JP7428688B2 - Place recommendation methods, systems, and programs - Google Patents

Description

The present invention relates to a method, system, and program for recommending places.

As technology develops, electronic devices (e.g., smartphones, tablet PCs, etc.) become more prevalent, and thus many parts of daily life become increasingly dependent on the Internet.

With the development of various technologies including the Internet, the traditional consumption pattern that was highly dependent on offline has gradually shifted to online, and now consumption is centered around online. is increasing exponentially.

Furthermore, even in industry fields that are offline-based due to the nature of their services, such as the travel industry and the food industry, it has become commonplace to attract customers through various online marketing methods.

Along with these changes, in the offline-based industrial field, reservation services using the Internet have been introduced in order to improve user convenience. Therefore, customers, that is, users, can now make reservations for various places such as accommodation facilities and restaurants with just a few clicks on an electronic device connected to the Internet.

As such a technology, Patent Document 1 discloses a general-purpose reservation method and system based on service collaboration, and technology related to a recording medium, and provides an online service reservation and payment process.

On the other hand, when a user makes a reservation for a specific place using an electronic device, the user exhibits a behavior pattern in which they find food, things to play with, things to enjoy, etc. by searching surrounding places related to the specific place.

Therefore, there is a need for a method that not only provides a reservation service for a specific location to a user, but also recommends various locations related to the specific location.

Korean Publication Patent No. 10-2016-0105090

The present invention provides a place recommendation method, system, and program that can recommend information related to reservation information based on reservation information.

More specifically, the present invention provides a place recommendation method, system, and program that can recommend meaningful related places to the user, centering on the reserved place corresponding to the reservation information.

Further, the present invention provides a place recommendation method, system, and program that can recommend related places that match the purpose of use of a user who uses a reserved place.

Further, the present invention provides a place recommendation method, system, and program that can extract meaningful related information based on date information defined by reservation information.

In order to solve the above problems, the place recommendation method according to the present invention includes the steps of: collecting reserved place information corresponding to the reserved event based on the reserved event performed on the user terminal; and searching with the user terminal. Collecting related place information related to the reserved event based on the search information, and generating sentences including words respectively corresponding to the reserved place information and the related place information. , performing word embedding on the sentence to represent the words included in the sentence on a vector space; The method may include the step of extracting recommended place information related to the reserved place information from the space.

Further, the place recommendation system according to the present invention includes: a communication unit that communicates with a user terminal; a control unit that receives reserved place information corresponding to the reserved event based on the reserved event performed on the user terminal; including. The control unit receives related place information related to the reserved event based on search information searched by the user terminal, and selects words corresponding to the reserved place information and the related place information, respectively. A sentence containing the sentence is generated, a word embedding is performed on the sentence, the words included in the sentence are expressed on a vector space, and the distance between the words on the vector space is calculated. Based on the vector space, recommended place information related to the reserved place information can be extracted from the vector space.

Further, the program executed by one or more processes in the electronic device according to the present invention and storable in a computer-readable recording medium is configured to generate reservation place information corresponding to the reservation event based on the reservation event performed on the user terminal. a step of collecting related location information related to the reserved event based on the search information searched by the user terminal; and a step of collecting related location information related to the reserved event based on search information searched by the user terminal; a step of generating a sentence (sentence) containing a word); a step of performing word embedding on the sentence and expressing the words included in the sentence on a vector space; extracting recommended place information related to the reserved place information from the vector space based on the distance between the words.

As described above, the place recommendation method, system, and program according to the present invention collects related place information searched on the user terminal, centering on the reserved event that occurred on the user terminal, and makes a reservation corresponding to the reserved event. It is possible to recommend location information that is related to the location information.

At this time, in the present invention, by expressing the reserved place information and related place information on a vector space based on word embedding, it is possible to extract related information based on the reserved place information.

In addition, the place recommendation method, system, and program according to the present invention secure information related to the reserved place information by collecting related information centering on the date on which the reserved event occurred and the date corresponding to the reserved event. can do.

1 is a conceptual diagram for explaining a place recommendation system according to the present invention. 3 is a flowchart for explaining a place recommendation method according to the present invention. FIG. 3 is a conceptual diagram for explaining a method of collecting reserved place information and related place information according to the present invention. FIG. 2 is a conceptual diagram for explaining a word embedding method for place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining a word embedding method for place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining a word embedding method for place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining a word embedding method for place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining a word embedding method for place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining a word embedding method for place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining a word embedding method for place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining the use of place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining the use of place recommendation according to the present invention. FIG. 2 is a conceptual diagram for explaining the use of place recommendation according to the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, but the same or similar components will be denoted by the same reference numerals regardless of the drawing numbers, and overlapping description thereof will be omitted. . The suffixes "module" and "part" used in the following explanations are given or mixed to facilitate the preparation of specifications, and they themselves have significance and usefulness. It's not something you have. Furthermore, when describing the embodiments of the present invention, if it is determined that detailed explanation of related known techniques would obscure the gist of the embodiments of the present invention, the detailed explanation will be omitted. Furthermore, the attached drawings are only for helping understanding of the embodiments of the present invention, and the technical idea of the present invention is not limited by the attached drawings, and any changes that fall within the idea and technical scope of the present invention may be made. , equivalents or substitutes.

Terms including ordinal numbers such as "first", "second", etc. are used to describe various components, but the components are not limited to the above terms. These terms are only used to distinguish one component from another.

When a component is referred to as being "coupled" or "connected" to another component, it may be directly coupled or connected to the other component, or there may be further intermediate components. It should be understood that it may exist. On the other hand, when a component is referred to as being "directly coupled" or "directly connected" to another component, there are no intermediate components. .

Singular expressions include plural expressions unless otherwise specified.

As used herein, terms such as "comprising" and "having" are intended to specify that the features, numbers, steps, acts, components, parts, or combinations thereof described herein are present. It should be understood that this does not exclude the existence or possible addition of one or more other features, figures, steps, acts, components, parts or combinations thereof.

The present invention is for generating and providing useful information to the user and any third user based on information collected from the user's terminal, and in particular, for generating and providing useful information to the user and any third user. and how to process data.

In the present invention, for convenience of explanation, "reservation of a place" will be mainly explained, but the present invention is not necessarily limited to this. In other words, "reservation" in the present invention includes not only a reservation for a place, but also all kinds of reservations, such as reservations for services (including reservations for services made on the Internet), reservations for providing food and beverages, and reservations for participation in any event. Includes reservations.

Note that the term "place" in the present invention refers to a place where a certain thing is done or occurs, regardless of its type, and may be a place that is geographically, conceptually, or physically divided.

As an example, a "place" in the present invention may be a place for commercial purposes, and may include a specific business office classified by business office number or the like. In this case, locations include accommodation facilities (e.g., hotels, pensions, guesthouses, condominiums, resorts, etc.), restaurants, theme parks, amusement facilities, beauty management salons (e.g., hair salons, nail salons, etc.), and real estate agencies. , including establishments with specific commercial purposes, such as hospitals. In this case, the "reservation of a place" in the present invention includes a reservation for lodging to use the business office, a reservation for the use of a restaurant, etc.

For example, as shown in FIG. 1, a user uses a terminal 1000 to reserve a place (for example, an accommodation facility called "Pension Yongli").

As another example, a "place" in the present invention may be a place for non-commercial purposes. In this case, places include places that are operated non-commercially for public or public benefit purposes or other purposes (for example, museums, buildings of certain foundations, etc.). In this case, the "reservation of a place" in the present invention refers to a reservation for viewing a museum, a reservation for using a public facility, etc. to use the place.

As described above, in the present invention, "place" includes various concepts, and "reserved place" is not limited in type and location as long as it is a place that can be reserved for use by a user.

Therefore, the term "reservation place" in the present invention refers to a place that can be reserved using a reservation system that can be accessed from a user's terminal. Users can access various reservation systems and make reservations for specific locations through applications, web browsers, programs, software, etc. installed on their terminals.

In the present invention, there is no particular limitation on the reservation system that provides reservations for places.

Further, the term "related place" in the present invention refers to a place related to the reserved place, and refers to a place included in information searched by a user's terminal.

Here, the searched information (or search information) refers to information searched by a search service portal or the like that provides a search service.

Also, the searched information refers to specific information selected or clicked by the user among the search results of the search term input into the user's terminal.

Although it has been explained above that reservations for places and information searches are performed on the "user's terminal," this also means a user account. That is, a user's terminal refers to a terminal connected to or logged into a user account.

Here, the user account refers to an account already registered in the place recommendation system according to the present invention. Such a user account refers to a user ID (identification, identification number).

Note that, as shown in FIG. Parks, etc.) regarding how to recommend places. Hereinafter, the place recommendation system and method according to the present invention will be explained in more detail together with the concept of place mentioned above. FIG. 1 is a conceptual diagram for explaining a place recommendation system according to the present invention.

First, as shown in FIG. 1, a user terminal (or a user's terminal) 1000 means an electronic device. Here, the electronic device includes at least one of a smartphone, a mobile phone, a tablet PC, a kiosk, a computer, a notebook computer, a digital broadcast terminal, a PDA (Personal Digital Assistant), and a PMP (Portable Multimedia Player). include. Further, the user terminal 1000 may be an electronic device to which a user account is logged in, connected, or registered.

Hereinafter, a repeated explanation of the user account logged into the user terminal 1000 will be omitted. It goes without saying that the user terminal 1000 described below is understood to be an electronic device with a logged-in user account or an electronic device from which information can be collected by the place recommendation system 100 according to the present invention.

Meanwhile, the place recommendation system 100 according to the present invention may include at least one of a communication unit 110, a storage unit 120, and a control unit 130.

The communication unit 110 can communicate with at least one of the user terminal 1000, an external storage (for example, a database (DB) 140), an external server, and a cloud server.

On the other hand, an external server or a cloud server may be configured to play at least a part of the control unit 130. That is, execution of data processing, data calculation, etc. may be performed on an external server or a cloud server, and the method is not particularly limited in the present invention.

On the other hand, the communication unit 110 can support various communication methods in accordance with communication standards of communication targets (for example, electronic equipment, external servers, devices, etc.).

For example, the communication unit 110 supports WLAN (Wireless LAN), Wi-Fi (registered trademark) (Wireless Fidelity), Wi-Fi Direct (Wireless Fidelity Direct), DLNA (Digital Living Network Alliance), WiBro (Wireless Broadband), WiMAX (World Interoperability for Microwave Access), HSDPA (High-Speed Downlink Packet Access), HSUPA (High-Speed Uplink Packet Access), LTE (Long Term Evolution), LTE-A (Long Term Evolution-Advanced), 5G (5th Generation) Mobile Telecommunication), Bluetooth (registered trademark) (Bluetooth), RFID (Radio Frequency Identification), IrDA (Infrared Data Association), UWB (Ultra Wide Band), ZigBee, NFC (Near Field Communication), and Wireless Universal Serial Communication with the communication target may be performed using at least one of the following technologies.

Next, the storage unit 120 may store various information related to the present invention. In the present invention, the storage unit 120 may be included in the place recommendation system 100 itself. On the other hand, at least a portion of the storage unit 120 may be at least one of the database (DB) 140 and cloud storage (or cloud server). That is, the storage unit 120 is understood to have no physical space limitations as long as it is a space in which information necessary for the place recommendation system and method according to the present invention is stored. Therefore, hereinafter, the storage unit 120, database 140, external storage, and cloud storage (or cloud server) will not be differentiated, and all will be referred to as the storage unit 120.

The information stored in the storage unit 120 to provide the location recommendation method according to the present invention may include reserved location information regarding a reserved location and related location information regarding a related location.

First, "reservation place information" refers to information about a reservation place corresponding to a reservation event that occurs on a user terminal. Become.

A "reservation event" in the present invention refers to a reservation for a specific location made on a user terminal, and occurs based on a user's selection.

A user may use a user terminal to perform a reservation process for use of a particular location.

Reservations for such places can be made on a reservation service page provided by the reservation system. A reservation system allows a user to reserve a place on a page accessible via the Internet (eg, a web page, an application execution screen, etc.). Such a reservation service page may be provided by the place recommendation system according to the present invention, or may be provided by another entity.

On the other hand, the storage unit 120 stores not only the reserved place information, but also information about the reservation date (for example, the date, period, time, etc. for using the reserved place) and information about the person who made the reservation (for example, the user) (for example, At least one of the following information may also be stored: name, address, telephone number, user account (ID), payment method, payment date, etc.).

Next, "related place information" refers to information about a place related to a reserved event, and is a place searched by the user terminal where the reserved event occurred or the user account logged into the user terminal where the reserved event occurred. Contains information about.

As described above, searched information (or search information) refers to information searched by a user terminal, and refers to information searched by a search service portal that provides a search service. In addition, the searched information refers to specific information selected or clicked by the user among the search results of the search term input to the user terminal.

Meanwhile, the reserved location information and related location information may be matched and stored in the storage unit 120. Matching of these pieces of information may be performed based on user terminals or user accounts.

The storage unit 120 may match and store reserved place information regarding reserved events occurring under the same user account and searched related place information.

Next, the control unit 130 is configured to control the overall operation of the place recommendation system 100 according to the present invention. The control unit 130 may process signals, data, information, etc. input or output from the above components, or may provide or process appropriate information or functions to the user.

The control unit 130 includes at least one central processing unit (CPU) and can perform functions according to the present invention. Further, the control unit 130 can perform artificial intelligence-based data processing, and in particular, can execute the place recommendation method according to the present invention by embedding words. Furthermore, the control unit 130 can execute the place recommendation method according to the present invention using at least one of machine learning and deep learning.

Hereinafter, the location recommendation method according to the present invention will be specifically explained with reference to the above-mentioned configuration and the accompanying drawings.

FIG. 2 is a flowchart for explaining a place recommendation method according to the present invention, and FIG. 3 is a conceptual diagram for explaining a method for collecting reserved place information and related place information according to the present invention. 4, FIG. 5, FIG. 6, FIG. 7, FIG. 8, FIG. 9, and FIG. 10 are conceptual diagrams for explaining the method of embedding words for place recommendation in the present invention. Furthermore, FIGS. 11, 12, and 13 are conceptual diagrams for explaining the use of place recommendation in the present invention.

In the place recommendation method according to the present invention, as shown in FIG. 2, first, a process of collecting reserved place information corresponding to a reserved event is performed (S210).

When the control unit 130 determines that a reserved event regarding a specific location has occurred in the user terminal, the controller 130 collects reserved location information corresponding to the reserved event.

In (a) of FIG. 3, a reservation event regarding a specific place (eg, "Jeju Guest House") occurs on the user terminal 1000.

The control unit 130 includes not only the reserved place information but also information regarding the reservation date (for example, the reservation confirmation date, the reservation execution date, the date, period, and time of using the reserved place) and the person making the reservation (for example, the user). Collect at least one of related information (for example, name, address, telephone number, user account (ID), payment method, payment date, etc.). The control unit 130 stores at least one of the collected reserved place information and related information in the storage unit 120.

The reserved place information collected by the control unit 130 includes not only information related to a specific user but also information related to an unspecified number of users. Information can be collected.

Next, in the present invention, a process of collecting related location information related to the reserved event is performed (S220).

As mentioned above, related location information refers to information about a location related to a reserved event, and is a location searched by the user terminal where the reserved event occurred or the user account logged into the user terminal where the reserved event occurred. Contains information about.

Also, searched information (or search information) refers to information searched by a user terminal, and refers to information searched by a search service portal that provides a search service. In addition, the searched information refers to specific information selected or clicked by the user among the search results of the search term input to the user terminal.

The control unit 130 collects at least part of the information searched by the user terminal (or user account) where the reservation event occurred as related location information.

The control unit 130 may collect information that satisfies preset criteria from among the searched information as related location information. That is, the related location information may be information that satisfies preset conditions.

Here, the preset conditions may relate to at least one of date, time, and period. For example, the control unit 130 may specify location information included in search information searched by a user terminal during a preset period related to a reserved event as related location information.

That is, the control unit 130 may collect location information included in search information searched by a user terminal during a preset period related to a reserved event as related location information. The related location information includes information regarding at least one location.

A pre-defined time period associated with a Booked Event means: i) the specific date on which the Booked Event occurs; ii) the booking date corresponding to the Booking Event (the date on which the use of the Booking Location is reserved); and iii) the date on which the Booking Event occurs. iv) a reservation period that includes the corresponding reservation date (for example, if one or more nights are booked); iv) dates before or after a specified period relative to the specified date on which the booking event occurred; v) a reservation that corresponds to the booking event. Refers to at least one of the following dates within a specific period based on the date.

In the present invention, the reason why location information included in search information searched during a preset period is collected as related location information is to reflect the user's behavior pattern. That is, the user searches for the date on which a reservation was made for a specific location or the date on which the specific location is actually used, and other locations to use in addition to the specific location before and after those dates. Here, other places to be used are places corresponding to food, things to play with, things to enjoy, etc.

Therefore, when a reservation event regarding a specific place occurs on a specific date (for example, October 30, 2020) as shown in (a) of FIG. and (c), at least one piece of search information searched by the user terminal during a preset period that satisfies preset criteria based on the specific date (date on which the reservation event occurred). can be collected as related location information.

As an example, as shown in FIG. 3(b), the control unit 130 may specify a specific period (e.g., 3 days) based on a specific date (e.g., October 30, 2020) on which the reservation event occurred. At least some of the search information 321, 322, and 323 searched by the user terminal on dates before and after (for example, October 31, 2020) may be collected as related location information.

The control unit 130 collects the information searched by the user terminal, and from the searched information, the control unit 130 collects the information searched in a period that satisfies preset conditions based on the date on which the reserved event occurred (or the searched information). The location information included in the information provided can be extracted. Further, the extracted information can be specified as related location information.

Referring to the figure, the extracted information 320 is information regarding locations such as Hyobchae Haenyeo House 321, Hyobchae Beach 322, and Hanrim Park 323.

As another example, the control unit 130 may control the reservation period including the reservation date (for example, November 20, 2020, reference numeral 311) corresponding to the reservation event, as shown in FIGS. At least some of the search information 331, 332, 333, and 334 searched by the user terminal before and after the period (for example, when the reservation is made for one night or more) or the period of time can be collected as related place information.

The control unit 130 collects the information searched by the user terminal, and from the searched information, determines the reservation period including the reservation date corresponding to the reservation event, or the information searched before and after the period (or the searched information). (location information contained in) can be extracted. Further, the extracted information can be specified as related location information.

Referring to the figure, the extracted information 330 is information regarding locations such as Seongsan Ilchulbong 331, Hamdeok Beach 332, Hyupchae Beach 333, and Camellia Hill 334.

In this way, when information is searched by the user terminal, the control unit 130 collects the searched information and extracts (or sorts) relevant place information related to the reserved event from the collected information. Can be done.

On the other hand, in order to extract information related to the reserved event, the control unit 130 extracts related location information from information searched on a specific date (or specific period) based on the date related to the reserved event. Information can be extracted.

In addition, in the present invention, noise information that is not related place information related to a reserved event can be removed from the information searched by the user terminal, and the technique for removing noise information is not particularly limited.

Meanwhile, when the reserved place information and the related place information are extracted from the user terminal, the control unit 130 may match or associate the reserved place information and the related place information so as to satisfy at least one condition. .

As an example, the control unit 130 may associate reserved place information and related place information for each reserved event. In that case, mutually associated information exists for each reserved event. That is, even if the reservation events have the same reservation location, if they are different reservations, they will be processed as different information. In that case, there are multiple data sets (or groups) corresponding to each reserved event.

Therefore, if there are multiple reservation events for the same reservation location, there are multiple data sets (or groups) that each contain associated location information that matches the same reservation location for the number or more of the multiple reservation events.

As another example, the control unit 130 may associate reserved place information and related place information on a per reserved place basis.

In that case, the control unit 130 associates all related place information corresponding to different reserved events with each other even if the reserved events are different, if the reserved locations are the same.

In that case, even if there are multiple reservation events, if all the reservation locations are the same, there is a single data set (or group) containing related location information that is correlated with each other based on the reservation location.

Furthermore, as another example, if there are multiple reserved events and all the reserved locations are the same, the control unit 130 divides the data into multiple data sets (or groups) according to preset criteria, and associates the data sets with the reserved location information. It may also be stored in association with location information. In that case, the preset criteria vary widely, such as the gender, age, residential area (or search area), and reservation date of the user who made the reservation or searched for information.

That is, the control unit 130 can group the reserved place information and the related place information by reflecting the characteristics of the user or the characteristics of the date.

Further, in addition to the above example, the control unit 130 can store reserved place information and related place information in association with each other.

On the other hand, when reserved place information and related place information related to a reserved event are collected by at least one of the various methods described above, in the present invention, a sentence including words corresponding to the reserved place information and related place information, respectively, is collected. A process of generating is performed (S230).

As shown in FIG. 4, the control unit 130 combines a word 430 corresponding to a reserved place (for example, Jeju Guest House) corresponding to the reserved place information 410 and a word 440 corresponding to the related place information 420, and As shown, a sentence 510 can be generated.

As illustrated, when there are a plurality of related places, the control unit 130 stores words 441, 442, 443, 444, and 445 corresponding to the plurality of related places 421, 422, 423, 424, and 425, respectively, and reservation place information 410. A sentence 510 can be generated using the word (word 1) 430 corresponding to . Referring to FIG. 5, the generated sentence 510 includes words 511 (511a, 511b, 511c) corresponding to the reserved place information 410, and words 512, 513, 514, 515, 516 corresponding to the related place information 420, respectively. and is included.

In the following, for convenience of explanation, "words corresponding to reserved place information" are referred to as "reserved place words", and "words corresponding to related place information" are referred to as "related place words".

In the present invention, the control unit 130 generates a sentence centering on the reserved place word, and the reserved place word is understood as a global context. When generating a sentence based on the reserved event or the reserved place, the control unit 130 can generate the sentence by globally inputting the reserved place word.

To this end, as shown in FIG. can be generated. In that case, as shown in the figure, the reserved place word 511 (511a, 511b, 511c) is arranged multiple times in the generated sentence 510.

The control unit 130 can arrange the reserved place word 511 (511a, 511b, 511c) multiple times via at least one of the plurality of related place words 512, 513, 514, 515, 516. As shown in the figure, at least one related location word is arranged between reserved location words 511 (511a, 511b, 511c).

Furthermore, related location words may be placed multiple times in the generated sentence. When the user terminal 1000 searches for search information corresponding to a specific related place word multiple times, the same related place word is arranged multiple times in the generated sentence.

On the other hand, the number of times the reserved place words 511 (511a, 511b, 511c) are arranged in the generated sentence and the arrangement interval are determined based on various criteria.

For example, the distance between the reservation location word and each of the plurality of related location words is determined based on the date on which the search information corresponding to each of the plurality of related location words was retrieved and the date associated with the reservation event.

The control unit 130 can give a weight value to the related place word by adjusting the distance between the reserved place word and the related place word, taking into consideration the date when the search information corresponding to the related place word was retrieved. can.

Here, the distance between the reserved place word and each of the plurality of related place words may relate to how many other related place words are placed between the reserved place word and the particular related place word. The distance between words exists from 0 to n depending on the number of related place words placed between them.

The control unit 130 determines how close to the related place word the reserved place word is to be placed and how many reserved place words to be placed around it, depending on the date on which the search information corresponding to the related place word has been retrieved. be able to.

For example, as shown in FIG. 7, the control unit 130 selects the reservation date (for example, November 20, 2020, The word 726 (for example, Hamdeok Beach) corresponding to the search information retrieved in FIG.

In this way, in the present invention, a sentence can be generated using the reserved place word and the related place word, and a sentence to be the target of word embedding can be secured.

In particular, the control unit 130 can assign a relationship between the reserved place word and the related place word by arranging the reserved place word multiple times in the sentence.

On the other hand, as shown in FIG. 5, the control unit 130 can generate various sentences 510, 520, and 530 mainly using various reserved place words, and can change the number of sentences and the length of the sentences depending on the situation. It can be adjusted as appropriate.

As shown in FIG. 5, when there are different reserved events related to different locations, the control unit 130 can generate sentences 510, 520, and 530 respectively corresponding to the different reserved events. As mentioned above, each of the texts 510, 520, 530 may all relate to different reservation locations, or at least one may relate to the same reservation location.

Furthermore, it goes without saying that when a reservation event regarding a different reservation location or another reservation event regarding the same location occurs on the same user terminal, the control unit 130 can generate a different text for each reservation event. .

In addition, in the present invention, the reserved place word and the related place word are the name of the place that can be identified by the user (or the name of the business, for example, Jeju Guest House, Hyupchee), as shown in FIGS. It may also consist of a diver's house, etc.). In addition, in the present invention, the reserved place word and the related place word are information in a form that can be processed by the control unit 130, as shown in FIG. It may be made to consist of an ID (identification number) of a place (or an ID of a business office) corresponding to a word. That is, the words shown in each box shown in FIGS. 4, 5, and 7 are for explanation, and the actual data corresponding to the words are as shown in FIG. 6 (or in a similar format). ).

Referring to FIG. 6, in the sentence 610, the reserved place word (word 1) 611 (611a, 611b, 611c) is arranged multiple times. Furthermore, in the sentence 610, at least one related place word (word 2, word 3) 612, 613 is included at least once.

As mentioned above, when sentences related to reserved place information and related place information are generated, the present invention includes a process of embedding words in the generated sentences and converting the words included in the sentences into vectors. This is done (S240).

The control unit 130 embeds words in the generated sentence, and can express the words included in the sentence on a vector space as shown in FIGS. 8 and 9.

On the other hand, in step S230, including the reserved place word multiple times to generate a sentence together with the related place word means that when embedding the word, there is a relationship between the reserved place word and the related place word as related information. This corresponds to work for giving.

The control unit 130 can embed words using at least one sentence 510, 520, 530, 610, and 620, as shown in FIGS. 5 and 6.

Word embedding is a method for Natural Language Processing (NLP), the most core technology of artificial intelligence that allows computers to understand and analyze human language. It is.

According to such word embedding, each word included in a sentence can be expressed as a vector, and such a vector is called a "word vector" (or word vector). A word represented by a vector has specific coordinates corresponding to the vector in vector space. Therefore, the positions of words in the vector space are relatively close or far depending on whether the words are similar or not.

The control unit 130 corresponds to each of the reserved place words and related place words expressed on a vector space, which is derived based on the result of embedding words in sentences containing the reserved place words and related place words. Using the distance between word vectors (or word vectors), it is possible to extract related place information that is related to the reserved place information.

The control unit 130 can embed words based on a POI (Point of Interest) 2vec algorithm. Furthermore, the control unit 130 can embed words based on the word2vec algorithm.

The control unit 130 can embed words that convert each of the reserved place word and the related place word into vectors based on the arrangement relationship between the reserved place word and the related place word in the sentence. Here, the dimensions of the vectors respectively corresponding to the reserved place word and the related place word are defined in various ways.

It goes without saying that the control unit 130 can embed words using at least one of various algorithms that can derive the relationships between words included in a sentence, in addition to the above-mentioned algorithm. .

After converting the words included in the text into vectors, the present invention performs a process of extracting recommended place information related to the reserved place information (S250).

Here, the reserved place information from which the recommended place information is extracted may be the reserved place information collected based on the reserved event in step S210, or the reserved place information specified by a different reserved event. It may be.

That is, in the present invention, a database for extracting recommended place information is constructed by utilizing reserved place information and related place information collected through various reservation events, and based on the database, useful information for various users is created. can be provided. That is, the reserved place information from which the recommended place information is extracted may be reserved place information corresponding to a reserved event of any user.

Therefore, it goes without saying that the reserved place information from which the recommended place information is extracted may be a place corresponding to the collected related place information.

On the other hand, the process of extracting recommended place information does not necessarily have to be performed consecutively with the process of embedding words. That is, when it is necessary to provide recommended place information due to a user's request or the occurrence of a new reserved event, the control unit 130 may perform a process of extracting recommended place information from the vector space.

In order to extract recommended place information, the control unit 130 may perform a cosine similarity calculation between vectors expressed on a vector space.

The control unit 130 may perform a cosine similarity calculation between the reserved place word vector and the related place word vector corresponding to the reserved place word and the related place word, respectively.

That is, the control unit 130 can calculate the similarity between the reserved place word vector and the related place word vector by performing a cosine similarity calculation between the reserved place word vector and the related place word vector. The control unit 130 can perform a cosine similarity calculation between all vectors converted by word embedding or between vectors that satisfy preset conditions, centering on the reservation place word vector.

By such cosine similarity calculation, the reserved place word vector and the related place word vector are expressed on vector spaces 800 and 900, as shown in FIGS. 8 and 9.

The control unit 130 calculates the cosine similarity between the reserved place word vector and the related place word vector to calculate the reserved place word vectors 810, 910 and the related place word vectors 821, 822, 823, 824, 825, 826, 921, The relative distance between 922, 923, 924, 925, 926 can be derived.

The control unit 130 also controls at least one word vector 821 , 822 , 823 , 824 , 825 , 921 , 922 , 923 , 924 located within a preset distance 820 , 920 from the reserved place word vector 810 , 910 , 925 and 926 can be specified. The word vector specified in this way is a word vector corresponding to a place to be recommended, and is hereinafter referred to as a "recommended place word vector."

The control unit 130 can extract place information corresponding to at least one of the recommended place word vectors 821, 822, 823, 824, 825, 921, 922, 923, 924, 925, and 926 as recommended place information. .

In this way, the recommended place word vectors 821, 822, 823, 824, 825, 921, 922, 923, 924, 925, 926 located within the preset distance from the reserved place word vectors 810, 910 are reserved. It can be specified by cosine similarity calculation between the location word vectors 810, 910 and a plurality of word vectors included on the vector spaces 800, 900.

The control unit 130 can extract locations corresponding to a preset number of word vectors arranged in order of proximity to the reserved location word vectors 810 and 910 as recommended locations.

The control unit 130 can rank the word vectors arranged in order of proximity to the reservation place word vectors 810 and 910. Furthermore, the control unit 130 can extract locations corresponding to the recommended location word vectors as recommended locations (or recommended location information) based on the ranking.

For example, as shown in FIG. 10(a), based on the reserved place word vector, cafe 1 1011, cafe 2 1012, theme park 1 1013, cafe 3 1014, cafeteria 1 1015, cafeteria 2 1016, museum 1 1017, Recommended place information is extracted in this order.

On the other hand, when recommended place word vectors located in order of proximity to the reserved place word vector are extracted in this way, the recommended places corresponding to the extracted recommended place word vectors may have low diversity. For example, when four highly ranked recommended locations are recommended, three of the four recommended locations may fall under the same category (for example, "cafe").

Therefore, the control unit 130 may normalize the recommended place information based on the category so that the recommended place information includes a variety of places that fall under different categories. Such normalization may be performed on all word vectors included in the vector space, or may be performed on recommended place word vectors specified as recommended place information.

On the other hand, the control unit 130 specifies a preset number of recommended place word vectors located close to the reserved place word vector for each different category based on the reserved place word vector, and then uses the recommended place word vectors specified for each category. Recommended place information can be determined by normalizing between place word vectors.

That is, the control unit 130 can normalize the recommended place word vector so that the recommended place information includes place information corresponding to different categories.

The control unit 130 can normalize the recommended place word vectors by dividing the cosine similarity value of each recommended place word vector by log ₂ (1+rank). Here, the rank is information corresponding to a ranking given based on how close each recommended place word vector is to the reserved place word vector within the category in which each recommended place word vector is included.

As a result of normalization, the control unit 130 selects place information that falls under various categories, such as a cafe, a theme park, a restaurant, and an art museum, as a recommended place, as shown in FIG. 10(b). It can be extracted as information.

As described above, in the place recommendation system and method according to the present invention, useful place information can be recommended to any user based on the similarity between reserved place information and related place information collected from various users. Can be done.

As an example of utilizing the place recommendation method according to the present invention, as shown in FIGS. 11(a) and 12(a), when a reservation event regarding a specific place occurs by an arbitrary user, the control unit 130 Recommended place information can be recommended. Such recommended place information may be provided based on the user's selections 1110, 1210.

The control unit 130 identifies a word vector corresponding to the reserved place information (for example, "Pension Yongli") corresponding to the reserved event in the vector space constructed by the above method, and selects a preset word vector from the identified word vector. Location information 1120 and 1220 corresponding to the word vectors located within the specified distance can be provided to the user terminal 1000, as shown in FIG. 11(b) and FIG. 12(b).

As another example of utilization, as shown in FIG. 13, when a user terminal requests a search for a specific search term on the search service page, the control unit 130 stores location information related to the specific search term in a vector. It can be extracted from the space and provided as recommended place information. As shown in FIGS. 13(a) and 13(b), the control unit 130 not only retrieves information 1311 and 1321 corresponding to the input specific search word, but also the user terminal (or Recommended location information 1322 regarding reserved events associated with the user account (user account) may further be provided. At this time, the control unit 130 may provide the recommended place information 1322 regarding the reserved event only when the reserved place information corresponding to the reserved event is related to a specific search term. For example, as shown in FIG. 13(b), the specific search term is a gourmet restaurant on "Jeju Island," and the reservation location information corresponding to the reserved event is an accommodation facility on "Jeju Island." The reservation location information corresponding to ``Jeju Island'' and the specific search term have a common feature in the area.

In this way, if the information regarding the location included in the search term is related to the reserved location information corresponding to the reserved event, the control unit 130 controls the user terminal or the user of the user account logged in to the user terminal. By providing a place related to reserved place information corresponding to a reserved event as recommended place information, useful information can be provided to the user at various times.

As described above, the place recommendation method, system, and program according to the present invention collects related place information searched on the user terminal, centering on the reserved event that occurred on the user terminal, and makes a reservation corresponding to the reserved event. It is possible to recommend location information that is related to the location information.

At this time, in the present invention, by expressing the reserved place information and related place information on a vector space based on word embedding, it is possible to extract related information based on the reserved place information.

In addition, the place recommendation method, system, and program according to the present invention secure information related to the reserved place information by collecting related information centering on the date on which the reserved event occurred and the date corresponding to the reserved event. can do.

On the other hand, the present invention described above can be implemented as a program that is executed by one or more processes on a computer and can be stored on a computer-readable medium (or recording medium).

Moreover, the present invention described above can be realized as computer readable code or commands on a program recording medium. That is, the present invention can be provided in the form of a program.

On the other hand, the computer-readable medium includes any type of recording device on which data that can be read by a computer system is recorded. Examples of computer readable media include HDD (Hard Disk Drive), SSD (Solid State Disk), SDD (Silicon Disk Drive), ROM, RAM, CD-ROM, magnetic tape, floppy (registered trademark) disk, optical data storage. equipment, etc.

The computer-readable medium also includes storage and can be a server or cloud storage that the electronic device can access via communication. In this case, the computer can download the program according to the invention from the server or cloud storage by wired or wireless communication.

Furthermore, in the present invention, the above-mentioned computer is an electronic device equipped with a processor, that is, a central processing unit (CPU), and its type is not particularly limited.

On the other hand, the detailed description of the present invention is illustrative and should not be construed as restrictive in any respect. The scope of the present invention should be determined by reasonable interpretation of the appended claims, and all modifications within the range of equivalency of the present invention are included within the scope of the present invention.

110 communication section 120 storage section 130 control section 140 database (DB)
1000 user terminals

Claims (9)

A place recommendation method, comprising:
a processor in the computer collects reservation location information corresponding to the reservation event based on the reservation event that occurs in the user account;
the processor collecting relevant location information related to the reserved event based on search information searched for the user account;
the processor generating a sentence including words corresponding to the reserved location information and the related location information, respectively;
the processor embeds words in the sentence and expresses the words included in the sentence on a vector space;
the processor extracting recommended location information related to the reserved location information from the vector space;
including;
The step of extracting the recommended place information includes:
Identifying at least one recommended place word vector located within a preset distance from the reserved place word vector corresponding to the reserved place information on the vector space;
normalizing the recommended place word vector for each category that includes information about the recommended place word vector;
A method characterized by: The related location information includes a plurality of different related location information,
The step of generating the sentence includes:
generating the sentence by combining a reserved place word corresponding to the reserved place information and a plurality of related place words respectively corresponding to the plurality of related place information;
The method according to claim 1. The step of generating the sentence includes:
arranging at least one of the reserved place words between the plurality of related place words and generating the sentence;
The method according to claim 2. In the above sentence,
The distance between the reserved place word and the plurality of related place words is
Search information corresponding to each of the plurality of related place words is determined based on a search date and a date associated with the reserved event.
The method according to claim 3. The step of collecting relevant location information includes:
Collecting location information included in search information searched for the user account during a preset period related to the reserved event as the related location information,
The method according to any one of claims 1 to 4. The step of generating the sentence includes:
When a plurality of different reservation events occur in at least one account including the user account,
Generating a plurality of sentences corresponding to each of the plurality of reserved events,
embedding words in the plurality of sentences and expressing words included in each of the plurality of sentences on the vector space;
including
A method according to any one of claims 1 to 5. The step of extracting the recommended place information includes:
extracting place information corresponding to at least one of the recommended place word vectors as the recommended place information;
including,
A method according to any one of claims 1 to 6. a communication unit that communicates with a user terminal with a logged-in user account;
a control unit that receives reservation location information corresponding to the reservation event based on the reservation event performed in the user account;
The control unit includes:
receiving relevant location information related to the booked event based on search information searched for the user account;
generating sentences including words corresponding to the reserved place information and the related place information, respectively;
embedding words in the sentence, expressing the words included in the sentence on a vector space,
extracting recommended place information related to the reserved place information from the vector space;
Identifying at least one recommended place word vector located within a preset distance from the reserved place word vector corresponding to the reserved place information on the vector space;
normalizing the recommended place word vector for each category that includes information about the recommended place word vector;
A place recommendation system characterized by: A place recommendation program comprising a plurality of computer instructions, the location recommendation program comprising:
The computer instructions, when executed by the computer's processor, cause the computer to:
collecting reservation location information corresponding to the reservation event based on the reservation event made in the user account;
collecting relevant location information related to the booked event based on search information searched for the user account;
generating a sentence including words corresponding to the reserved place information and the related place information, respectively;
embedding words in the sentence and expressing the words included in the sentence on a vector space;
extracting recommended place information related to the reserved place information from the vector space;
run the
The step of extracting the recommended place information includes:
Identifying at least one recommended place word vector located within a preset distance from the reserved place word vector corresponding to the reserved place information on the vector space;
normalizing the recommended place word vector for each category that includes information about the recommended place word vector;
A place recommendation program featuring: