KR20200069113A - Method, apparatus and computer program for translating location - Google Patents

Abstract

According to an embodiment of the present invention, provided is a location representation system which comprises: a target location obtaining unit for obtaining a target location; a user location obtaining unit for obtaining a user current location from a user terminal; a reference point information obtaining unit for obtaining information of one or more candidate reference points existing within a predetermined radius from the target position; a reference point selection unit for selecting a first reference point having a fitness score equal to or greater than a reference value among the one or more candidate reference points; and a semantic description generating unit for generating a natural language sentence describing the target location or a path to the target location using the reference point based on the current location of the user.

Description

METHOD, APPARATUS AND COMPUTER PROGRAM FOR TRANSLATING LOCATION}

The present invention relates to a location expression method, system and computer program, and more particularly, to a method, system and computer program written in natural language so that a specific coordinate on a map can be easily recognized offline.

Recently, with the advent of various online to offline (O2O) services and the emergence of location-based services, users are increasingly required to share a specific offline location in real time in a digital environment. In particular, location-based services include various location-based services, such as connecting a person moving through a location or a person having a vehicle (taxi, carpool service, etc.), delivering to a specific location, or messaging through a specific location. Is being developed and provided.

Due to this, it is becoming increasingly important to provide accurate information about offline locations, but since maps are only schematic drawings of the real world, there are real offline spaces and gaps, and even if additional information such as street views is used, it is possible to It was difficult to accurately specify the offline location.

An object of the present invention is to convert coordinate data on a map designated by a location sharer into a semantic description and provide it to a location sharer.

Another object of the present invention is to generate a semantic expression using one or more reference points selected in consideration of visibility and relative distance.

According to an embodiment of the present invention, a target position obtaining unit for obtaining a target position; A user location obtaining unit for obtaining a user current location from a user terminal; A reference point information obtaining unit for obtaining information of one or more candidate reference points existing within a predetermined radius from the target position; A reference point selection unit for selecting a first reference point having a fitness score equal to or greater than a reference value among the one or more candidate reference points; A semantic description generator for generating a natural language sentence describing the target location or a path to the target location based on the user's current location using the reference point; A location representation system is provided.

In the present invention, the reference point selection unit may calculate the suitability score by considering at least one of a visibility weight and a relative distance weight from the target position from the obtained information of the candidate reference point.

In the present invention, when the relative position from the first reference point to the user's current position is less than a preset value, the reference point selection unit is a candidate that exists within a second radius in a path from the first reference point to the user's current position. It may include an additional reference point selection unit for selecting one or more of the additional reference points.

In the present invention, the additional reference point selection unit may calculate the number of the additional reference points in consideration of the coordinates of the first reference point, the coordinates of the user's current location, and a reference distance suitable for the location description.

In the present invention, the additional reference point selection unit considers the suitability score of each candidate additional reference point among the paths from the user's current location consisting of the candidate additional reference points to the reference point to match the number of the additional reference points. Accordingly, one or more candidate additional reference points included in the path having the optimal description order may be selected as additional reference points.

In the present invention, the additional reference point may be a reference point associated with a direction change point on a path from the user's current location to the first reference point.

In the present invention, the candidate reference point is one of a POI reference point or an object reference point, and the POI reference point may be a reference point based on POI included in map data.

In the present invention, the semantic expression can learn a description sentence corresponding to the information about the first reference point and the path, divide a path or sentence for each reference point, and learn each phrase to generate a semantic expression. have.

According to an embodiment of the present invention, a target position acquisition step of acquiring a target position; A user location obtaining step of obtaining a user current location from a user terminal; A reference point information obtaining step of acquiring information of one or more candidate reference points existing within a predetermined radius from the target position; A reference point selection step of selecting a first reference point having a fitness score equal to or greater than a reference value among the one or more candidate reference points; A semantic description generation step of generating a natural language sentence describing the target location or a path to the target location based on the user's current location using the reference point; A method for expressing a position including a is provided.

In the present invention, the reference point selection unit may calculate the suitability score by considering at least one of a visibility weight and a relative distance weight from the target position from the obtained information of the candidate reference point.

In the present invention, when the relative position from the first reference point to the user's current position is less than a preset value, the reference point selection unit is a candidate that exists within a second radius in the path from the first reference point to the user's current position. Including the additional reference point selection unit for selecting one or more additional reference points among the additional reference points, in the present invention, the candidate reference point is one of a POI reference point or an object reference point, and the POI reference point is a reference point based on POI included in map data. Can be.

Additionally, a computer program recorded on a computer readable recording medium for carrying out the method of the present invention is provided.

According to the present invention, by converting the coordinate data on the map to a semantic description (semantic description) made of natural language to provide to the location shareee, it is possible to provide a human-friendly accurate location sharing method.

1 is a diagram illustrating an example of a network environment according to an embodiment of the present invention.
2 is a block diagram illustrating an internal configuration of a user terminal and a server in an embodiment of the present invention.
3 shows an internal configuration of a processor according to an embodiment of the present invention.
4 is a view showing a method for expressing a position in time series according to an embodiment of the present invention.
5 illustrates a map showing a target location and an actual offline space according to an embodiment of the present invention.
6 illustrates an example of obtaining reference point information according to an embodiment of the present invention.
7 is a view showing a method of preferentially considering a POI reference point in time series.
8 is a view showing a method of selecting an additional reference point according to an embodiment of the present invention in time series.
9 shows an example of a path that can be generated by a combination of additional reference points according to an embodiment of the present invention.

For a detailed description of the present invention, which will be described later, reference is made to the accompanying drawings that illustrate, by way of example, specific embodiments in which the present invention may be practiced. These embodiments are described in detail enough to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention are different, but need not be mutually exclusive. For example, specific shapes, structures, and characteristics described in this specification may be implemented by changing from one embodiment to another without departing from the spirit and scope of the present invention. In addition, it should be understood that the position or arrangement of individual components within each embodiment may be changed without departing from the spirit and scope of the present invention. Therefore, the following detailed description is not intended to be done in a limiting sense, and the scope of the present invention should be taken to cover the scope claimed by the claims of the claims and all equivalents thereto. In the drawings, similar reference numerals denote the same or similar components throughout several aspects.

1 is a diagram illustrating an example of a network environment according to an embodiment of the present invention.

The network environment of FIG. 1 includes a plurality of first user terminals 111, 112, 113, 114, second user terminals 121, 122, 123, 124, a plurality of servers 150, 160, and a network ( 160). 1 is not limited to the number of the first and second user terminals or the number of servers as an example for describing the invention.

The plurality of first user terminals 111, 112, 113, 114 or the second user terminals 121, 122, 123, 124 may be a fixed terminal or a mobile terminal implemented as a computer device. Examples of the plurality of user terminals include a smart phone, a mobile phone, navigation, a computer, a laptop, a terminal for digital broadcasting, a PDA (Personal Digital Assistants), a portable multimedia player (PMP), and a tablet PC. For example, the user terminal 1 111 may communicate with other user terminals 112, 113, 114, 121, 122, 131, 141 and/or the server 150 through the network 160 using a wireless or wired communication method. Can communicate.

The communication method is not limited, and a communication method using a communication network (for example, a mobile communication network, a wired Internet, a wireless Internet, and a broadcasting network) that the network 160 may include may include short-range wireless communication between devices. For example, the network 160 includes a personal area network (PAN), a local area network (LAN), a campus area network (CAN), a metropolitan area network (MAN), a wide area network (WAN), and a broadband network (BBN). , Any one or more of the networks such as the Internet. In addition, the network 160 may include any one or more of a network topology including a bus network, a star network, a ring network, a mesh network, a star-bus network, a tree or a hierarchical network, etc. It is not limited.

The server 150 communicates with a plurality of first user terminals 111, 112, 113, 114 or second user terminals 121, 122, 123, 124 through the network 160, and commands, codes, and files. , It may be implemented as a computer device that provides content, services, or a plurality of computer devices. The server 150 may be an apparatus for implementing the location expression method or the location expression system of the present invention.

In one example, the server 150 may provide a file for installing an application to the user terminal 1 111 connected through the network 160. In this case, the user terminal 1 111 may install an application using a file provided from the server 150. In addition, the service provided by the server 150 by accessing the server 150 under the control of an operating system (OS) included in the user terminal 1 111 and at least one program (for example, a browser or an installed application) or Content can be provided. For example, when the user terminal 1 111 transmits a location sharing request through the network 160 to the server 150 under the control of the application, the server 150 transmits a code corresponding to the location sharing request to the user terminal 1 ( 111), and the user terminal 1 111 may provide a semantic expression to the user by configuring and displaying a screen according to a code according to application control.

2 is a block diagram illustrating an internal configuration of a first user terminal, a second user terminal, and a server in an embodiment of the present invention.

In FIG. 2, the internal configuration of the first user terminal 110 and the second user terminal 120 and the server 150 as an example for one server will be described as an integrated example for the first and second user terminals.

The first user terminal 110, the second user terminal 120 and the server 150 include memory 211, 221, 251, processors 212, 222, 252, communication modules 213, 223, 253, and input/output Interfaces 214, 224, and 254. The memory 211, 221, 251 is a computer-readable recording medium and may include a non-permanent mass storage device such as random access memory (RAM), read only memory (ROM), and a disk drive. have. In addition, an operating system and at least one program code (for example, a code for a browser or an application described above, which is installed and driven in the user terminal 1 111) may be stored in the memories 211, 221, and 251. These software components can be loaded from a computer-readable recording medium separate from the memories 211, 221, 251 using a drive mechanism. Such a separate computer-readable recording medium may include a computer-readable recording medium such as a floppy drive, disk, tape, DVD/CD-ROM drive, and memory card. In other embodiments, software components may be loaded into memory 211, 221, 251 through communication modules 213, 223, 253 rather than a computer-readable recording medium. For example, at least one program is a program (an example) in which a file distribution system (for example, the server 160 described above) for distributing installation files of developers or applications is installed by files provided through the network 160. It may be loaded into the memory (211, 221, 251) based on the above-described application).

Processors 212, 222, and 252 may be configured to process instructions of a computer program by performing basic arithmetic, logic, and input/output operations. Instructions may be provided to the processors 212, 222, 252 by memory 211, 221, 251 or communication modules 213, 223, 253. For example, the processors 212, 222, and 252 may be configured to execute instructions received according to program codes stored in a recording device, such as memories 211, 221, and 251.

The communication modules 213, 223, and 253 may provide a function for the first user terminal 110 or the second user terminal 120 and the server 150 to communicate with each other through the network 160, and other users A function for communicating with a terminal or another server (for example, the server 160) may be provided. In one example, the request (eg, location sharing request) generated by the processor 212 of the first user terminal 110 according to the program code stored in the recording device such as the memory 211 is controlled according to the control of the communication module 213 It may be delivered to the server 150 through the network 160. Conversely, control signals, commands, contents, files, etc. provided under the control of the processor 252 of the server 150 are communicated through the communication module 223 and the network 160 to the communication module of the first user terminal 110 It may be received by the first user terminal 110 through (213). For example, control signals or commands of the server 150 received through the communication module 213 may be transmitted to the processor 212 or the memory 211, and contents or files may be transmitted to the first user terminal 110. It may be stored as a storage medium that may further include.

The input/output interfaces 214, 224, and 254 may be means for interfacing with the input/output device 215. For example, the input device may include a device such as a keyboard or mouse, and the output device may include a device such as a display for displaying a communication session of an application. As another example, the input/output interface 214 may be a means for an interface with a device in which functions for input and output are integrated into one, such as a touch screen. As a more specific example, the processor 212 of the first user terminal 110 uses data provided by the server 150 or the second user terminal 120 in processing a command of a computer program loaded in the memory 211. The service screen or the content configured by may be displayed on the display through the input/output interface 214.

In addition, in other embodiments, the first user terminal 110, the second user terminal 120, and the server 150 may include more components than those in FIG. 2. However, there is no need to clearly show most prior art components. For example, the first user terminal 110 or the second user terminal 120 is implemented to include at least a portion of the input/output device 215 described above, or a transceiver, a global positioning system (GPS) module, and a camera , It may further include other components such as various sensors, databases, and the like.

3 shows an internal configuration of a processor according to an embodiment of the present invention.

The configuration for performing the location expression function according to an embodiment of the present invention in the processor 232 of the server 150 is a target location acquisition unit 310, a user location acquisition unit 320, as shown in FIG. A reference point information acquisition unit 330, a reference point selection unit 340 and a semantic expression generation unit 350 may be included. Depending on the embodiment, the components of the processor 232 may be selectively included or excluded from the processor 232. Further, according to an embodiment, the components of the processor 232 may be separated or merged to express the function of the processor 232.

Here, the components of the processor 232 may be expressions of different functions of the processor 232 performed by the processor 232 according to an instruction provided by the program code stored in the server 150. .

The processor 232 and the components of the processor 232 may control the server 150 to perform steps S1 to S5 included in the location representation method of FIG. 4. For example, the processor 232 and components of the processor 232 may be implemented to execute instructions according to code of an operating system included in the memory 231 and code of at least one program.

4 is a view showing a method for expressing a position in time series according to an embodiment of the present invention. Hereinafter, the location expression method, system, and computer program of the present invention will be described in detail with reference to FIGS. 3 and 4 together.

The location sharing method presented in the coordinates on the existing 2D map is not compatible with the actual offline environment, and may cause confusion. For example, when coordinates are set to a specific roadside, it is difficult for the user to intuitively know where the point of the roadside is. Such confusion is exacerbated in environments where the POI occupies a large area, such as the Han River, a plaza, fishing grounds, and large-scale festivals, or does not map to a specific POI. Therefore, there is a need to describe the location in natural language so that both the location sharer and the location shareee can intuitively understand the location without confusion. To this end, the method and system for expressing a location of the present invention may provide coordinate information on a map by converting it to a semantic description. Hereinafter, a method of generating a position semantic expression according to an embodiment of the present invention will be described in detail.

First, the target location acquisition unit 310 acquires a target location (S1). In this case, the target location may be obtained from a user terminal or may be obtained from another server. The target location may be the location of the user terminal itself, or may be another location selected by the user terminal. Another location selected by the user terminal may be a specific location on the map or a specific object location on the map. The target location acquisition unit 310 may acquire the target location as coordinates on the map. In the present invention, the target location may be a location that a location sharer wants to share. According to an embodiment of the present invention, the target location may be three-dimensional coordinates including not only two-dimensional coordinates, but also altitude information derived from information obtained from three or more satellites. Alternatively, according to another embodiment of the present invention, the target location may be a location desired to be shared by a specific service, and the target location may be obtained from a server of the corresponding service. For example, in the case of a mobility service, a location that guides a user to a boarding location may be a target location, and the target location may be obtained from a server of the corresponding mobility service.

According to an embodiment of the present invention, there may be two users using the location expression service. It is a first user who wants to share his/her location or a target location that is his/her designated location with another user, and a second user who wants to know the location shared by another user or how to get to the location. That is, the first user is a location sharer who wants to easily explain the target location, and the second user is a location sharer who wants to describe the target location. In the following embodiments, a location designated by the first user may be referred to as a target location, and a current location of the second user may be referred to as a user current location. In addition, when described as a user without first and second distinction, this may mean a second user. The first user may be a user of the first user terminal 110 described above, and the second user may be a user of the second user terminal 120.

According to an embodiment of the present invention, the target location acquisition unit 310 may receive feedback from the first user whether the target location is accurately specified. To this end, the target location acquiring unit 310 may display the target location obtained on the map, query whether the target location is accurately specified, and correct the target location according to whether or not consent is given. Alternatively, in another embodiment, the target location acquisition unit 310 may present the acquired target location to the first user as a semantic expression, and query whether the target location is accurately specified. In this case, the position semantic expression may be a language representation of a position coordinate at a human level. The detailed description of the position semantic expression will be described later.

In addition, if it is difficult to generate the semantic expression for the acquired target location by the location expression method of the present invention, the target location acquisition unit 310 recommends another location that is easy to generate the semantic expression, and is recommended by the first user If you agree to this, you can correct the position to the target position. For example, if there is no POI or facility around the acquired target location and it is not easy to explain, we recommend the location where the '3m right side pedestrian crossing moves forward', and if the first user agrees, correct the location to the target location can do.

In more detail, the target location obtaining unit 310 determines a case where it is difficult to generate a semantic expression for the acquired target location. The positional expression method of the present invention, which will be described later, generates a semantic expression for a location, and in certain conditions, it may not be possible to generate a semantic expression. For example, if there is no reference point within a predetermined radius from the target position, or if there is no reference point having a fitness score equal to or greater than the reference value, it may be determined that semantic expression of the acquired target position is impossible. In addition, even if a reference point having a fitness score equal to or greater than a reference value exists within a predetermined radius, it may be determined that semantic expression of the acquired target position is impossible when generation of a natural language sentence using the reference point is impossible.

In addition, when the semantic expression for the target location is not possible, the target location obtaining unit 310 selects an alternate location for generating the semantic expression and transmits it to the user terminal, and obtains a consent signal from the user terminal to target the corresponding alternate location Position can be corrected. For example, a position of a reference point having the highest suitability score within a predetermined radius from the acquired target position may be selected as an alternative position for generating a semantic expression. Alternatively, the position of the reference point having a fitness score equal to or greater than the reference value and closest to the target position may be selected as an alternative position for generating a semantic expression.

Next, the user location acquisition unit 320 acquires the user's current location from the user terminal. The user's current location may be physical coordinates obtained using a location measurement module (GPS, etc.) of the user terminal, or may be 3D coordinates like a target location. The user location acquisition unit 320 may acquire the current location of the user in real time.

5 illustrates a map showing a target location and an actual offline space according to an embodiment of the present invention.

5(a) is an example of a map screen in which the target location A and the current location B of the second user (hereinafter, the user current location) are displayed. As shown in (a) of FIG. 5, the target location and the user location on the map may be clearly displayed by coordinate information. However, as can be seen in (b) of FIG. 5 showing the offline world reflected by the map, it is not easy to find a location that exactly corresponds to the coordinates on the map in actual offline. This is because GPS providing coordinate information on a map has errors ranging from a few centimeters to as many km as possible, and it is also difficult for users to quickly match a schematic map with an offline location. In particular, as shown in (b) of FIG. 5, when the range of the POI is wide or the location is not specified as a specific POI, it is more difficult for the user to find an accurate location.

Therefore, in services that need to provide very specific location information such as delivery, carpool, pickup, and mobility services, it is cumbersome for the location sharer and the shareee to share each other's location through direct connection such as phone or messaging in order to accurately share the target location There was. For example, the delivery service user had to share his location with the delivery person by specifying his or her location in words, such as '100m to the right at a traffic light in front of a crossroad gas station'. In the present invention, expressing a specific location, such as '10m to the right at a traffic light in front of a crossroad gas station' in human language, may be referred to as a location semantic description in the present invention.

Next, the reference point information acquisition unit 330 acquires information of one or more candidate reference points existing within a preset radius from the target position (S3). At this time, the preset radius may be the first radius. The reference point of the present invention is used to describe a location or a path in a semantic expression, and may be a point of interest (POI) or an object that actually exists offline. The reference point may be a POI reference point or an object reference point in which a fitness score considering visibility, relative distance, or area is greater than or equal to a reference value.

That is, a POI included in the map data or an object that is not included in the map data but enables location semantic representation may be a reference point. In a specific example, if the positional semantic expression indicating the target location is'in front of the mailbox between the fire station and the post office of the sperm crossroad','sperm crossroad','fire station','post office', and'postbox' may all be reference points. . At this time,'sperm crossroads','fire station','post office' is a POI reference point already registered in the map data, and'post office' may be an object reference point used for location semantic expression although it is not included in the existing map data.

In more detail, according to an embodiment of the present invention, the reference point may be POI. The general map data may include POI data. POI data may be data representing buildings, major facilities, stations, airports, terminals, hotels, etc. located on a map on an electronic map. In the example as described above,'sperm crossroad','fire station', and'post office' are all POI reference points. Since the POI is often more than a certain level of visibility, POI information may be referred to in order to provide a semantic expression to the user.

According to another embodiment of the present invention, the reference point may be an object other than POI. The object reference point is not a POI such as a fire hydrant, a street tree, a traffic light, a flower bed, a telephone pole, etc., but a fixture with high visibility may be an object that can be used for position semantic expression. Since the information of the object reference point is often not included in the map data, it can be obtained through deep learning image analysis of a street view, human engineering, and user experience (eg, a survey). have. The information of the object reference point may be obtained by using possible methods in addition to those illustrated herein.

In addition, the reference point information acquisition unit 330 not only one or more candidate reference points existing in the first radius from the target location, but also information of one or more candidate additional reference points existing in the second radius from the path from the first reference point to the user's current location Can be obtained. That is, the reference point information acquisition unit 330 may set a certain area around the target location or path as the reference point information acquisition area, and acquire information of reference points existing in the reference point information acquisition area.

On the other hand, in order to build the information data of the reference points acquired by the reference point information acquisition unit 330, the location representation system according to an embodiment of the present invention can store the information of the reference point information in the database by semantic analysis and labeling of the map data have. More specifically, the semantics of map data is analyzed using deep learning and POIs or objects related to location representation are extracted. To this end, various elements that can express the environment of roads, places, and spaces as well as basic POIs on a map can be extracted. For example, street trees, street lights, signs, fire hydrants, flower beds, etc. can be extracted and stored in a database, all of the terrain features fixed in the corresponding place. At this time, when storing an object that is a terrain feature, it is possible to store relative information by mapping with POI. For example, a street light can be stored as a'street light in front of McDonald'. In addition, variable objects (people, cars, etc.) that can be noise in location expression can be excluded from the reference point after analysis.

6 illustrates an example of obtaining reference point information according to an embodiment of the present invention.

FIG. 6 is a diagram illustrating reference point information and a selection reference point for semantic expression in the embodiment of FIG. 5.

6(a) is a target position (A), a path (C) from the first reference point to the user's current position (B) in the above-mentioned FIG. 5(a) drawing, and a preset radius from the target position (A) It is the figure which showed the 2nd radius range E from the range D and the path C. At this time, the radius ranges for determining the D region and the E region may be different from each other. As illustrated in FIG. 5, when there is a target location A and a user current location B, and the path C is set as shown in the drawing, the reference point information acquisition unit 330 acquires information of reference points in the D area or the E area. Can be.

In more detail, the reference point information acquisition unit 330 first acquires information of candidate reference points existing in the first radius range D from the target position A. At this time, a selected reference point among candidate reference points existing in the radius range D may be the first reference point. In addition, the reference point information acquisition unit 330 acquires information of candidate additional reference points existing in the second radius range E from the path C. One or more selected reference points among candidate additional reference points existing in the radius range E may be additional reference points.

In the example of (b) of FIG. 6, the reference point within the first radius from the target location (A) may include a subway entrance 612, a bus sign 625, and a traffic light 622 for a vehicle. In addition. The reference point within the second radius (E) from the path (C) may include a bus sign 625, a pedestrian crossing 615, a flowerbed 623, a pedestrian traffic light 621, a vehicle traffic light 622.

In addition, referring to (b) of FIG. 6, the reference points may include a crossroad 611, a subway entrance 612, a church 613, a building 614, and a pedestrian crossing 615 as POI reference points. In addition, as an object reference point, there may be a pedestrian traffic light 621, a vehicle traffic light 622, a flower bed 623, a power pole 624, and a bus sign 625.

Next, the reference point selection unit 340 selects one or more reference points having a fitness score equal to or greater than the reference value. At this time, the selected reference point is a reference point used to generate the position semantic expression of the present invention. In the above example, the reference points used in the positional semantic expression of'before the mailbox between the fire station and the post office at the sperm crossroad' are'sperm crossroad','fire station','post office', and'postbox'. That is, the reference point selection unit 340 selects a descriptive reference point to be used for the actual position semantic expression among the numerous reference point information acquired by the reference point information acquisition unit 330.

According to an embodiment of the present invention, the reference point selection unit 340 may include a fitness score calculation unit 341, a first reference point selection unit 342, and an additional reference point selection unit 343.

First, the fitness score calculation unit 341 calculates the fitness score of the reference points by considering at least one of the visibility weight, the relative distance weight, and the area weight of the reference points. According to an embodiment of the present invention, the fitness score calculation unit 341 assigns a higher weight as the visibility is higher, the relative distance from the target position (or other reference point) is shorter, and the smaller the floor area occupied by the reference point is, the more suitable it is. You can calculate your score. Also, the fitness score calculation unit 341 may calculate a fitness score using different calculation methods for the POI reference point and the object reference point. In particular, the method of calculating the visibility weight of the POI reference point and the object reference point may be different.

In more detail, the fitness score calculating unit 341 may consider the popularity of the corresponding POI when calculating the visibility weight with respect to the POI reference point. For example, when the POI is a franchise, the popularity of the franchise may be reflected in calculating the weight of the POI. Popularity is a value that reflects how many times the POI has been viewed by users, and may be a value based on QC (Quality Count)/CC (Click Count), Review count/score in the integrated search or map search. Alternatively, the visibility weight of the POI reference point may be determined based on the physical size or color of the POI. For example, if the POI is a building, the larger the size of the building, the higher the visibility weight, or the primary or chromatic color, the higher the weight than the achromatic color. In addition, the larger the size of the signboard, the higher the visibility weight can be assigned.

In addition, the fitness score calculation unit 341 may consider how much the object is preferred when describing the position of the object reference point when calculating the visibility weight. Since there is no way to measure the popularity of the object reference point, the visibility weight can be calculated based on the experience data on how much the object is preferred when describing the location. For example, the visibility weight of the object reference point can be calculated by reflecting how much preferred object is generally used for location description, and such preference can be based on a survey result. In addition, the size and color of the object reference point can be considered, or the visibility and weight can be calculated by tagging and machine learning a suitable and prominent object to describe the position in the image. For example, after learning the distance view image itself using machine learning, an object can be extracted to calculate the visibility weight or overall fitness score of each object.

In addition, the fit score calculating unit 341 may give the area weight higher as the area of the physical space occupied by the corresponding reference point is smaller. For example, in the case of a building, although visibility is very high, there are various spaces such as a building edge, a building front door, a building back door, and a front face of the building. The same is true for a reference point such as a range. However, in the case of a pedestrian traffic light, although the visibility is less than that of a building, it is rather advantageous to specify an accurate location due to the small space occupied. Accordingly, according to an embodiment of the present invention, when selecting a reference point score, the area distance as well as the relative distance and visibility may be considered and calculated.

Also, the fitness score calculation unit 341 may calculate the fitness score by further considering the history of location sharing by the semantic expression using the corresponding reference point. In more detail, when there is a history provided to other users of the semantic expression generated by the location expression method of the present invention, the suitability scores of reference points included in the semantic expression may be adjusted. For example, if the semantic expression using the reference point A is shared with other users, the suitability of the reference point A based on the history of other users agreeing or recommending the semantic expression or moving the other user in the path indicated by the semantic expression You can increase your score. That is, the fitness score calculating unit 341 may adjust the fitness score of the reference point A in consideration of the user's feedback on the semantic expression using the reference point A. Accordingly, according to an embodiment of the present invention, since the suitability scores of the reference points are adjusted based on the feedback of other users, the reference point having the adjusted suitability score can be used when generating the next semantic expression, thereby improving the accuracy of the semantic expression. Next, the first reference point selection unit 342 selects a first reference point having a fitness score equal to or greater than a reference value from one or more candidate reference points existing within the first radius from the target location. In more detail, the first reference point may be selected as a reference point having the highest suitability score among candidate reference points around the target location. According to the above-described embodiment of FIG. 6, the bus sign 625 having the highest suitability score among candidate reference points located in the first radius D from the target position A may be selected as the first reference point. That is, according to the embodiment of FIG. 6, the first reference point may be a bus sign 625 that is an object reference point.

In an optional embodiment of the present invention, the POI reference point may be preferentially set as the first reference point. That is, the first reference point is first considered whether there is a POI reference point, and if there is no suitable POI reference point, it may be an object reference point. In step S3 and S4 of FIG. 3 described above, information of the reference point is acquired regardless of the type of the reference point (POI reference point, object reference point) and a first reference point is selected among them. On the other hand, in another embodiment of the present invention, POI reference point information is first acquired and selected, and then, if there is no suitable POI reference point, object reference point information may be obtained and selected. In this regard, steps S3 and S4 of FIG. 4 may be modified as shown in FIG. 7.

7 is a view showing a method of preferentially considering a POI reference point in time series.

Referring to FIG. 7, the reference point information acquisition unit 330 acquires information of one or more POI reference points existing within a preset radius from the target position (S31 ). At this time, the preset radius may be the first radius. Further, the reference point selection unit 340 determines whether there is a POI reference point having a fitness score equal to or greater than a predetermined value among the one or more POI reference points (S32), and if so, the first reference point of the POI reference point having the highest fitness score among the one or more POI reference points It can be selected (S33).

If there is no POI reference point having a fitness score equal to or greater than a predetermined value, the reference point information acquisition unit 330 obtains information of one or more object reference points existing within a predetermined radius from the target location (S34), and the reference point selection unit ( In step 340, the object reference point having the highest suitability score among the one or more object reference points is selected as the first reference point (S35).

Next, the additional reference point selection unit 343 determines the number of additional reference points based on the distance between the first reference point and the user, and selects the additional reference points. The additional reference point is a reference point that is additionally selected in consideration of the difficulty of the user finding the first reference point itself when the distance between the first reference point and the user is greater than or equal to the reference value.

8 is a view showing a method of selecting an additional reference point according to an embodiment of the present invention in time series. The process of FIG. 8 may be performed by an additional reference point selection unit 343.

Referring to FIG. 8, after step S4 of FIG. 4, first, the additional reference point selection unit 343 determines whether the relative distance between the first reference point and the user's current position is less than or equal to the reference value (S41 ). This is because even if the target position is semanticly expressed based on the first reference point, if the user cannot find the first reference point itself, it is difficult to accurately recognize the target position with the corresponding expression. Accordingly, according to an embodiment of the present invention, when the first reference point is more than a predetermined distance from the user's current location, an additional reference point is selected. If the relative distance is less than the reference value, the process proceeds to step S5 without selecting an additional reference point.

Next, when the relative distance is not equal to or less than the reference value, information of one or more reference points existing within a preset radius is obtained from a path between the first reference point and the user's current position (S42). In more detail, when the relative position from the first reference point to the user's current position is less than a predetermined value, candidate additional reference point information existing within a second radius is acquired in a path from the first reference point to the user's current position. At this time, information about the candidate additional reference point may be obtained by the reference point information acquisition unit 330.

Next, the number of additional reference points is calculated in consideration of the relative distance between the first reference point and the user's current position (S44). According to an embodiment of the present invention, the number of additional reference points may be determined by a formula such as [Equation 1] below.

[Equation 1]

In Equation 1, N _AP is the number of additional reference points, BP is the coordinates of the reference points, UP is the coordinates of the user, SD is a reference distance suitable for location description (eg, 10m), and distance(p1, p2) is It may be a relative distance between p1 and p2. For example, when the distance (BP, UP) between the reference point and the user is 50 m and SD is 10 m, the number of additional reference points calculated may be four.

Next, an additional reference point having a fitness score equal to or greater than a predetermined value is selected to match the calculated number of reference points (S45). In more detail, among the paths from the first reference point consisting of candidate additional reference points to the user's current location, the candidate included in the path having the optimal description order in consideration of the suitability score of each candidate additional reference points to match the number of additional reference points Additional reference points can be selected as additional reference points. At this time, when selecting an additional reference point, an optimal description order may be determined by considering a relative distance between each additional reference point.

According to an embodiment of the present invention, the additional reference point selection unit 343 may select a reference point associated with a corresponding rotation point as an additional reference point when a user needs to change direction when moving. For example, if you need to turn right with a traffic light, the traffic light can be selected as an additional reference point.

9 shows an example of a path that can be generated by a combination of additional reference points according to an embodiment of the present invention.

Referring to FIG. 9, p0 may be a first reference point and p7 may be a user's location. In addition, p1 to p6 are candidate addition reference points, and arrows connecting candidate addition reference points indicate paths from p0 to p7 that can be derived by connecting each candidate addition reference point. More specifically, in the embodiment of FIG. 7, a possible path may be p0->p1->p2->p3->p7.

According to an embodiment of the present invention, it is possible to derive an optimal description order among all possible paths and select additional reference points included in a path having an optimal description order as additional reference points.

At this time, the optimal description sequence P can be expressed by the following [Equation 2].

[Equation 2]

In Equation 2, P is an optimal description order, path is a possible path from a first reference point to a user, distance (pi, pj) is a relative distance between pi and pj, and w _v (pi) is This is the fitness score of pi. Considering the values of all possible paths, the optimal order of explanation can be found. In the example of FIG. 7, the first reference point p0 is a bus sign 625, p1 is a flower bed 623, p2 is a pedestrian traffic light 621, p3 is a pedestrian crossing 615, and p4 is a subway entrance 612. , p5 may be a power pole 624 and p6 may be a subway entrance 621. That is, all arrow connections p0 to p7 shown in FIG. 7 may be in the order of description, but in the following embodiment, it is assumed that p0->p1->p2->p3->p7 is a path satisfying the optimal description order. can do. That is, the target location may be generated as a natural language sentence including a sequence of a bus sign->flower bed->pedestrian traffic light->crosswalk.

Next, the description generation unit 350 generates a semantic expression, which is a natural language sentence describing a target location or a path to the target location, using the reference point based on the user's current location (S5). As described above, since the optimal description order corresponds to a path from the first reference point to the user's current location, the description generation unit 350 may invert the corresponding path to generate a semantic expression based on the user. That is, in the previous example, if p0->p1->p2->p3->p7 is the selected reference point and path, the user's current location p7 is referenced in the order of p7->p3->p2->p1->p0. You can create a semantic expression made up.

According to an embodiment of the present invention, the semantic expression may be generated with a phrase including a reference point. According to an embodiment of the present invention, the semantic expression is obtained by learning information after obtaining the information on the first reference point and the route and the corresponding description sentence, and dividing the route or sentence according to the first reference point or the additional reference point, and for each phrase. You can learn to generate semantic expressions. In addition, learning by the first reference point or additional reference points, information on the previous or next reference point on the path may be considered. In addition, according to an embodiment of the present invention, when learning, a Recurrent Neural Network (RNN) may be used.

In an optional embodiment, when it is determined that an additional reference point is not necessary in addition to the first reference point, the natural language sentence may be generated in a format that describes the relative position from the first reference point, not the path. For example, it may be a semantic expression in which'right 1m from a traffic light' and'in front of a subway entrance stairs' are generated.

According to another embodiment of the present invention, it may be determined that an additional reference point is required, and an optimal description sequence combining the first reference point and the additional reference point may be determined. The path may be described using first reference points and additional reference points included in the optimal description order. Continuing with the examples of FIGS. 6 and 7, the '2m right point on the bus sign 625 that crosses the pedestrian crossing 615 and turns to the left with a flower bed existing on the left side of the pedestrian traffic light 621' is targeted It may be a semantic expression describing the position A.

At this time, the description may be expressed based on the user moving the path. For example, when describing the right side, the left side, etc., the corresponding path may be described based on the direction the user looks when moving. Continuing with the example of FIG. 6, when the user is moving in the direction of crossing the pedestrian crossing 615, the flower bed 623 is located on the left with respect to the pedestrian traffic light 621 and reflects this point to reflect the natural language sentence You can create The user can more clearly grasp the target position by receiving the description.

The embodiment according to the present invention described above may be implemented in the form of a computer program that can be executed through various components on a computer, and such a computer program can be recorded on a computer-readable medium. In this case, the medium may continuously store a program executable on a computer or may be stored for execution or download. In addition, the medium may be various recording means or storage means in the form of a combination of single or several hardware, and is not limited to a medium directly connected to a computer system, but may be distributed on a network. Examples of the media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks, And program instructions including ROM, RAM, flash memory, and the like. In addition, examples of other media include an application store for distributing applications, a site for distributing or distributing various software, and a recording medium or storage medium managed by a server.

In the above, the present invention has been described by specific matters such as specific components and limited examples and drawings, but it is provided to help a more comprehensive understanding of the present invention, and the present invention is not limited to the above embodiments, and Those skilled in the art to which the invention pertains may seek various modifications and changes from these descriptions.

Therefore, the spirit of the present invention is not limited to the above-described embodiments, and should not be determined, and the scope of the spirit of the present invention as well as the claims to be described later, as well as all ranges equivalent to or equivalently changed from the claims. Would belong to

Claims (13)

A target location obtaining unit for obtaining a target location;
A user location obtaining unit for obtaining a user current location from a user terminal;
A reference point information obtaining unit for obtaining information of one or more candidate reference points existing within a predetermined radius from the target position;
A reference point selection unit for selecting a first reference point having a fitness score equal to or greater than a reference value among the one or more candidate reference points;
A semantic description generator for generating a natural language sentence describing the target location or a path to the target location using the reference point based on the current location of the user;
Location representation system comprising a. According to claim 1,
The reference point selecting unit calculates the suitability score by considering at least one of a visibility weight and a relative distance weight from the target position from the obtained information of the candidate reference point. According to claim 1,
The reference point selection unit,
If the relative position from the first reference point to the user's current position is less than a preset value, one or more additional reference points among candidate additional reference points existing within a second radius in a path from the first reference point to the user's current position are selected. Positioning system, including an additional reference point selection. The method of claim 3,
The additional reference point selection unit calculates the number of the additional reference points in consideration of the coordinates of the first reference point, the coordinates of the user's current location, and a reference distance suitable for the location description. The method of claim 4,
The additional reference point selection unit, in order to match the number of the additional reference points, among the paths from the user's current location made up of the candidate additional reference points to the reference point, considers a suitability score of each candidate additional reference point to determine an optimal description order A location representation system that selects one or more candidate additional reference points included in a path to have as an additional reference point. The method of claim 3,
And the additional reference point is a reference point associated with a turning point on a path from the user current position to the first reference point. According to claim 1,
The candidate reference point is one of a POI reference point or an object reference point, and the POI reference point is a reference point based on POI included in map data. According to claim 1,
The semantic expression is a location expression system for learning a description sentence corresponding to the information about the first reference point and the path, dividing a path or a sentence for each reference point, and learning for each phrase to generate a semantic expression. A target position acquisition step of acquiring a target position;
A user location obtaining step of obtaining a user current location from a user terminal;
A reference point information obtaining step of acquiring information of one or more candidate reference points existing within a predetermined radius from the target position;
A reference point selection step of selecting a first reference point having a fitness score equal to or greater than a reference value among the one or more candidate reference points;
A semantic description generation step of generating a natural language sentence describing the target location or a path to the target location using the reference point based on the current location of the user;
Position representation method comprising a. The method of claim 9,
In the step of selecting the reference point, the suitability score is calculated by considering at least one of a visibility weight and a relative distance weight from the target position from the obtained information of the candidate reference point. The method of claim 9,
The reference point selection step,
If the relative position from the first reference point to the user's current position is less than a preset value, one or more additional reference points among candidate additional reference points existing within a second radius in a path from the first reference point to the user's current position are selected. A method for expressing a location, comprising the step of selecting an additional reference point. The method of claim 9,
The candidate reference point is one of a POI reference point or an object reference point, and the POI reference point is a reference point based on POI included in map data. A computer program recorded on a computer-readable recording medium for carrying out the method according to any one of claims 9 to 12.

Images