the sequence of the scores from high to low, recommending the sequenced M front-located candidate travel destinations to the user, wherein M is a positive integer. By applying the scheme of the invention, the accuracy of the recommendation result can be improved.
Description
[ technical field ] A method for producing a semiconductor device
The invention relates to the internet technology, in particular to a travel destination recommendation method and device.
[ background of the invention ]
In order to intelligently and secretarily help a user to complete the whole travel process and improve the viscosity and satisfaction degree of the user using a map product, the possible travel destination of the user needs to be intelligently found and recommended.
In the prior art, generally, only the home and the company of the mined user can be recommended to the user as a travel destination, and the recommendation cannot be performed according to the actual requirements of the user, so that the accuracy of the recommendation result is reduced.
[ summary of the invention ]
The invention provides a travel destination recommendation method and device, which can improve the accuracy of a recommendation result.
The specific technical scheme is as follows:
a travel destination recommendation method comprising:
determining regular travel destinations of users according to user track data, and determining recent interest travel destinations of the users according to user retrieval data;
determining candidate travel destinations according to the regular travel destinations and the recent interest travel destinations;
and respectively determining the confidence score of each candidate travel destination, sequencing the candidate travel destinations according to the sequence of the scores from high to low, recommending the sequenced M front-located candidate travel destinations to the user, wherein M is a positive integer.
A travel destination recommendation apparatus comprising: the recommendation system comprises a first processing unit, a second processing unit, a third processing unit and a recommendation unit;
the first processing unit is used for determining a regular travel destination of the user according to the user trajectory data and sending the regular travel destination to the third processing unit;
the second processing unit is used for determining a recent interest travel destination of the user according to user retrieval data and sending the recent interest travel destination to the third processing unit;
the third processing unit is used for determining candidate travel destinations according to the regular travel destinations and the recent interest travel destinations;
and the recommending unit is used for respectively determining the confidence score of each candidate travel destination, sequencing the candidate travel destinations according to the sequence of the scores from high to low, and recommending the sequenced M front-located candidate travel destinations to the user, wherein M is a positive integer.
According to the scheme, regular travel destinations of the user can be determined according to user track data, recent interest travel destinations of the user are determined according to user retrieval data, and then candidate travel destinations are determined according to the regular travel destinations and the recent interest travel destinations.
[ description of the drawings ]
Fig. 1 is a flowchart of an embodiment of a method for recommending a travel destination according to the present invention.
Fig. 2 is a flowchart of an embodiment of the method for determining a regular travel destination of a user according to user trajectory data according to the present invention.
FIG. 3 is a flowchart of an embodiment of a method for determining a recently-interested travel destination of a user according to user search data.
Fig. 4 is a schematic diagram of a process for determining candidate travel destinations and strong time attribute weights for the candidate travel destinations according to the present invention.
Fig. 5 is a schematic diagram illustrating a process of recommending a candidate travel destination to a user according to the present invention.
Fig. 6 is a schematic structural diagram of a trip destination recommendation device according to an embodiment of the present invention.
[ detailed description ] embodiments
In order to make the technical solution of the present invention clearer and more obvious, the solution of the present invention is further described in detail below by referring to the drawings and examples.
Example one
Fig. 1 is a flowchart of an embodiment of a method for recommending a travel destination according to the present invention, as shown in fig. 1, including the following specific implementation manners:
in 11, determining a regular travel destination of the user according to the user trajectory data;
at 12, determining a recent interest travel destination of the user according to the user retrieval data;
at 13, determining candidate travel destinations according to the regular travel destinations and the recent interest travel destinations;
in 14, the confidence score of each candidate trip destination is determined, the candidate trip destinations are ranked according to the order of the scores from high to low, the ranked candidate trip destinations at the top M positions are recommended to the user, and M is a positive integer.
Specific implementations of the above steps are described in detail below.
1) Determining regular travel destinations of users according to user trajectory data
In order to obtain the regular travel destination, user track data can be obtained firstly, then the user track data is analyzed, stop points in the user track are determined, and then frequent places with time regularity are screened out from the stop points and used as the regular travel destination of the user.
Fig. 2 is a flowchart of an embodiment of a method for determining a regular travel destination of a user according to user trajectory data according to the present invention, as shown in fig. 2, including the following specific implementation manners.
In 21, user trajectory data is acquired.
The acquired user trajectory data may include: usage data of a map, such as a hundredth map, by a user, and trajectory data of the user collected by other applications by means of, for example, a hundredth Software Development Kit (SDK), etc.
How to obtain user track data is the prior art, and all user track data of the user can be obtained.
At 22, a dwell point in the user trajectory is determined by analyzing the user trajectory data.
How to analyze the user trajectory data to determine the staying point in the user trajectory is also the prior art, for example, the staying point in the user trajectory can be mined according to the information of the user positioning networking type, whether the base station is switched, and the like, and the information includes the position coordinate of the staying point, the occurrence time of the user at the staying point, the staying time, and the like.
In 23, the noise type dwell points are filtered out.
Whether to perform an operation of filtering out the noise type of the stop point is optional.
Filtering rules may be preset to filter the dwell points for noise types.
What is specifically included in the filtering rules may be determined according to actual needs, for example, the noise type stop point may be filtered according to the time length of the user's stay at the stop point or the time of the user's occurrence at the stop point.
For example, if a user only stays at a certain stop for three minutes, it may be that the user goes to a convenience store on the roadside to buy a bottle of beverage on the way, and for such a stop, it can be considered as a noise-type stop.
In 24, geographically close stopover points are clustered.
After the processing in step 23, the remaining stopover points are all valuable stopover points, and then the stopover points with similar geographic positions can be clustered by adopting the existing clustering algorithm according to the position coordinates of each stopover point.
Optionally, for any clustering result, if the number of remaining points included in the clustering result is too small, the clustering result may be filtered out, that is, the clustering result is not subjected to subsequent processing.
In 25, 26 to 28 are performed for each clustering result, respectively.
At 26, the average value of the position coordinates of each stop point in the clustering result is used as the position coordinate of the trip destination corresponding to the clustering result.
If the position coordinates of each stop point in the clustering result are known, the mean value of the position coordinates can be calculated, and the calculation result is used as the position coordinates of the travel destination corresponding to the clustering result.
At 27, the day-level regularity weight and week-level regularity weight of the travel destination corresponding to the clustering result are determined according to the time distribution of each stop point appearing in the clustering result by the user.
How to determine the day-level regularity weight and the week-level regularity weight may be determined according to actual needs, and is not limited to the following manner.
For the day-level regularity weight, the number of days in the last predetermined time period, for example, the last 14 days (two weeks), in which the user has reached the stop point in the clustering result, may be counted first according to the time of the user appearing at each stop point in the clustering result, and then the counted number of days is divided by 14, so as to obtain the day-level regularity weight of the travel destination corresponding to the clustering result.
Assuming that the user has reached the stop point in the clustering result for 7 days in the last 14 days, the obtained day-level regularity weight is 0.5, and it can be seen that, according to the above manner, the value of the obtained day-level regularity weight is at least 0 and at most 1.
For the week-level regularity weights, respectively counting the number of days that the last user with a preset time length, such as the last 4 weeks, reaches the stop point in the clustering result according to the time of the user at each stop point in the clustering result, dividing the number of days by 4 according to the counting result to obtain the week-level regularity weights of the Monday, similarly, respectively obtaining the week-level regularity weights of the Tuesday to the Sunday, then selecting a maximum value from the week-level regularity weights of the Monday to the Sunday, and taking the maximum value as the week-level regularity weight of the travel destination corresponding to the clustering result.
Assuming that 2 monday users reach the stop point in the clustering result in 4 mondays in the last 4 weeks, the obtained week-level regularity weight of mondays is 0.5, and similarly, the value of the obtained week-level regularity weight is minimum 0 and maximum 1 according to the above manner.
After 7 week-level regularity weights of monday to sunday are obtained, the maximum value of the week-level regularity weights can be used as the week-level regularity weight of the travel destination corresponding to the clustering result.
At 28, the day-level regularity weights are compared with the corresponding first threshold values, the week-level regularity weights are compared with the corresponding second threshold values, if any one of the weights is greater than the corresponding threshold value, the travel destination corresponding to the clustering result is determined as a regular travel destination, the weight greater than the corresponding threshold value is used as the regularity weight of the regular travel destination, and the weight less than or equal to the corresponding threshold value is discarded.
After the day-level regularity weight and the week-level regularity weight of the route destination corresponding to the clustering result are obtained, the two weights can be respectively compared with corresponding threshold values, namely the day-level regularity weight is compared with a first threshold value, the week-level regularity weight is compared with a second threshold value, and if the day-level regularity weight is larger than the first threshold value, or the week-level regularity weight is larger than the second threshold value, or the day-level regularity weight is larger than the first threshold value and the week-level regularity weight is larger than the second threshold value, the route destination corresponding to the clustering result is determined as the regular route destination.
For the regular trip destination, the regularity weight may include only a day-level regularity weight, only a week-level regularity weight, and both the day-level regularity weight and the week-level regularity weight.
If the comparison result is that the day-level regularity weight is greater than the first threshold value but the week-level regularity weight is less than or equal to the second threshold value, the regularity weight of the regular trip destination only comprises the day-level regularity weight; if the comparison result is that the day-level regularity weight is less than or equal to the first threshold value but the week-level regularity weight is greater than the second threshold value, the regularity weight of the regular trip destination only comprises the week-level regularity weight; if the comparison result is that the day-level regularity weight is greater than the first threshold and the week-level regularity weight is greater than the second threshold, the regularity weight of the regular trip destination simultaneously comprises the day-level regularity weight and the week-level regularity weight.
The specific values of the first threshold and the second threshold can be determined according to actual needs.
2) Determining the recent interest travel destination of the user according to the user retrieval data
In order to obtain the recently-interested travel destination, user retrieval data may be first obtained, where the user retrieval data described in this embodiment refers to retrieval data of a user in a map, and then the recently-interested travel destination of the user may be determined according to the user retrieval data.
Fig. 3 is a flowchart of an embodiment of a method for determining a recently-interested travel destination of a user according to user search data according to the present invention, as shown in fig. 3, including the following specific implementation manners.
In 31, user search data is acquired.
For example, the user may perform route planning, location viewing, and the like by using a map, so as to search a point of interest (poi), such as beijing south station, which is search data.
How to obtain the user retrieval data is the prior art, and all the user retrieval data which the user passes can be obtained.
At 32, noise-type search data is filtered out.
Whether to perform an operation of filtering out the retrieved data of the noise type is optional.
The filtering rules can be preset so as to filter the retrieved data of the noise type, and the specific content included in the filtering rules can be determined according to the actual needs.
At 33, the search data corresponding to the same poi are clustered.
After the processing in 32, the rest of the retrieval data are valuable retrieval data, and on the basis, the retrieval data corresponding to the same poi can be clustered.
For example, for a poi of beijing south station, a user performs multiple searches, and the search data corresponding to the multiple searches are clustered together.
Optionally, for any clustering result, if the number of search data included in the clustering result is too small, the clustering result may be filtered out, that is, the clustering result is not subjected to subsequent processing.
In 34, 35 to 36 are performed for each clustering result, respectively.
In 35, according to the retrieval type and the retrieval time of each retrieval data in the clustering result, the recent interest weight of the poi corresponding to the clustering result is determined.
If it is determined that the recent interest weight of the poi corresponding to the clustering result can be determined according to actual needs, the method is not limited to the following manner.
And aiming at each retrieval data in the clustering result, determining the retrieval type score and the retrieval time score of the retrieval data according to the retrieval type and the distance between the retrieval time and the current time. The retrieval types can comprise route planning, site viewing and the like, and the retrieval type scores corresponding to different retrieval types can be preset respectively. The closer the retrieval time is to the current time, the higher the corresponding retrieval time score is, whereas the farther the retrieval time is from the current time, the lower the corresponding retrieval time score is, and the corresponding relationship between the distance between the retrieval time and the current time and the corresponding retrieval time score can be preset, for example, a plurality of continuous value-taking intervals can be preset, each value-taking interval corresponds to different retrieval time scores, and the corresponding retrieval time score is determined according to the value-taking interval to which the distance between the retrieval time and the current time belongs.
After the retrieval type score and the retrieval time score of each retrieval data are obtained, the retrieval type score and the retrieval time score of the retrieval data can be multiplied by corresponding coefficients respectively and then added, so that the score of the retrieval data is obtained, and the specific value of each coefficient can be determined according to actual needs.
In this way, for the clustering result, the scores of the retrieval data included in the clustering result can be added as the recent interest weight of the poi corresponding to the clustering result.
At 36, the recent interest weight is compared with a corresponding third threshold, and if the recent interest weight is greater than the third threshold, the poi corresponding to the clustering result is determined as the recent interest travel destination.
After the recent interest weight of the poi corresponding to the clustering result is obtained, the recent interest weight can be compared with a third threshold, and if the recent interest weight is greater than the third threshold, the poi corresponding to the clustering result is determined as a recent interest travel destination.
The specific value of the third threshold can be determined according to actual needs.
3) Determining candidate travel destinations according to regular travel destinations and recent interest travel destinations
After the regular travel destinations and the recently-interested travel destinations are respectively determined according to the manners in 1) and 2), candidate travel destinations can be further determined, and in order to realize subsequent recommendation, the strong time attribute weight of each candidate travel destination needs to be respectively determined.
Fig. 4 is a schematic diagram of a process for determining candidate travel destinations and strong time attribute weights for the candidate travel destinations according to the present invention.
As shown in fig. 4, for each regular trip destination, the following processes may be performed:
according to the position coordinates of the regular travel destination, posi in an area with a preset size around the regular travel destination on a map is determined;
and according to the distance from the regular travel destination and the poi heat degree, selecting one representative poi from the determined pois, taking the representative poi as a candidate travel destination, and taking the regularity weight of the regular travel destination as the regularity weight of the candidate travel destination.
For example, poi in a circular area on a map with the position coordinates of the regular travel destination as the center and the radius of 100 meters may be determined, and then a representative poi may be selected from the determined poi according to the distance from the regular travel destination and the heat of the poi, and the representative poi is used as a candidate travel destination, that is, one regular travel destination corresponds to one representative poi, and one representative poi is one candidate travel destination.
How to select the representative poi may be determined according to actual needs, for example, there are two pois for selection, one is a beijing south station, the other is a roadside convenience store, and the heat degree of the beijing south station is significantly higher than that of the roadside convenience store, so even if the roadside convenience store is closer to the regular travel destination than the beijing south station, the beijing south station is usually selected as the representative poi, and therefore, different weights may be set for the distance and the heat degree, and the representative poi may be selected by integrating the actual distance, the heat degree and the weight.
As shown in fig. 4, all the recent interest travel destinations determined in 2) may be used as candidate travel destinations, and the recent interest weight of the recent interest travel destination may be used as the recent interest weight of the candidate travel destination.
Then, for each candidate travel destination, description information of the candidate travel destination, which may include types, labels, and the like, may be obtained, and then a preset strong time attribute weight table is queried, and a strong time attribute weight corresponding to the description information of the candidate travel destination is found, and the found strong time attribute weight is used as the strong time attribute weight of the candidate travel destination.
The strong time attribute weight table can be generated according to actual experience, the requirements of different travel destinations on time and the like, the specific form of the table can be determined according to actual needs, but it is required to ensure that only one strong time attribute weight can be found for one candidate travel destination according to the description information of the candidate travel destination, but different candidate travel destinations can correspond to the same strong time attribute weight.
4) Recommending candidate travel destinations to a user
Fig. 5 is a schematic diagram of a process of recommending a candidate travel destination to a user according to the present invention, and as shown in fig. 5, when a user opens a map and enters a recommendation interface, recommendation of the candidate travel destination can be performed for the user.
Specifically, for each candidate travel destination, weight information of the candidate travel destination may be first obtained, where the obtained weight includes a strong time attribute weight and one or all of the following: and then determining the confidence score of the candidate travel destination according to all the weights of the candidate travel destination.
For each candidate travel destination, the weight that must be included is a strong temporal attribute weight, while other weights may or may not be included, such as for a travel destination of near-term interest, a weight of near-term interest but typically not a weight of regularity, and for a travel destination of regularity, a weight of regularity but typically not a weight of near-term interest.
A confidence calculation formula may be empirically generated in advance, so that for each candidate travel destination, the confidence score of the candidate travel destination may be calculated according to the confidence calculation formula and all the weights of the candidate travel destination, and the specific form of the confidence calculation formula may be determined according to actual needs, for example, each weight may be multiplied by a corresponding coefficient and then added, and if a certain weight is lacked, the weight may be considered as 0.
After the confidence scores of the candidate travel destinations are obtained respectively, the candidate travel destinations can be ranked according to the order of the scores from high to low, and the ranked candidate travel destinations at the top M positions are recommended to the user, wherein M is a positive integer, and the specific value can also be determined according to the actual requirement, such as 3.
It should be noted that, since the user trajectory data and the user search data of the user are continuously updated, the operations 1) to 3) and the like may be periodically executed, and for example, at the zero point of each day, the operations 1) to 3) and the like are executed again according to the latest user trajectory data and user search data, so that the recommendation result is more and more accurate.
The above is a description of method embodiments, and the embodiments of the present invention are further described below by way of apparatus embodiments.
Example two
Fig. 6 is a schematic diagram of a composition structure of an embodiment of the travel destination recommendation device according to the present invention, as shown in fig. 6, including: a first processing unit 61, a second processing unit 62, a third processing unit 63 and a recommending unit 64.
The first processing unit 61 is configured to determine a regular travel destination of the user according to the user trajectory data, and send the regular travel destination to the third processing unit 63;
the second processing unit 62 is configured to determine a recent interest trip destination of the user according to the user retrieval data, and send the recent interest trip destination to the third processing unit 63;
a third processing unit 63, configured to determine a candidate route destination according to the regular route destination and the route destination of recent interest;
and the recommending unit 64 is configured to determine the confidence score of each candidate route destination, sort the candidate route destinations in the order from high to low according to the scores, and recommend the sorted top M-numbered candidate route destinations to the user, where M is a positive integer.
The first processing unit 61 may first obtain the user trajectory data, then determine the stop points in the user trajectory by analyzing the user trajectory data, and further screen out frequent places with time regularity from the stop points as regular travel destinations of the user.
How to obtain the user trajectory data is the prior art, the first processing unit 61 may obtain all user trajectory data of the user, and dig a stopping point in the user trajectory, including a position coordinate of the stopping point, an appearance time of the user at the stopping point, a stopping time, and the like.
Optionally, after determining the stopping points in the user trajectory, the first processing unit 61 may also filter out the stopping points which are noise types.
Then, the first processing unit 61 may cluster the stop points with similar geographic locations, and optionally, for any clustering result, if the number of stop points included in the clustering result is too small, the clustering result may be filtered, that is, the clustering result is not subjected to subsequent processing.
After that, the first processing unit 61 may perform the following processing for each clustering result respectively:
taking the mean value of the position coordinates of each stop point in the clustering result as the position coordinates of the travel destination corresponding to the clustering result;
determining the day-level regularity weight and week-level regularity weight of the travel destination corresponding to the clustering result according to the time distribution of each stop point appearing in the clustering result by the user;
comparing the day-level regularity weight with a corresponding first threshold value, comparing the week-level regularity weight with a corresponding second threshold value, if any weight is larger than the corresponding threshold value, determining the travel destination corresponding to the clustering result as a regular travel destination, taking the weight larger than the corresponding threshold value as the regularity weight of the regular travel destination, and discarding the weight smaller than or equal to the corresponding threshold value.
In order to obtain the route destination of recent interest, the second processing unit 62 may first obtain user search data, where the user search data in this embodiment refers to search data of a user in a map, and then may determine the route destination of recent interest of the user according to the user search data.
How to obtain the user retrieval data is the prior art, the second processing unit 62 may obtain all the user retrieval data of the user.
Optionally, after the user retrieval data is acquired, the retrieval data which is noise type can be filtered.
Then, the second processing unit 62 may perform clustering on the search data corresponding to the same poi, and optionally, for any clustering result, if the number of search data included in the clustering result is too small, the clustering result may be filtered, that is, the clustering result is not subjected to subsequent processing.
Thereafter, for each clustering result, the second processing unit 62 may perform the following processing:
determining the recent interest weight of the poi corresponding to the clustering result according to the retrieval type and the retrieval time of each retrieval data in the clustering result;
and comparing the recent interest weight with a corresponding third threshold, and if the recent interest weight is greater than the third threshold, determining the poi corresponding to the clustering result as a recent interest travel destination.
After determining the regular travel destination and the recently interested travel destination, respectively, in the above manner, the third processing unit 63 may further determine candidate travel destinations.
Specifically, the third processing unit 63 may perform the following processing for each regular trip destination, respectively:
according to the position coordinates of the regular travel destination, posi in an area with a preset size around the regular travel destination on a map is determined;
and according to the distance from the regular travel destination and the poi heat degree, selecting one representative poi from the determined pois, taking the representative poi as a candidate travel destination, and taking the regularity weight of the regular travel destination as the regularity weight of the candidate travel destination.
The third processing unit 63 may take all of the recent interest trip destinations as candidate trip destinations, and take the recent interest weight of the recent interest trip destination as the recent interest weight of the candidate trip destination.
In addition, the third processing unit 63 needs to determine a strong time attribute weight for each candidate trip destination, respectively.
For each candidate travel destination, the third processing unit 63 may respectively obtain the description information of the candidate travel destination, and look up the strong time attribute weight corresponding to the description information of the candidate travel destination by querying a preset strong time attribute weight table, and use the found strong time attribute weight as the strong time attribute weight of the candidate travel destination.
When a travel destination needs to be recommended to the user, the recommending unit 64 may determine the confidence score of each candidate travel destination, that is, the confidence score of each candidate travel destination may be determined according to all the weights of the candidate travel destination.
The weight for each candidate travel destination may include a strong temporal attribute weight and one or all of the following: regularity weight, recent interest weight.
The first processing unit 61 and the second processing unit 62 send the regular route destination and the recent interest route destination to the third processing unit 63, and simultaneously send the regular weight and the recent interest weight to the third processing unit 63, and accordingly, the recommending unit 64 may obtain each candidate route destination and the corresponding weight from the third processing unit 63, so as to respectively calculate the confidence score of each candidate route destination according to the weight and the like.
After the confidence scores of the candidate trip destinations are obtained, the recommending unit 64 may rank the candidate trip destinations according to the order of the scores from high to low, and recommend the ranked candidate trip destinations M-th before to the user, where M is a positive integer, and a specific value may be determined according to actual needs, and may be 3, for example.
For the specific work flow of the above device embodiment, please refer to the corresponding description in the above method embodiment, which is not described herein again.
In short, by adopting the scheme of the invention, regular travel destinations of a user can be determined according to user track data, recent interest travel destinations of the user can be determined according to user retrieval data, and then candidate travel destinations can be determined according to the regular travel destinations and the recent interest travel destinations, so that when travel destinations need to be recommended to the user, confidence scores of each candidate travel destination can be determined respectively, all candidate travel destinations are ranked according to the sequence of scores from high to low, and then the ranked candidate travel destinations at the top M positions are recommended to the user; moreover, by adopting the scheme of the invention, the confidence score of each candidate travel destination can be determined through the regularity weight, the recent interest weight, the strong time attribute weight and the like, and then the candidate travel destination is selected and recommended according to the score result, so that the recommendation result has strong time attribute, and the subsequent travel planning and the like of the user are more convenient.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described device embodiments are merely illustrative, and for example, the division of the units is only one logical functional division, and other divisions may be realized in practice.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (12)
1. A travel destination recommendation method, comprising:
determining regular travel destinations of users according to user track data, and determining recent interest travel destinations of the users according to user retrieval data; the user retrieval data is retrieval data of the user in a map;
determining candidate travel destinations according to the regular travel destinations and the recent interest travel destinations;
respectively determining the confidence score of each candidate travel destination, sequencing the candidate travel destinations according to the sequence of the scores from high to low, recommending the sequenced M front-located candidate travel destinations to the user, wherein M is a positive integer;
wherein, the determining the regular travel destination of the user according to the user trajectory data comprises:
acquiring the user track data; determining a stopping point in the user track by analyzing the user track data;
screening out frequent places with time regularity from the stopping points as the regular travel destination, wherein the frequent places with time regularity comprise: clustering stop points with similar geographic positions; and aiming at each clustering result, respectively carrying out the following processing: determining the day-level regularity weight and week-level regularity weight of the travel destination corresponding to the clustering result according to the time distribution of each stop point of the user appearing in the clustering result; and comparing the day-level regularity weight with a corresponding first threshold, comparing the week-level regularity weight with a corresponding second threshold, and if any weight is greater than the corresponding threshold, determining the travel destination corresponding to the clustering result as the regular travel destination.
2. The method of claim 1,
the method further comprises the following steps:
taking the mean value of the position coordinates of each stop point in the clustering result as the position coordinates of the travel destination corresponding to the clustering result;
and taking the weight larger than the corresponding threshold value as the regularity weight of the regular travel destination, and discarding the weight smaller than or equal to the corresponding threshold value.
3. The method of claim 2,
the determining the recent interest travel destination of the user according to the user retrieval data comprises:
clustering the retrieval data corresponding to the same interest point poi;
and aiming at each clustering result, respectively carrying out the following processing:
determining the recent interest weight of the poi corresponding to the clustering result according to the retrieval type and the retrieval time of each retrieval data in the clustering result;
and comparing the recent interest weight with a corresponding third threshold, and if the recent interest weight is greater than the third threshold, determining the poi corresponding to the clustering result as the recent interest travel destination.
4. The method of claim 3,
the determining a candidate travel destination according to the regular travel destination and the recent interest travel destination comprises:
for each regular trip destination, the following processing is performed:
according to the position coordinates of the regular travel destination, posis in a region with a preset size around the regular travel destination on a map are determined;
selecting a representative poi from the determined pois according to the distance from the regular travel destination and the poi popularity, taking the representative poi as the candidate travel destination, and taking the regularity weight of the regular travel destination as the regularity weight of the candidate travel destination;
taking the recent interest travel destination as the candidate travel destination, and taking the recent interest weight of the recent interest travel destination as the recent interest weight of the candidate travel destination.
5. The method of claim 4,
before determining the confidence score of each candidate travel destination, the method further includes: respectively determining the strong time attribute weight of each candidate trip destination;
the determining the confidence score for each candidate travel destination comprises: and for each candidate travel destination, determining the confidence score of the candidate travel destination according to all the weights of the candidate travel destination.
6. The method of claim 5,
the determining the strong time attribute weight for each candidate travel destination comprises:
and respectively acquiring the description information of the candidate travel destinations aiming at each candidate travel destination, searching the strong time attribute weight corresponding to the description information of the candidate travel destination by inquiring a preset strong time attribute weight table, and taking the searched strong time attribute weight as the strong time attribute weight of the candidate travel destination.
7. A travel destination recommendation device, comprising: the recommendation system comprises a first processing unit, a second processing unit, a third processing unit and a recommendation unit;
the first processing unit is used for determining a regular travel destination of the user according to the user trajectory data and sending the regular travel destination to the third processing unit;
the second processing unit is used for determining a recent interest travel destination of the user according to user retrieval data and sending the recent interest travel destination to the third processing unit; the user retrieval data is retrieval data of the user in a map;
the third processing unit is used for determining candidate travel destinations according to the regular travel destinations and the recent interest travel destinations;
the recommending unit is used for respectively determining the confidence score of each candidate travel destination, sequencing the candidate travel destinations according to the sequence of the scores from high to low, and recommending the sequenced M front-located candidate travel destinations to the user, wherein M is a positive integer;
the first processing unit acquires the user track data; determining a stopping point in the user track by analyzing the user track data; screening out frequent places with time regularity from the stopping points as the regular travel destination, wherein the frequent places with time regularity comprise: clustering stop points with similar geographic positions; and aiming at each clustering result, respectively carrying out the following processing: determining the day-level regularity weight and week-level regularity weight of the travel destination corresponding to the clustering result according to the time distribution of each stop point of the user appearing in the clustering result; and comparing the day-level regularity weight with a corresponding first threshold, comparing the week-level regularity weight with a corresponding second threshold, and if any weight is greater than the corresponding threshold, determining the travel destination corresponding to the clustering result as the regular travel destination.
8. The apparatus of claim 7,
the first processing unit takes the mean value of the position coordinates of each stop point in the clustering result as the position coordinates of the travel destination corresponding to the clustering result;
the first processing unit regards a weight larger than a corresponding threshold as a regularity weight of the regularity travel destination, and discards a weight smaller than or equal to the corresponding threshold.
9. The apparatus of claim 8,
the second processing unit clusters the retrieval data corresponding to the same interest point poi;
and aiming at each clustering result, respectively carrying out the following processing:
determining the recent interest weight of the poi corresponding to the clustering result according to the retrieval type and the retrieval time of each retrieval data in the clustering result;
and comparing the recent interest weight with a corresponding third threshold, and if the recent interest weight is greater than the third threshold, determining the poi corresponding to the clustering result as the recent interest travel destination.
10. The apparatus of claim 9,
the third processing unit performs the following processing for each regular trip destination, respectively:
according to the position coordinates of the regular travel destination, posis in a region with a preset size around the regular travel destination on a map are determined;
selecting a representative poi from the determined pois according to the distance from the regular travel destination and the poi popularity, taking the representative poi as the candidate travel destination, and taking the regularity weight of the regular travel destination as the regularity weight of the candidate travel destination;
taking the recent interest travel destination as the candidate travel destination, and taking the recent interest weight of the recent interest travel destination as the recent interest weight of the candidate travel destination.
11. The apparatus of claim 10,
the third processing unit is further configured to determine a strong time attribute weight for each candidate trip destination, respectively;
and the recommending unit determines the confidence degree scores of the candidate travel destinations according to all the weights of the candidate travel destinations respectively aiming at each candidate travel destination.
12. The apparatus of claim 11,
the third processing unit respectively acquires the description information of the candidate travel destinations aiming at each candidate travel destination, searches the strong time attribute weight corresponding to the description information of the candidate travel destination by inquiring a preset strong time attribute weight table, and takes the found strong time attribute weight as the strong time attribute weight of the candidate travel destination.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610844639.7A CN106446157B (en) | 2016-09-22 | 2016-09-22 | Travel destination recommendation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610844639.7A CN106446157B (en) | 2016-09-22 | 2016-09-22 | Travel destination recommendation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106446157A CN106446157A (en) | 2017-02-22 |
| CN106446157B true CN106446157B (en) | 2020-01-21 |
Family
ID=58167129
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610844639.7A Active CN106446157B (en) | 2016-09-22 | 2016-09-22 | Travel destination recommendation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106446157B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110020111B (en) * | 2017-09-21 | 2023-02-24 | 腾讯科技（深圳）有限公司 | Travel recommendation method and device, computer equipment and storage medium |
| CN112262557B (en) | 2018-06-11 | 2023-12-05 | 北京嘀嘀无限科技发展有限公司 | System and method for providing information for on-line to off-line services |
| CN110059260B (en) * | 2019-04-29 | 2020-07-31 | 北京字节跳动网络技术有限公司 | Recommendation method, device, equipment and medium |
| CN110516017B (en) * | 2019-08-02 | 2022-05-20 | Oppo广东移动通信有限公司 | Location information processing method and device based on terminal equipment, electronic equipment and storage medium |
| CN114041129A (en) * | 2019-11-04 | 2022-02-11 | 北京嘀嘀无限科技发展有限公司 | System and method for determining name of boarding point |
| CN111611500A (en) * | 2020-04-09 | 2020-09-01 | 中国平安财产保险股份有限公司 | Frequent place identification method and device based on clustering and storage medium |
| CN113590674A (en) * | 2021-06-29 | 2021-11-02 | 北京百度网讯科技有限公司 | Travel purpose identification method, device, equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102719A (en) * | 2014-07-18 | 2014-10-15 | 百度在线网络技术（北京）有限公司 | Track information pushing method and device |
| CN105183800A (en) * | 2015-08-25 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Information prediction method and apparatus |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104520881B (en) * | 2012-06-22 | 2017-12-19 | 谷歌公司 | Predict that the future to place accesses based on possibility is accessed to neighbouring destination ranking and from position history |
| JP5944770B2 (en) * | 2012-07-17 | 2016-07-05 | 株式会社デンソーアイティーラボラトリ | Destination proposal system, destination proposal method, and program |
| KR101405434B1 (en) * | 2012-11-23 | 2014-06-11 | 현대엠엔소프트 주식회사 | Device and method for generating guiding information about crossroad |
| CN104410717A (en) * | 2014-12-18 | 2015-03-11 | 百度在线网络技术（北京）有限公司 | Information pushing method and device |
| CN109241439A (en) * | 2015-05-12 | 2019-01-18 | 北京嘀嘀无限科技发展有限公司 | A kind of method and device of information push |
-
2016
- 2016-09-22 CN CN201610844639.7A patent/CN106446157B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102719A (en) * | 2014-07-18 | 2014-10-15 | 百度在线网络技术（北京）有限公司 | Track information pushing method and device |
| CN105183800A (en) * | 2015-08-25 | 2015-12-23 | 百度在线网络技术（北京）有限公司 | Information prediction method and apparatus |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106446157A (en) | 2017-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106446157B (en) | Travel destination recommendation method and device | |
| CN107291888B (en) | Machine learning statistical model-based living recommendation system method near living hotel | |
| CN110726418B (en) | Method, device and equipment for determining interest point region and storage medium | |
| CN106570722B (en) | Intelligent recommendation system and intelligent recommendation method | |
| US10136249B2 (en) | Information distribution apparatus and method | |
| CN107167136B (en) | Position recommendation method and system for electronic map | |
| TWI564738B (en) | Method and apparatus for recommending candidate words based on geographic location | |
| TWI584137B (en) | Search, determine the active area of the method with the server | |
| Ceikute et al. | Routing service quality--local driver behavior versus routing services | |
| CN103389979B (en) | Recommend system, the device and method of classified lexicon in input method | |
| US10228258B2 (en) | Determining top venues from aggregated user activity location data | |
| KR20160010279A (en) | Method and apparatus for pushing track information | |
| Xu et al. | DESTPRE: a data-driven approach to destination prediction for taxi rides | |
| KR20190124524A (en) | Method and system for recommending point of interest | |
| CN106126544B (en) | Internet content delivery method and device | |
| JP6756744B2 (en) | Location information provision method and equipment | |
| CN113486927B (en) | Priori probability-based unsupervised track access place labeling method | |
| KR20140093772A (en) | Method for recommending point of interest using user preferences and moving patterns | |
| WO2018192506A1 (en) | Method and apparatus for recommending social information, and storage medium | |
| CA3033210A1 (en) | Method and apparatus for interacting with information distribution system | |
| WO2017008653A1 (en) | Poi service provision method, poi data processing method and device | |
| CN106779204B (en) | Personalized path recommendation method and device | |
| CN109699003B (en) | Position determination method and device | |
| JP2013257668A (en) | Interest analysis method, interest analyzer and program of the same | |
| CN116823535B (en) | Journey planning and intelligent navigation system based on multi-mode large model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |