KR20210157073A - System and method for recommending travel destination combination based on hybrid filtering - Google Patents

Abstract

Disclosed are a hybrid filtering-based travel destination combination recommendation system and method, and a recording medium for recommending an optimal travel destination combination to a user by hybrid filtering in which content-based filtering and collaborative filtering are mixed. A hybrid filtering-based travel destination combination recommendation system according to an embodiment comprises: a travel destination attribute data generation unit that collects travel destination contents related to a plurality of travel destinations and generates attribute data for each travel destination from collected travel destination contents; a data input unit that receives, from a first user, a first travel destination history including a travel destination or a combination of travel destinations, and travel destination recommendation request data including a travel period and travel attributes; a content filtering recommendation module for recommending a first travel destination on the basis of the similarity between the travel destination recommendation request data and the attribute data for each travel destination; a collaborative filtering recommendation module that compares the first travel destination history with second travel destination histories of a plurality of second users by collaborative filtering and recommends a second travel destination from the travel destination histories of a second user having a high degree of relevance to the first user; a travel destination history frequency calculation module for calculating a travel destination history frequency that is a frequency included in the second travel destination histories of the second users with respect to the travel destination included in the first travel destination history; and a travel destination combination recommendation module that provides travel destination recommendation results by combining the recommended results of the first and second travel destinations on the basis of the frequency of travel destination histories.

Description

Hybrid filtering-based travel destination combination recommendation system and method

The present invention relates to a hybrid filtering-based travel destination combination recommendation system and method, and more particularly, by hybrid filtering in which content based filtering and collaborative filtering are mixed. It relates to a hybrid filtering-based travel destination combination recommendation system and method capable of recommending a travel destination combination.

In general, the conventional travel destination recommendation service that recommends a travel destination to a user selects a travel destination that matches the travel-related information input by the user, and provides a travel destination recommendation list in order of similarity to the travel-related information input by the user. is ending at

Due to this, when the user has a plan to travel to several destinations, the user does not consider whether the distance between the destinations and the combination according to the travel order of the destinations satisfy the user's needs, so that the user can It is inconvenient to have to make a travel plan for various destinations by checking the route, etc.

In addition, filtering algorithms such as Contents based Filtering or Collaborative Filtering are known, but the method of recommending a travel destination by such a specific filtering algorithm is limited in recommending the optimal combination of travel destinations to the user for each filtering algorithm. has a

Since content-based filtering is a method of recommending travel destinations based on the similarity between travel-related information and travel destinations input by the user, there is a limitation in not recommending a new travel destination combination based on the travel history of other users similar to the user’s travel taste (preference). have

Collaborative filtering can recommend a combination of travel destinations to the user based on the travel history of other users similar to the user's travel taste, but recommends based on the collected travel-related data. may not proceed.

In addition, the travel destination recommendation method based on collaborative filtering has a problem in that a travel area or combination of travel destinations that are not selected for any user may not be recommended forever. Due to this, there is a limitation in that it is not possible to recommend a suitable travel destination for the user at an initial point in time when sufficient data for travel destination recommendation based on collaborative filtering is not secured.

The present invention provides a hybrid filtering-based travel destination combination recommendation system and method that can recommend an optimal travel destination combination to a user by hybrid filtering in which content based filtering and collaborative filtering are mixed; To provide a recording medium.

In addition, the present invention provides a hybrid filtering-based travel destination combination recommendation system, method, and record capable of recommending an optimal travel destination combination for each user based on the average staying time at a travel destination, time required for a movement route between travel destinations, and a means of transportation to provide a medium.

A hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention includes: a travel destination attribute data generator configured to collect travel destination contents related to a plurality of travel destinations, and generate attribute data for each travel destination from the collected travel destination contents; a data input unit configured to receive, from the first user, a first travel destination history including a travel destination combination according to a time sequence of one travel destination or two or more different travel destinations, and travel destination recommendation request data including a travel period and travel attributes; a content filtering recommendation module configured to recommend one or more first travel destinations among the plurality of travel destinations based on a similarity between the travel destination recommendation request data and the attribute data for each travel destination; Comparing the first travel destination history with second travel destination histories of a plurality of second users by collaborative filtering, and recommending one or more second travel destinations from the travel destination histories of a second user with a high degree of relevance to the first user Collaborative filtering recommendation module; a travel destination history frequency calculation module, configured to calculate a travel destination history frequency, which is a frequency included in second travel destination histories of the plurality of second users, for each travel destination included in the first travel destination history; and a travel destination combination recommendation module configured to provide a travel destination recommendation result to the first user by combining a recommendation result of the one or more first travel destinations and a recommendation result of the one or more second travel destinations based on the travel destination history frequency do.

The travel destination combination recommendation module, when the travel destination history frequency is less than a set reference frequency, applies a weight higher than that of the one or more second travel destinations to the recommendation results of the one or more first travel destinations to obtain a first travel destination recommendation result create; And when the travel destination history frequency is equal to or greater than the reference frequency, the second travel destination recommendation result may be generated by applying a weight equal to or less than the recommendation result of the one or more second travel destinations to the recommendation result of the one or more first travel destinations. .

The travel destination combination recommendation module obtains central positions of recommended travel destinations including the first travel destination and the second travel destination, and applies a higher weight as the distance between the central location and the recommended travel destinations decreases to obtain the travel destination recommendation result create; And based on the travel destination transportation means input by the first user, transportation means convenience and parking-related information of each travel destination, the travel destination recommendation result may be verified to recommend a travel destination combination to the first user.

The travel destination combination recommendation module calculates a travel required time by applying an average staying time at each recommended travel destination and a time required for a moving route between the recommended travel destinations, and recommends when the travel required time exceeds the travel period may be configured to exclude.

The hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention receives feedback of the time the first user stayed in each travel destination and the time taken for movement between the travel destinations according to the travel destination combination selected by the first user, and the travel destination It may further include a feedback module configured to be reflected in the travel destination recommendation by the combination recommendation module.

A hybrid filtering-based travel destination combination recommendation method according to an embodiment of the present invention includes, by a travel destination attribute data generator, collecting travel destination contents related to a plurality of travel destinations, and generating attribute data for each travel destination from the collected travel destination contents; Step of receiving, by the data input unit, from the first user, a first travel destination history including a travel destination combination according to time sequence of one travel destination or two or more different travel destinations, and travel destination recommendation request data including a travel period and travel attributes ; recommending, by a content filtering recommendation module, one or more first travel destinations among the plurality of travel destinations based on the similarity between the travel destination recommendation request data and the attribute data for each travel destination; By the collaborative filtering recommendation module, the first travel destination histories are compared with the second travel destination histories of a plurality of second users by the collaborative filtering, and one or more second destinations are selected from the travel destination histories of the second users having a high degree of relevance to the first user. 2 step of recommending a travel destination; calculating, by the travel destination history frequency calculation module, a travel destination history frequency that is a frequency included in the second travel destination histories of the plurality of second users for each travel destination included in the first travel destination history; and providing, by the travel destination combination recommendation module, a travel destination recommendation result to the first user by combining the recommendation result of the one or more first travel destinations and the recommendation result of the one or more second travel destinations based on the travel destination history frequency includes

The step of providing the travel destination recommendation result may include, when the travel destination history frequency is less than a set reference frequency, applying a weight higher than that of the one or more second travel destinations to the recommendation result of the one or more first travel destinations to obtain a first travel destination generating a recommendation result; and generating a second travel destination recommendation result by applying a weight equal to or less than the recommendation result of the one or more second travel destinations to the recommendation result of the one or more first travel destinations when the travel destination history frequency is equal to or greater than the reference frequency can

The step of providing the travel destination recommendation result includes obtaining central locations of recommended travel destinations including the first travel destination and the second travel destination, and applying a higher weight to the travel destination as the distance between the central location and the recommended travel destinations is shorter. generating a recommendation result; calculating a travel required time by applying an average staying time at each recommended travel destination and a time required for a movement route between the recommended travel destinations, and excluding the travel time from recommendation when the travel required time exceeds the travel period; and

The method may include recommending a travel destination combination to the first user by verifying the travel destination recommendation result based on the travel destination moving means input by the first user, transportation means convenience of each destination, and parking-related information.

In the hybrid filtering-based travel destination combination recommendation method according to an embodiment of the present invention, the time the first user stayed in each travel destination and the time taken for movement between the travel destinations according to the travel destination combination selected by the first user by the feedback module The method may further include the step of receiving feedback and reflecting in the travel destination recommendation by the travel destination combination recommendation module.

According to an embodiment of the present invention, there is provided a computer-readable recording medium in which a program for executing the hybrid filtering-based travel destination combination recommendation method is recorded.

According to an embodiment of the present invention, there is provided a hybrid filtering-based travel destination combination recommendation system and method, and a recording medium that can recommend an optimal travel destination combination to a user by hybrid filtering in which content-based filtering and collaborative filtering are mixed do.

In addition, according to an embodiment of the present invention, a hybrid filtering-based travel destination combination recommendation that can recommend an optimal travel destination combination for each user based on an average staying time in a travel destination, a time required for a movement route between travel destinations, and a means of transportation A system, method, and recording medium are provided.

1 is a configuration diagram of a hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention.
2 and 3 are conceptual diagrams for explaining a travel destination recommendation method of a travel destination combination recommendation module constituting a hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention.
4 is a flowchart of a hybrid filtering-based travel destination combination recommendation method according to an embodiment of the present invention.
5 is a flowchart of step S170 of FIG. 4 .
6 is a flowchart of a travel destination combination recommendation method according to another embodiment of the present invention.
7 is a flowchart of a travel destination combination recommendation method according to another embodiment of the present invention.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and common knowledge in the art to which the present invention pertains It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout.

In the present specification, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless otherwise stated. As used herein, '~ unit, ~ module' is a unit that processes at least one function or operation, and may refer to, for example, software, FPGA, or hardware component. The functions provided by '~ unit, ~ module' may be performed separately by a plurality of components, or may be integrated with other additional components. 'Part, ~ module' in this specification is not necessarily limited to software or hardware, and may be configured to be in an addressable storage medium, or may be configured to reproduce one or more processors. Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

The hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention recommends an optimal travel destination combination to the user by hybrid filtering in which content-based filtering and collaborative filtering are mixed. It may be provided as a travel destination combination recommendation server.

1 is a block diagram of a hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention. Referring to FIG. 1 , a hybrid filtering-based travel destination combination recommendation system 100 according to an embodiment of the present invention may be provided to a server that provides travel destination recommendation information to a user terminal through a network having a wired/wireless communication interface. .

Wired / wireless communication interface, for example, LAN (Local Area Network), ISDN (Integrated Services Digital Network), CDMA (Code Division Multiple Access), WCDMA (Wideband Code Division Multiple Access), GSM (Global System for Mobile communication) , Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wireless Local Area Network (LAN), Wide Area Network (WAN), Wireless Fidelity (WIFI), and the like, but is not limited thereto.

In addition, the hybrid filtering-based travel destination combination recommendation system 100 according to an embodiment of the present invention may be provided in the form of an application that can be downloaded to a user terminal. The user terminal is a terminal used by a user, and may be implemented as, for example, a smartphone, a smart pad, a tablet PC, a notebook computer, a desktop PC, a laptop PC, a netbook computer, and the like, but is not limited thereto.

The hybrid filtering-based travel destination combination recommendation system 100 according to an embodiment of the present invention includes a travel destination attribute data generator 110 , a data input unit 120 , a content filtering recommendation module 130 , a collaborative filtering recommendation module 140 , and a travel destination It includes a history frequency calculation module 150 , a travel destination combination recommendation module 160 , a feedback module 170 , and a database 180 .

The travel destination attribute data generator 110 may collect travel destination contents related to a plurality of travel destinations, and generate attribute data for each travel destination from the collected travel destination contents. The travel destination may be, for example, a specific area with GPS coordinates.

The travel destination content may be searched and collected by a web crawler based on, for example, travel or travel destination related keywords, or may be input by a user terminal, a server management terminal, or an administrator's terminal.

Travel destination contents include, for example, travel destination related blogs, text data from social network services such as Facebook, Instagram, and Twitter, review data related to various destinations, comments, etc. may include Attribute data generated for each travel destination by the travel destination attribute data generator 110 may be stored in the database 180 .

The travel destination attribute data generation unit 110 may generate attribute data on the travel destination by natural language processing on explanatory data related to the travel destination and extracting the travel destination related attribute keyword.

The travel destination attribute data generator 110 may generate attribute data for each destination by performing pre-processing such as morphological analysis and pre-processing (removing stopwords, extracting roots, etc.) on text data included in the travel destination content.

The travel destination attribute data generation unit 110 may generate attribute data in the form of a travel destination attribute word-set by natural language processing text data from travel destination-related content collected from blog text data, review data, comments, etc. about the travel destination. have.

Attribute data may be defined as words representing unique attributes of each travel destination, for example, local location, tourist attraction, park, beach, lake, amusement park, water park, ski resort, museum, art gallery, movie theater, restaurant, It may be defined as various attribute words collected from travel destination contents, such as cafes and activities.

The data input unit 120 includes, from a first user who wants to recommend a travel destination, a first travel destination history including one travel destination or a combination of travel destinations according to the time sequence of two or more different travel destinations, a travel period and travel attributes (eg, family Travel, traveler's age group, activity/recreation/tourism preference, etc.) may be input for recommendation request data.

The travel destination combination may be a combination of travel destinations in an order according to time that a user can visit and move within a specified period within a specific radius. The travel destination combination may consist of a movement route between each travel destination and adjacent destinations.

In an embodiment, when a first travel destination is input from the first user, the data input unit 120 selects one or more travel destinations that can be visited within a travel period within an area set based on the distance to the first travel destination from among a plurality of travel destinations. A user interface that allows the user to input as a second travel destination following the first travel destination may be provided.

The content filtering recommendation module 130 is based on the similarity between the user's travel destination recommendation request data and the attribute data for each travel destination. One or more first travel destinations having similar attributes may be recommended.

The similarity between the user's destination recommendation request data and the attribute data for each destination is, for example, based on Cosine Similarity, Euclidean Distance, Manhattan Distance, Minkowski Distance, etc. may be calculated, but is not limited thereto.

The collaborative filtering recommendation module 140 compares the first travel destination histories input by the first user with the second travel destination histories of a plurality of second users based on collaborative filtering, and the degree of relevance with the first user is One or more second travel destinations may be recommended from the high second user's travel destination history.

In this case, the second user may be a user or users different from the first user. The collaborative filtering recommendation module 140 may recommend a second travel destination to the user through collaborative filtering using the visit histories of people who prefer a travel route of a specific travel destination combination.

Therefore, by the collaborative filtering recommendation module 140 , a travel destination that has not yet been visited between users who have a similar travel taste or a similar travel route is recommended, or a related travel route is recommended according to a classified travel route attribute or theme. can do.

The collaborative filtering recommendation module 140 generates an attribute vector of a combination of travel destinations that the user is selecting or selected, extracts a combination of travel destinations with similar properties through a collaborative filtering algorithm, By extracting the combination, it is possible to recommend a travel destination to the user.

Here, the attribute vector of the travel destination combination may be a vector listing the travel destinations input by the user as being of interest in the order of travel or a vector listing the properties of each of the travel destinations.

For example, the combination of travel destinations selected by user A is "Travel A1 -> Destination A2 -> Destination A3", User B's travel history is "Travel A2 -> Destination A3 -> Destination A4", User A and User B If the travel tastes are similar, a new travel destination combination may be recommended to the user A in consideration of the user B's travel history.

Specifically, in the above example, for user A, taking user B's travel history into consideration, "Travel A1 -> Destination A2 -> Destination A4" or "Travel A1 -> Destination A3 -> Destination A4" or "Travel A2 -> Destination A4" A new travel destination combination can be recommended, such as A3 -> Destination A4" or "Location A1 -> Destination A2 -> Destination A3 -> Destination A4".

In addition, for user B, in consideration of user A's travel history, "Travel A1 -> Destination A3 -> Destination A4" or "Travel A1 -> Destination A2 -> Destination A3" or "Travel A1 -> Destination A2 -> Destination A4" " or "Location A1 -> Destination A2 -> Destination A3 -> Destination A4".

The collaborative filtering recommendation module 140 selects another user having an attribute vector similar to the attribute vector of the travel destination combination input by the user, identifies a travel destination that the user does not input from among the selected attribute vectors of other users, and recommends it to the user. can

Since content-based filtering is a method of recommending a travel destination based on the similarity between travel-related information input by a user and a travel destination, it has a limitation in not recommending a new travel destination combination based on the travel history of other users similar to the user's travel taste.

Collaborative filtering can recommend a combination of travel destinations to the user based on the travel history of other users similar to the user's travel taste, but recommends based on the collected travel-related data. may not proceed.

In addition, in the case of recommending travel destinations based only on collaborative filtering, a travel area or combination of travel destinations that have never been selected by any user may not be recommended to the user forever.

The hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention utilizes all the advantages of such content-based filtering and collaborative filtering and recommends a travel destination history frequency calculation module 150 and travel destination combination to supplement the problems of each filtering technique A module 160 is provided.

The travel destination history frequency calculation module 150 calculates the travel destination history frequency, which is a frequency included in the second travel destination histories of a plurality of second users, for each travel destination included in the first travel destination history recommended by the content filtering recommendation module 130 can be calculated.

The travel destination combination recommendation module 160 may provide a travel destination recommendation result to the first user by combining the recommendation results of one or more first travel destinations and the recommendation results of one or more second travel destinations based on the travel destination history frequency.

Accordingly, the travel destination combination recommendation module 160 may recommend a travel destination combination expected to have a higher user preference among the destination extraction result of the content-based filtering and the destination extraction result of the collaborative filtering to the user.

When the travel destination history frequency calculated by the travel destination history frequency calculation module 150 is less than the set reference frequency, the travel destination combination recommendation module 160 is configured to recommend one or more first travel destinations recommended by the content filtering recommendation module 130 to the recommendation result. The first travel destination recommendation result may be generated by applying a higher weight than the recommendation result of one or more second travel destinations recommended by the collaborative filtering recommendation module 140 .

When the travel destination history frequency is equal to or greater than the reference frequency, the travel destination combination recommendation module 160 is a second travel destination recommended by the collaborative filtering recommendation module 140 with respect to the recommendation result of the first travel destination recommended by the content filtering recommendation module 130 A second travel destination recommendation result may be generated by applying a weight equal to or less than the recommendation result of .

For example, the travel destination combination recommendation module 160 applies the first weight calculated according to the travel history frequency to the travel destination A and the travel destination B recommended by the content filtering recommendation module 130, respectively, to the travel destination of the travel destination A and the destination B Recommendation suitability is calculated, and the second weight calculated according to travel history frequency is applied to travel destination B and travel destination C recommended by the collaborative filtering recommendation module 140 to calculate travel destination recommendation suitability of travel destination B and travel destination C have.

The travel destination combination recommendation module 160 compares the travel destination recommendation suitability calculated by applying the weight as described above, selects one or more travel destinations in the order of travel destination recommendation suitability, and reflects the travel order in consideration of the optimal travel route between one or more travel destinations A combination can be recommended.

2 and 3 are conceptual diagrams for explaining a travel destination recommendation method of a travel destination combination recommendation module constituting a hybrid filtering-based travel destination combination recommendation system according to an embodiment of the present invention.

1 and 2 , the travel destination combination recommendation module 160 obtains a central location (C) of recommended travel destinations including a first travel destination and a second travel destination, and a distance between the central location of the recommended travel destinations and recommended travel destinations As (D1, D2, D3, D4) is shorter, a higher weight (recommended result value) may be applied to generate a travel destination recommendation result.

That is, the travel destination combination recommendation module 160 reflects the distance (D1, D2, D3, D4) between the center location of the travel destinations and each travel destination together with the travel destination history frequency calculated by the travel destination history frequency calculation module 150. By calculating it, it is possible to calculate the recommended fitness of each travel destination and travel destination combination.

Referring to FIGS. 1 and 3 , the travel destination combination recommendation module 160 takes the average time spent in each recommended travel destination (L1, L2, L3, L4) and the travel route (P1, P2, P3) between the recommended travel destinations The travel time is calculated by applying the required time, and if the travel time exceeds the travel period, it may be excluded from the recommendation. The travel destination combination recommendation module 160 may determine whether to exclude the recommendation of the travel destination combination according to Equation 1 below.

[Formula 1]

는 여행지에 머무르는 시간,

는 이동에 소요되는 최대 시간, n은 여행지 조합에 포함된 각 여행지의 순번, k는 여행지 또는 이동 경로를 나타내는 변수이다.In Equation 1,

is the time spent in the destination,

is the maximum time required for movement, n is the sequence number of each travel destination included in the travel destination combination, and k is a variable representing a travel destination or a moving route.

That is, the travel destination combination recommendation module 160 applies the average time of staying at each travel destination among the weighted results, and when the recommended destination destination is added to the combination, the travel required time that is a threshold value corresponding to the travel period input by the user Exceeding this may result in exclusion from travel recommendations.

In addition, the travel destination combination recommendation module 160 provides a means of movement between destinations input by the user (car, train, airplane, bus, subway, taxi, rental car, etc.), convenience of transportation means of each destination, and parking-related information (whether or not a parking lot is provided) , parking cost, etc.), it is possible to recommend a combination of destinations to the user.

Convenience of transportation means and parking related information of a travel destination can be analyzed by extracting a word corresponding to a transportation method from content data about the destination by a web crawler. If the moving means input by the user is a car or a rental car and there is not enough parking in the travel destination, the travel destination combination recommendation module 160 may exclude the recommendation of the corresponding travel destination or lower the recommendation ranking.

As another example, if the means of transportation entered by the user is a bus, but in an area where the bus does not pass and it is difficult to catch a taxi to the travel destination, the recommendation of the destination is excluded or the recommendation ranking is downgraded, and access is easy through the bus You can raise the recommendation priority of one other travel destination.

The feedback module 170 receives feedback of the time the first user stayed in each travel destination and the time spent moving between the travel destinations according to the travel destination combination finally selected by the first user, and reflected in the travel destination recommendation by the travel destination combination recommendation module 160 can do.

The feedback module 170 may generate the final combination selected by the user as a new combination and add it as travel destination combination history data. Also, the feedback module 170 may add a preferred new travel destination combination as feedback data through evaluation of the recommended result.

The database 180 may store attribute data for each travel destination generated by the travel destination attribute data generation unit 110 , feedback data of the user generated by the feedback module 170 , and the like, and the stored travel-related data may be stored later by the user. It can be used for travel recommendations.

According to an embodiment of the present invention, a travel destination combination is recommended based on hybrid filtering that mixes content-based filtering according to travel destination attributes and collaborative filtering according to travel destination combination history, so that in a situation where travel destination-related data is insufficient, content-based filtering is more It is possible to derive the result value with high weight.

In addition, according to an embodiment of the present invention, it is possible to recommend a combination of consecutive travel destinations (travel destination history), and the combination of destinations that have never been selected also derives a corrected result value by content-based filtering, so that the initial data about the destination is When there is not enough, it is possible to recommend a suitable destination for the user based on the properties of the content.

In addition, according to an embodiment of the present invention, it is possible to recommend a travel destination combination in consideration of the travel route and the travel means between the travel destinations and the travel required time according to the travel route to recommend an optimal travel destination combination in consideration of the travel period for each user. have.

4 is a flowchart of a hybrid filtering-based travel destination combination recommendation method according to an embodiment of the present invention. 1 and 4 , the hybrid filtering-based travel destination combination recommendation method according to an embodiment of the present invention recommends an optimal travel destination combination to a user by hybrid filtering in which content-based filtering and collaborative filtering are mixed.

First, the travel destination attribute data generating unit 110 may collect travel destination contents related to a plurality of travel destinations and generate attribute data for each travel destination from the collected travel destination contents ( S110 and S120 ).

The travel destination attribute data generation unit 110 may generate attribute data on the travel destination by natural language processing of explanatory data related to the travel destination and extracting the travel destination related attribute keyword.

The data input unit 120 receives, from the first user, a first travel destination history including a travel destination combination according to a time sequence of one travel destination or two or more different travel destinations, and travel destination recommendation request data including a travel period and travel attributes. can be (S130).

The travel destination combination may be a combination of travel destinations in an order according to time that a user can visit and move within a specified period within a specific radius. The travel destination combination may consist of a movement route between each travel destination and adjacent destinations.

In an embodiment, when a first travel destination is input from the first user, the data input unit 120 determines one or more travel destinations that can be visited within a travel period within an area set based on the distance to the first travel destination among a plurality of travel destinations, and The user may be selected as a selection candidate for a second travel destination following the first travel destination.

The content filtering recommendation module 130 may recommend one or more first travel destinations among a plurality of travel destinations based on the similarity between the travel destination recommendation request data and the attribute data for each travel destination ( S140 ).

The content filtering recommendation module 130 may recommend the first travel destination by content-based filtering for extracting content having similar properties based on the properties of the travel destination-related content.

The collaborative filtering recommendation module 140 compares the first travel destination history with the second travel destination histories of a plurality of second users by collaborative filtering, and receives one or more second A travel destination may be recommended (S150).

The collaborative filtering recommendation module 140 may recommend a second travel destination to the user through collaborative filtering using the visit histories of people who prefer a travel route of a specific travel destination combination.

The collaborative filtering recommendation module 140 may recommend a travel destination that has not yet been visited to users who have a similar taste or preference for a similar route, or recommend a related travel destination route according to attributes or themes of the classified route.

The collaborative filtering recommendation module 140 generates an attribute vector of a combination of travel destinations that the user is selecting or selected, extracts a combination of travel destinations with similar properties through a collaborative filtering algorithm, By extracting the combination, it is possible to recommend a travel destination to the user.

The travel destination history frequency calculation module 150 may calculate a travel destination history frequency, which is a frequency included in the second travel destination histories of a plurality of second users, for each travel destination included in the first travel destination history ( S160 ).

The travel destination combination recommendation module 160 may provide a travel destination recommendation result to the first user by combining the recommendation result of one or more first travel destinations and the recommendation result of one or more second travel destinations based on the travel destination history frequency (S170) .

The travel destination combination recommendation module 160 may recommend a result having a higher preference among the destination extraction result of the content-based filtering and the destination extraction result of the collaborative filtering to the user.

5 is a flowchart of step S170 of FIG. 4 . 1 to 3 , the travel destination combination recommendation module 160 applies a weight higher than that of one or more second travel destinations to a recommendation result of one or more first travel destinations when the travel destination history frequency is less than a set reference frequency to generate a first travel destination recommendation result (S172, S174).

When the travel destination history frequency is equal to or greater than the reference frequency, the travel destination combination recommendation module 160 may generate a second travel destination recommendation result by applying a weight equal to or less than the recommendation result of one or more second travel destinations to the recommendation result of one or more first travel destinations. There is (S172, S176).

6 is a flowchart of a travel destination combination recommendation method according to another embodiment of the present invention. 1, 4 and 6 , the travel destination combination recommendation module 160 obtains central positions of recommended travel destinations including a first travel destination and a second travel destination, and a distance between the central locations of the recommended travel destinations and recommended travel destinations The shorter is, the higher weight (recommendation result) can be applied to generate a travel destination recommendation result (S182, S184).

The travel destination combination recommendation module 160 calculates the travel required time by applying the average staying time at each recommended travel destination and the time required for the movement route between the recommended travel destinations, and excludes from the recommendation if the travel required time exceeds the travel period It can be done (S186, S188).

That is, the travel destination combination recommendation module 160 applies an average time of staying at each travel destination among the weighted results to recommend a travel destination when the “total travel time required” exceeds the threshold value when a recommended destination is added to the combination. can be excluded from

Referring back to FIGS. 1 and 4 , the feedback module 170 receives feedback of the time the first user stayed in each travel destination and the time taken for movement between the travel destinations according to the travel destination combination selected by the first user, and the travel destination combination recommendation module It may be reflected in the travel destination recommendation by ( 160 ) ( S180 ).

The feedback module 170 may generate the final combination selected by the user as a new combination and add it as travel destination combination history data. Also, the feedback module 170 may add a preferred new travel destination combination as feedback data through evaluation of the recommended result.

7 is a flowchart of a travel destination combination recommendation method according to another embodiment of the present invention. Referring to FIGS. 1 and 7 , the travel destination combination recommendation module 160 analyzes transportation means of each travel destination, and parking-related information (whether or not there is a parking lot, parking cost, etc.) (S192), and between destinations input by the user Based on the means of transportation (car, train, airplane, bus, subway, taxi, rental car, etc.), the convenience of transportation in each destination, and parking-related information (whether or not parking lot is available, parking cost, etc.) A travel destination combination may be recommended (S194).

According to the above-described embodiment of the present invention, by mixing content-based filtering according to travel destination properties and collaborative filtering according to travel destination combination history, in a situation where travel destination-related data is insufficient, a result value with higher weight on content-based filtering can be derived. have.

In addition, according to an embodiment of the present invention, it is possible to recommend a combination of consecutive travel destinations (travel destination history), and the combination of destinations that have never been selected also derives a corrected result value by content-based filtering, so that the initial data about the destination is When there is not enough, it is possible to recommend a travel destination based on the properties of the content.

In addition, according to an embodiment of the present invention, it is possible to recommend a travel destination combination in consideration of the travel route and the travel means between the travel destinations and the travel required time according to the travel route to recommend an optimal travel destination combination in consideration of the travel period for each user. have.

The embodiments described above may be implemented by a hardware component, a software component, and/or a combination of a hardware component and a software component. For example, the apparatus, methods and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate (FPGA). Array), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions, may be implemented using one or more general purpose or special purpose computers.

The processing device may run an operating system and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, a processing device is sometimes described as being used, but a person of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It will be understood that this may include

For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a Parallel Processor. The software may include a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device.

The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or device, to be interpreted by or provide instructions or data to the processing device. , or may be permanently or temporarily embody in a transmitted signal wave. The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software.

Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CDROMs and DVDs, and ROM, RAM, and flash memory. Hardware devices specially configured to store and execute program instructions, such as, etc. are included. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with reference to the limited embodiments and drawings, various modifications and variations are possible from the above description by those skilled in the art. For example, the described techniques are performed in an order different from the described method, and/or the described components of the system, structure, apparatus, circuit, etc. are combined or combined in a different form than the described method, or other components Or substituted or substituted by equivalents may achieve an appropriate result. Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

100: Hybrid filtering-based travel destination combination recommendation system
110: travel destination attribute data generation unit
120: data input unit
130: content filtering recommendation module
140: collaborative filtering recommendation module
150: travel destination history frequency calculation module
160: travel destination combination recommendation module
170: feedback module
180: database

Claims (10)

a travel destination attribute data generator configured to collect travel destination contents related to a plurality of travel destinations and generate attribute data for each travel destination from the collected travel destination contents;
a data input unit configured to receive, from the first user, a first travel destination history including a travel destination combination according to a time sequence of one travel destination or two or more different travel destinations, and travel destination recommendation request data including a travel period and travel attributes;
a content filtering recommendation module configured to recommend one or more first travel destinations among the plurality of travel destinations based on the similarity between the travel destination recommendation request data and the attribute data for each travel destination;
Comparing the first travel destination history with second travel destination histories of a plurality of second users by collaborative filtering, and recommending one or more second travel destinations from the travel destination histories of a second user with a high degree of relevance to the first user Collaborative filtering recommendation module;
a travel destination history frequency calculation module, configured to calculate a travel destination history frequency, which is a frequency included in second travel destination histories of the plurality of second users, for each travel destination included in the first travel destination history; and
A travel destination combination recommendation module configured to provide a travel destination recommendation result to the first user by combining a recommendation result of the one or more first travel destinations and a recommendation result of the one or more second travel destinations based on the travel destination history frequency ,
Hybrid filtering-based travel destination combination recommendation system. According to claim 1,
The travel destination combination recommendation module is,
if the travel destination history frequency is less than a set reference frequency, generate a first travel destination recommendation result by applying a higher weight to the recommendation result of the one or more first travel destinations than the recommendation result of the one or more second travel destinations; and
configured to generate a second travel destination recommendation result by applying a weight equal to or less than the recommendation result of the one or more second travel destinations to the recommendation result of the one or more first travel destinations when the travel destination history frequency is equal to or greater than the reference frequency,
Hybrid filtering-based travel destination combination recommendation system. According to claim 1,
The travel destination combination recommendation module is,
obtaining central locations of recommended travel destinations including the first travel destination and the second travel destination, and generating the travel destination recommendation result by applying a higher weight as the distance between the central location and the recommended travel destinations decreases; and
configured to recommend a travel destination combination to the first user by verifying the travel destination recommendation result based on the travel destination transportation means input by the first user, transportation means convenience of each travel destination, and parking-related information,
Hybrid filtering-based travel destination combination recommendation system. 4. The method of claim 3,
The travel destination combination recommendation module is,
configured to calculate the travel required time by applying the average staying time at each recommended travel destination and the time required for the travel route between the recommended travel destinations, and to exclude from recommendation when the travel required time exceeds the travel period,
Hybrid filtering-based travel destination combination recommendation system. 5. The method of claim 4,
A feedback module configured to receive feedback of the time the first user stayed in each travel destination and the time taken for movement between the travel destinations according to the travel destination combination selected by the first user and reflect it in the travel destination recommendation by the travel destination combination recommendation module. Including, hybrid filtering-based travel destination combination recommendation system. collecting, by the travel destination attribute data generator, travel destination contents related to a plurality of travel destinations, and generating attribute data for each travel destination from the collected travel destination contents;
Step of receiving, by the data input unit, from the first user, a first travel destination history including a travel destination combination according to time sequence of one travel destination or two or more different travel destinations, and travel destination recommendation request data including a travel period and travel attributes ;
recommending, by a content filtering recommendation module, one or more first travel destinations among the plurality of travel destinations based on the similarity between the travel destination recommendation request data and the attribute data for each travel destination;
By the collaborative filtering recommendation module, the first travel destination histories are compared with the second travel destination histories of a plurality of second users by the collaborative filtering, and one or more second destinations are selected from the travel destination histories of the second users having a high degree of relevance to the first user. 2 step of recommending a travel destination;
calculating, by the travel destination history frequency calculation module, a travel destination history frequency, which is a frequency included in second travel destination histories of the plurality of second users, for each travel destination included in the first travel destination history; and
The step of providing, by the travel destination combination recommendation module, a travel destination recommendation result to the first user by combining the recommendation result of the one or more first travel destinations and the recommendation result of the one or more second travel destinations based on the travel destination history frequency containing,
A hybrid filtering-based travel destination combination recommendation method. 7. The method of claim 6,
The step of providing the travel destination recommendation result includes:
generating a first travel destination recommendation result by applying a weight higher than that of the one or more second travel destinations to the recommendation results of the one or more first travel destinations when the travel destination history frequency is less than a set reference frequency; and
generating a second travel destination recommendation result by applying a weight equal to or less than the recommendation result of the one or more second travel destinations to the recommendation result of the one or more first travel destinations when the travel destination history frequency is equal to or greater than the reference frequency,
A hybrid filtering-based travel destination combination recommendation method. 7. The method of claim 6,
The step of providing the travel destination recommendation result includes:
obtaining a central location of recommended travel destinations including the first travel destination and the second travel destination, and generating the travel destination recommendation result by applying a higher weight as the distance between the central location and the recommended travel destination decreases;
calculating a travel required time by applying an average staying time at each recommended travel destination and a time required for a movement route between the recommended travel destinations, and excluding the travel time from recommendation when the travel required time exceeds the travel period; and
Comprising the step of recommending a travel destination combination to the first user by verifying the travel destination recommendation result based on the travel destination transportation means input by the first user, transportation means convenience and parking-related information of each travel destination,
A hybrid filtering-based travel destination combination recommendation method. 9. The method of claim 8,
By the feedback module, receiving feedback of the time the first user stayed in each travel destination and the time taken for movement between the travel destinations according to the travel destination combination selected by the first user, and reflecting the feedback in the travel destination recommendation by the travel destination combination recommendation module Further comprising, hybrid filtering-based travel destination combination recommendation method. A computer-readable recording medium in which a program for executing the hybrid filtering-based travel destination combination recommendation method according to any one of claims 6 to 9 is recorded.

Images