Map generation device, map generation method, and computer program

Abstract

To provide technology that helps reduce cost for generating maps in a map generation device.SOLUTION: The map generation device comprises: an input unit to which is inputted natural language information in which position information on a plurality of points is represented by a natural language expression; and a conversion unit for converting the natural language expression of natural language information inputted to the input unit into a map expression, and thereby generating map information in which the position information is represented by the map expression. The conversion unit includes a language information conversion unit for converting the natural language expression of natural language information into an intermediate expression and thereby generating intermediate information in which the position information is represented by the intermediate expression, and a map information generation unit for converting the intermediate expression of the intermediate information into a map expression and thereby generating map information.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a map generating device, a map generating method, and a computer program. [Background technology]

[0002] 2. Description of the Related Art Conventionally, map generating devices that generate a map using information relating to distances to surrounding objects obtained by a sensor are known (for example, see Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2007-113918 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, even with the prior art such as Patent Document 1, there is still room for improvement in terms of technology for reducing the cost of generating a map in a map generating device.

[0005] The present invention has been made to solve the above-mentioned problems, and has an object to provide a technique for reducing the cost of generating a map in a map generating device. [Means for solving the problem]

[0006] The present invention has been made to solve at least part of the above-mentioned problems, and can be realized in the following forms.

[0007] (1) According to one aspect of the present invention, there is provided a map generating device, comprising: an input unit to which natural language information expressing position information of a plurality of points in a natural language expression is input; and a conversion unit that generates map information in which the position information is expressed in the map expression by converting the natural language expression of the natural language information inputted at the input unit into a map expression, the conversion unit having a language information conversion unit that converts the natural language expression of the natural language information into an intermediate representation to generate intermediate information in which the position information is expressed in the intermediate representation, and a map information generation unit that converts the intermediate representation of the intermediate information into a map representation to generate the map information.

[0008] According to this configuration, the conversion unit generates intermediate information expressed in an intermediate expression from natural language information that expresses the location information of multiple locations in a natural language expression, and then generates map information expressed in a map expression from the intermediate information. This makes it possible to make the location information of multiple locations that is in a person's brain and not specifically expressed into specific map information. Therefore, for example, compared to acquiring information by sensing with a sensor to generate a map, it is possible to reduce the cost of generating map information including the location information of multiple locations, and therefore the cost of generating a map can be reduced.

[0009] (2) In the map generating device of the above embodiment, the natural language information may include information about a movement action from a first point to a second point among the multiple points and information about space, and the language information conversion unit may generate the intermediate information by dividing the natural language information into multiple combinations of information about the movement action and information about space. According to this configuration, the language information conversion unit generates the intermediate information by dividing information about a movement action from a first point to a second point among the multiple points and information about space into multiple combinations of information about the movement action and information about space. This makes it easier for the map information generation unit, which generates map information from the intermediate information, to generate the map information. Therefore, the cost of generating a map can be further reduced.

[0010] (3) The map generating device of the above embodiment may further include a storage unit that stores the map information generated by the map information generating unit, and the map information generating unit may determine whether the newly generated map information is the same as the map information stored in the storage unit, and when it is determined that the generated map information is not the same as the map information stored in the storage unit, may add the generated map information to the storage unit and store it. According to this configuration, when it is determined that the newly generated map information is not the same as the map information stored in the storage unit, the map information generating unit adds the newly generated map information to the storage unit and stores it. This allows map information to be accumulated in the storage unit, thereby expanding the range covered by the map information.

[0011] (4) The map generating device of the above embodiment may further include a storage unit that stores the map information generated by the map information generating unit, the map information generating unit generates intermediate information from the map information stored in the storage unit, the language information converting unit generates natural language information from the intermediate information generated by the map information generating unit, and the map generating device may further include an output unit that outputs the natural language information generated by the language information converting unit to an external device. According to this configuration, the output unit can output the map information stored in the storage unit as natural language information expressed in a natural language expression that can be understood by humans. This allows the map information stored in the storage unit to be effectively used in fields where it is used by humans to understand it, such as route guidance for pedestrians and vehicle drivers.

[0012] (5) The map generating device of the above embodiment may further include a route information generating unit that generates a route between at least two of the multiple points by using the map information stored in the storage unit. According to this configuration, the route information generating unit can generate a route between at least two of the multiple points by using the map information stored in the storage unit. This makes it possible to use the map information stored in the storage unit to generate a travel route.

[0013] (6) In the map generating device of the above embodiment, the input unit may input information about a departure point and information about a destination point, and the route information generating unit may use the map information stored in the storage unit to determine whether or not a travel route from the departure point to the destination point can be generated, and when it is determined that the travel route cannot be generated, the output unit may output a notification that prompts the user to input information necessary for generating the travel route. According to this configuration, the route information generating unit uses the input information about the departure point and the destination point to determine whether or not a travel route from the departure point to the destination point can be generated. When it is determined that the travel route cannot be generated, the route information generating unit outputs a notification that prompts the user to input natural language information necessary for generating the travel route from the output unit. In this way, even when the map information stored in the storage unit cannot generate a travel route, the map generating device can generate a travel route by inputting natural language information necessary for generating the travel route.

[0014] (7) In the map generating device of the above embodiment, the input unit may be capable of inputting additional information regarding the positions of the plurality of points acquired by an external detection device, and the map generating device may further include a storage unit that stores the map information generated by the map information generating unit, and a map information correction unit that corrects the map information stored in the storage unit using the additional information. According to this configuration, the map information correction unit corrects the map information stored in the storage unit using the additional information regarding the positions of the plurality of points. This makes it possible to improve the accuracy of the map information stored in the storage unit.

[0015] (8) According to another aspect of the present invention, there is provided a map generating method for generating a map using a map generating device. This map generating method includes an input step of inputting natural language information expressing position information of a plurality of points in a natural language expression, and a conversion step of converting the natural language expression of the natural language information input in the input step into a map expression to generate map information expressing the position information in the map expression, the conversion step including a language information conversion step of converting the natural language expression of the natural language information into an intermediate expression to generate intermediate information expressing the position information in the intermediate expression, and a map information generating step of converting the intermediate expression of the intermediate information into a map expression to generate the map information. According to this configuration, the map generating device generates an intermediate expression expressed in the intermediate expression from the natural language information expressing the position information of a plurality of points in the natural language expression in the conversion step, and then generates map information expressed in the map expression from the intermediate information. This allows the position information of a plurality of points that is in the human brain and not specifically represented to be made into specific map information, thereby reducing the cost of generating a map.

[0016] (9) According to yet another aspect of the present invention, there is provided a computer program for causing a computer to generate a map. The computer program causes the computer to execute an input function for inputting natural language information expressing location information of a plurality of locations in a natural language expression, and a conversion function for converting the natural language expression of the natural language information input by the input function into a map expression to generate map information expressing the location information in the map expression, the conversion function having a language information conversion function for converting the natural language expression of the natural language information into an intermediate expression to generate intermediate information expressing the location information in the intermediate expression, and a map information generation function for converting the intermediate expression of the intermediate information into a map expression to generate the map information. According to this configuration, the map generation device generates an intermediate expression expressed in the intermediate expression from the natural language information expressing location information of a plurality of locations in a natural language expression in the conversion step, and then generates map information expressed in the map expression from the intermediate information. This allows the location information of a plurality of locations that is in a human brain and not specifically represented to become specific map information, thereby reducing the cost of generating a map.

[0017] The present invention can be realized in various forms, such as a map generated by a map generating device, a system including a map generating device, a method of operating these devices and systems, a navigation method using a map generating device, a server device for distributing computer programs, and a non-transitory storage medium on which the computer program is stored. [Brief description of the drawings]

[0018] [Figure 1] 1 is a schematic diagram showing a schematic configuration of a map generating device according to a first embodiment. [Diagram 2] 3 is a flowchart showing a map generating method according to the first embodiment. [Diagram 3] FIG. 2 is a diagram for explaining a map generating method according to the first embodiment. [Figure 4] FIG. 13 is a schematic diagram showing a schematic configuration of a map generating device according to a second embodiment. [Diagram 5] 10 is a flowchart showing a route generation method according to a second embodiment. [Figure 6] FIG. 13 is a schematic diagram showing a schematic configuration of a map generating device according to a third embodiment. [Figure 7] 13 is a flowchart showing a map correction method according to a third embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] First Embodiment FIG. 1 is a schematic diagram showing a schematic configuration of a map generating device 1 of the first embodiment. The map generating device 1 of the first embodiment generates a map by using an input natural language. The map generating device 1 is a computer including, for example, a read-only memory (ROM), a random access memory (RAM), a central processing unit (CPU), and the like, and includes a language information input unit 11, a calculation unit 20, and a storage unit 30. In this embodiment, the language information input unit 11, the calculation unit 20, and the storage unit 30 are located in different places, and the language information input unit 11, the calculation unit 20, and the storage unit 30 are connected via a network. Note that one device may include the language information input unit 11, the calculation unit 20, and the storage unit 30.

[0020] The language information input unit 11 receives natural language information that expresses location information of a plurality of locations in a natural language expression. Here, the term "natural language expression" refers to information expressed in a natural language that can be recognized by humans, such as speech between humans or text written in a document. The language information input unit 11 in this embodiment is, for example, a keyboard. Note that the language information input unit 11 is not limited to an input device such as a keyboard. The language information input unit 11 may be a microphone for inputting voice, a touch panel, a scanner for reading the contents written in a document, or an interface with an external recording device such as a USB memory or a flash memory in which location information is stored.

[0021] The calculation unit 20 is a CPU electrically connected to both the language information input unit 11 and the storage unit 30. The calculation unit 20 functions as a language information conversion unit 21 and a map information generation unit 22 by executing a computer program stored in the storage unit 30. The calculation unit 20 converts the natural language expression of the natural language information inputted at the language information input unit 10 into a map expression, thereby generating map information in which location information is expressed in a map expression. The calculation unit 20 corresponds to the "conversion unit" in the claims.

[0022] The language information conversion unit 21 converts the natural language expression of the natural language information into an intermediate representation, thereby generating intermediate information in which the location information is expressed in the intermediate representation. Here, the "intermediate representation" refers to natural language information expressed in a natural language expression being divided into a plurality of fixed phrases in order to facilitate conversion of the natural language expression into a map representation in the map information generation unit 22 described later. In this embodiment, the language information conversion unit 21 is a large-scale language model, and is constructed, for example, by a data set related to natural language expressions and deep learning technology. Details of the function of the language information conversion unit 21 and specific examples of intermediate information expressed in the intermediate representation of this embodiment will be described later.

[0023] The map information generating unit 22 generates map information by converting the intermediate representation of the intermediate information into a map representation. Here, the "map representation" refers to a data representation representing a map image, and is represented by, for example, a mathematical formula including symbols and descriptors. In this embodiment, the map information is represented by a mathematical formula including descriptors. Details of the function of the language information converting unit 21 and specific examples of map information represented by the map representation of this embodiment will be described later.

[0024] The storage unit 30 stores map information expressed by a map expression, i.e., a mathematical formula including a descriptor. In this embodiment, the storage unit 30 is a general term for a memory device, and various types of storage devices such as a ROM, a RAM, a solid-state drive (SSD), a hard disk, and a flash memory card can be used. In this embodiment, the storage unit 30 also stores a computer program for executing a map generation method described later.

[0025] 2 is a flowchart showing a map generation method in the map generation device 1. Next, the map generation method of this embodiment will be described in detail. The map generation method of this embodiment is started when a computer program stored in the storage unit 30 is loaded into the calculation unit 20 in response to input of natural language information expressed in a natural language expression by a user of the map generation device 1.

[0026] In the map generating method of this embodiment, first, natural language information expressing location information in a natural language expression is input (step S11). In step S11, a user of the map generating device 1 operates the language information input unit 11 to input, for example, natural language information expressing location information of a plurality of points stored by the user in a natural language expression to the map generating device 1.

[0027] FIG. 3 is a diagram for explaining a specific example of the map generation method of the present embodiment. In FIG. 3, as a specific example for explaining the position information of a plurality of points, a schematic diagram showing the positional relationship of a bedroom R1, a study R2, and a child's room R3 as a plurality of points is shown. In FIG. 3, two corners C1 and C2 are shown in a passage connecting the bedroom R1, the study R2, and the child's room R3. For convenience, the passage shown in FIG. 3 shows a passage A1 connecting the bedroom R1 and the corner C1, a passage A2 connecting the corner C1 and the study R2, a passage A3 connecting the corner C1 and the corner C2, and a passage A4 connecting the corner C2 and the child's room R3. When the user of the map generating device 1 has stored the positional relationships of a plurality of points as shown in Fig. 3, in step S11, natural language information expressed in natural language expressions such as "Leaving the bedroom and going straight down the corridor, there is the first corner. Turn right at the first corner and go straight ahead to reach the study" or "Leaving the bedroom and going straight down the corridor, there is the first corner. Going straight further past the first corner, there is the second corner. Turn left at the second corner and you will arrive at the child's room" is input. Hereinafter, for convenience, information on the route from the bedroom R1 to the study R2 shown in Fig. 3, i.e., the positional information between the bedroom R1 and the study R2, is referred to as first positional information, and information on the route from the bedroom R1 to the child's room R3, i.e., the positional information between the bedroom R1 and the child's room R3, is referred to as second positional information.

[0028] The map generating device 1 of this embodiment includes a keyboard as the language information input unit 11. In step S11, the user operates the keyboard to input, for example, the first location information (first natural language information) expressed in the above-mentioned natural language expression and the second location information (second natural language information) expressed in the natural language expression to the map generating device 1. If the language information input unit 11 is a microphone, the user of the map generating device 1 can input the above-mentioned first location information and second location information to the map generating device 1 by reading them into the microphone. In this case, the language information input unit 11 analyzes the input voice of the user and generates the first natural language information and the second natural language information. In addition, the user of the map generating device 1 may input the first location information and the second location information to the map generating device 1 by reading a document on which the first location information and the second location information are written with a scanner. In this case, the language information input unit 11 analyzes the contents written in the read document and generates the first natural language information and the second natural language information. The natural language information generated in the language information input unit 11 is output to a language information conversion unit 21 .

[0029] Next, the natural language expression of the natural language information is converted into an intermediate representation (step S12). In this embodiment, the language information conversion unit 21 divides the natural language information output by the language information input unit 11 into a plurality of fixed phrases.

[0030] Here, the intermediate representation of this embodiment will be described by taking the above-mentioned first natural language information and second natural language information as examples. Each of the first natural language information and second natural language information output by the language information input unit 11 includes information on movement actions from a specific place to another place and information on the space. In this embodiment, such natural language information is divided into multiple combinations of information on movement actions and information on the space. In this embodiment, it is divided into multiple combinations of information on one movement action and information on the space in which that one movement action is performed. This makes it easier for the map information generating unit 22, which will be described later, to generate map information expressed in a map representation.

[0031] When the first natural language information, "Leave the bedroom and go straight down the hallway, you will come to the first corner. Turn right at the first corner and go straight ahead to reach the study," is converted into an intermediate representation, it becomes as follows. "Leave the bedroom towards the first corner." "Follow the road from the bedroom to the first turning." "Turn right at the first corner." "At the first turning point, follow the road towards the study." "Reach the study." In this way, each of the fixed phrases expressed in the intermediate representation includes one movement action and information such as the location and direction in which the movement action is performed. For example, in "Leave the bedroom towards the first corner," the movement action is "Leave towards the first corner," and the location in which the movement action is performed is the "bedroom." Hereinafter, the information expressed in the intermediate representation of the first natural language information is referred to as first intermediate information.

[0032] Furthermore, when the second natural language information mentioned above, "If you leave the bedroom and go straight down the hallway, you will come to the first corner. If you go straight again after the first corner, you will come to the second corner. If you turn left at the second corner, you will arrive at the child's room," is converted into the intermediate representation as follows: "Leave the bedroom towards the first corner." "Follow the road from the bedroom to the first turning." "Follow the road from the first turn to the second turn." "Turn left at the second corner." "At the second turning point, follow the path towards the children's room." "Reach the children's room." In the second intermediate information that expresses the second natural language information in an intermediate representation, in “follow the road from the first corner to the second corner,” one movement action is “follow the road to the second corner,” and the location where one movement action is performed is “the first corner.” The generated intermediate information is output to the map information generating unit 22.

[0033] Next, the intermediate representation is converted into a map representation (step S13). In this embodiment, the map information generating unit 22 converts the intermediate representation of the intermediate information output by the language information converting unit 21 into a map representation. The map representation in this embodiment is represented by a mathematical expression using a descriptor, as shown below. Note that the "X" in parentheses of the descriptor indicates that the map representation is expressed by a mathematical expression using a descriptor. i ", "X j Each of the " indicates the name of a location, such as "study" or "children's room." a d (X i ,X j ):X j From X i Depart for a a (X i ,X j ):X j From X i Arrive at a s (X i ,X j ):X j From X i Continue along the road until a l (X i ):X i Turn left with a r (X i ):X i Turn right with a t : Turn around

[0034] In this embodiment, the following relational expressions (1) to (16) are defined for these descriptors. k " indicates the name of a location such as "study" or "children's room", and "A" and "B" in equation (16) are the above-mentioned descriptors a d , a a , a s , a l , a r , a t Either of the following is assigned. a s (X i ,X j ) -1=a s (X j ,X i ) ···(1) a a (X i ,X j ) -1 =a d (X j ,X i ) ···(2) a d (X i ,X j ) -1 =a a (X j ,X i ) ···(3) a l (X i ) -1 =a r (X i ) ···(4) a r (X i ) -1 =a l (X i ) ···(5) a l (X i )a t =a r (X i ) ···(6) a r (X i )a t =a l (X i ) ···(7) a d (X i ,X j )a a (X j ,X i )=a t ···(8) a s (X i ,X j )a t a s (X i ,X j )=a t ···(9) a r (X i )a l (Xi )=1 ···(10) a s (X i ,X j )a t a d (X k ,X j )=a s (X i ,X j )a d (X i ,X j ) ···(11) a s (X i ,X j )a r (X j )a d (X k ,X j )=a s (X i ,X j )a d (X i ,X j ) ···(12) a s (X i ,X j )a l (X j )a d (X k ,X j )=a s (X i ,X j )a d (X i ,X j ) ···(13) a a (X i ,X j )a r (X i )a d (X i ,X k )=a a (X i ,X k )a s (X i ,X k ) ···(14) aa (X i ,X j )a l (X i )a d (X i ,X k )=a a (X i ,X k )a s (X i ,X k ) (15) (AB) -1 =B -1 A -1 (16)

[0035] Regarding the first intermediate information mentioned above, bedroom X 1 , turn the corner with X 2 , X in the study 3 , the second turn is X 4 , children's room X 5 Then, the first map information generated by converting the intermediate representation into the map representation is expressed by the following formula (17). X 3 =a a (X 3 ,X 2 )a s (X 3 ,X 2 )a r (X 2 )a s (X 2 ,X 1 )a d (X 2 ,X 1 )X 1 (17) Moreover, the second intermediate information becomes the second map information expressed by the following formula (18) by converting the intermediate representation into the map representation in the map information generating unit 22. X 5 =a a (X 5 ,X 4 )a s (X 5 ,X 4 )al (X 4 )a s (X 2 ,X 1 )a d (X 2 ,X 1 )X 1 (18)

[0036] Next, it is determined whether or not the map information generated by the map information generating unit 22 is the same as the map information already stored in the storage unit 30 (hereinafter simply referred to as "stored map information") (step S14). In this embodiment, it is determined whether or not the map information generated by the map information generating unit 22 is the same as the stored map information. If the map information generating unit 22 determines that the newly generated map information is not the same as the stored map information (step S14: Yes), the map information generating unit 22 adds the newly generated map information to the storage unit 30 and stores it (step S15). If the map information generating unit 22 determines that the newly generated map information is not the same as the stored map information (step S14: No), the map information generating unit 22 discards the generated map information (step S16).

[0037] According to the map generating device 1 of the present embodiment described above, the calculation unit 20 generates intermediate information expressed in an intermediate expression from natural language information that expresses the position information of multiple points in a natural language expression, and then generates map information expressed in a map expression from the intermediate information. This makes it possible to make the position information of multiple points that is in the human brain and not specifically expressed into concrete map information. Therefore, for example, compared to the case where information is obtained by sensing with a sensor to generate a map, the cost of generating map information including the position information of multiple points can be reduced, and therefore the cost of generating a map can be reduced.

[0038] Furthermore, according to the map generating device 1 of this embodiment, the language information conversion unit 21 generates intermediate information by dividing information about movement actions and spaces between two of the multiple locations into multiple combinations of movement actions and spaces. This makes it easier for the map information generating unit that generates map information from the intermediate information to generate map information. Therefore, the cost of generating a map can be further reduced.

[0039] Furthermore, according to the map generating device 1 of this embodiment, when the map information generating unit 22 determines that the newly generated map information is not the same as the stored map information, the newly generated map information is added and stored in the storage unit 30. This allows map information to be accumulated in the storage unit 30, thereby expanding the range covered by the stored map information.

[0040] According to the map generating method of the present embodiment, the map generating device 1 generates an intermediate representation from natural language information that expresses position information of a plurality of points in a natural language expression, and then generates map information expressed in a map representation from the intermediate information in steps S12 and S13. This makes it possible to make the position information of a plurality of points that is in the human brain and not specifically expressed into specific map information, thereby reducing the cost of generating a map.

[0041] Furthermore, according to the computer program of this embodiment, the calculation unit 20 can convert the position information of multiple points in the human brain that are not specifically represented into specific map information, thereby reducing the cost of generating a map.

[0042] <Second embodiment> 4 is a schematic diagram showing a schematic configuration of a map generating device 2 of the second embodiment. The map generating device 2 of the second embodiment differs from the map generating device 1 (FIG. 1) of the first embodiment in that the map generating device 2 of the second embodiment generates a travel route between two points in response to an input.

[0043] The map generating device 2 of the second embodiment includes a language information input unit 12, a language information output unit 13, a calculation unit 40, and a storage unit 30. The calculation unit 40 includes a language information conversion unit 41, a map information generation unit 42, and a route information generation unit 43. Like the map generating device 1 of the first embodiment, the map generating device 2 generates map information by receiving natural language information that expresses position information of multiple points in a natural language expression. The map generating device 2 further generates route information regarding a travel route from the departure point to the destination by receiving information on each of a departure point and a destination in a movement. The calculation unit 40 corresponds to a "conversion unit" in the claims.

[0044] The language information input unit 12 receives input of natural language information expressing position information of a plurality of points in a natural language expression, similar to the language information input unit 11 of the first embodiment. The language information input unit 12 is also used to input information relating to each of the departure point and destination of a travel route, for example, when a user of the map generating device 2 searches for a travel route.

[0045] The language information output unit 13 is, for example, a display device for displaying characters and images, such as a display. The language information output unit 13 outputs natural language information expressed in a natural language expression, which is generated from intermediate information expressed in an intermediate expression by the language information conversion unit 41. Details of the contents output by the language information output unit 13 will be described later. The output of the language information output unit 13 may be a written output by a printing machine such as a printer, or may be an audio output by a speaker.

[0046] The calculation unit 40 is a CPU electrically connected to each of the language information input unit 12, the language information output unit 13, and the storage unit 30. The calculation unit 40 functions as a language information conversion unit 41, a map information generation unit 42, and a route information generation unit 43 by executing a computer program stored in the storage unit 30.

[0047] The language information conversion unit 41 is connected to the language information input unit 12, the language information output unit 13, the map information generation unit 42, and the route information generation unit 43. The language information conversion unit 41 is a large-scale language model, similar to the language information conversion unit 21 of the first embodiment, and converts the natural language expression of the natural language information into an intermediate representation, and generates intermediate information in which location information is expressed in the intermediate representation. The language information conversion unit 41 extracts the name of the departure place and the name of the destination from the information on the departure place and the destination input by the language information input unit 12. The language information conversion unit 41 can also generate natural language information expressed in a natural language expression from the intermediate information expressed in the intermediate representation. The function of the language information conversion unit 41 will be described in detail later.

[0048] The map information generation unit 42 is connected to the storage unit 30, the language information conversion unit 41, and the route information generation unit 43. Like the map information generation unit 22 in the first embodiment, the map information generation unit 42 converts the intermediate representation of the intermediate information into a map representation to generate map information. The map information generation unit 42 is also capable of generating intermediate information from stored map information. The functions of the map information generation unit 42 will be described in detail later.

[0049] The route information generating unit 43 is connected to the storage unit 30, the language information converting unit 41, and the map information generating unit 42. The route information generating unit 43 generates a route between at least two of the multiple locations using the stored map information stored in the storage unit 30. The route information generating unit 43 determines whether or not it is possible to generate a travel route from a departure point to a destination input by the language information input unit 12. The function of the route information generating unit 43 will be described in detail later.

[0050] Fig. 5 is a flowchart showing a route generation method in the map generation device 2. Next, the details of the route generation method of this embodiment will be described. The route generation method of this embodiment is started when a computer program stored in the storage unit 30 is loaded into the calculation unit 40 in response to input of information expressed in a natural language expression by a user of the map generation device 2. Note that the generation of map information in the map generation device 2 is performed in the same manner as in the map generation device 1 of the first embodiment.

[0051] In the route generation method of this embodiment, first, information on the departure point and destination is input (step S21). In step S21, the user of the map generation device 2 operates the language information input unit 12 to input, for example, information on the departure point and destination when the user moves to the map generation device 2 in natural language expressions. In this embodiment, for example, in the positional relationship between the bedroom R1, study R2, and child's room R3 shown in Fig. 3, information is input such as "What is the route to get from the child's room to the study?"

[0052] Next, the starting point and destination are extracted from the input information (step S22). In step S22, the language information conversion unit 41 extracts the name of the starting point and the name of the destination from the information on the starting point and the destination expressed in natural language expressions input by the language information input unit 12. For example, in the above-mentioned question "What is the route from the child's room to the study?", it is extracted that the name of the starting point is "child's room" and the name of the destination is "study". The language information conversion unit 41 outputs the extracted name of the starting point and the name of the destination to the route information generation unit 43.

[0053] Next, it is determined whether or not it is possible to generate a travel route from the departure point to the destination (step S23). In step S23, the route information generating unit 43 uses the stored map information stored in the storage unit 30 to determine whether or not it is possible to generate a travel route from the departure point to the destination. For example, if the stored map information includes information on the departure point and the destination, and the departure point and the destination are continuously connected in the stored map information, it is possible to generate a travel route from the departure point to the destination. If the route information generating unit 43 determines that it is possible to generate a travel route (step S23: Yes), it proceeds to step S24. On the other hand, if the stored map information does not include information on the departure point or the destination, or if the departure point and the destination are not continuously connected in the stored map information, it is not possible to generate a travel route from the departure point to the destination. If the route information generating unit 43 determines that it is not possible to generate a travel route (step S23: No), it proceeds to step S28.

[0054] In step S23, if the route information generating unit 43 determines that generation of a travel route is possible, the route information generating unit 43 generates route information from the departure point to the destination (step S24). In step S24, the route information generating unit 43 uses the stored map information stored in the storage unit 30 to generate information on the travel route from the departure point to the destination as route information expressed in a map representation.

[0055] Here, a method of generating a route in step S24 will be described. When the stored map information stored in the storage unit 30 contains information about the travel route from the departure point to the destination, the travel route is generated based on the map information. Even when the stored map information does not contain information about the travel route from the departure point to the destination, the map generating device 2 can generate a travel route from the departure point to the destination by combining the stored map information.

[0056] For example, in the positional relationship of the bedroom R1, the study R2, and the children's room R3 shown in FIG. 3, 5 ) to the study (X 3 ) is first calculated using equation (18) from the child's room (X 5 ) to the bedroom (X 1 ) is expressed by the following equation (19). X 1 ={a a (X 5 ,X 4 )a s (X 5 ,X 4 )a l (X 4 )a s (X 4 ,X 2 )a s (X 2 ,X 1 )a d (X 2 ,X 1 )} -1 X 5 =a a (X1 ,X 2 )a s (X 1 ,X 2 )a s (X 2 ,X 4 )a r (X 4 )a s (X 4 ,X 5 )a d (X 4 ,X 5 )X 5 (19) X in Eq. (17) 1 By substituting equation (19) into equation (20), the following equation (20) is generated. X 3 =a a (X 3 ,X 2 )a s (X 3 ,X 2 )a r (X 2 )a s (X 2 ,X 1 )a d (X 2 ,X 1 )a a (X 1 ,X 2 )a s (X 1 ,X 2 )a a (X 2 ,X 4 )a r (X 4 )a s (X 4 ,X 5 )a d (X 4 ,X 5 )X 5 =a a (X 3 ,X 2 )a s (X 3 ,X 2 )a l (X 2 )as (X 2 ,X 4 )a r (X 4 )a s (X 4 ,X 5 )a d (X 4 ,X 5 )X 5 (20) In this way, the map generating device 2 can generate a travel route from a departure point to a destination by combining the stored map information. Furthermore, the map generating device 2 can generate the shortest travel route by using information on the distance between a plurality of points included in the stored map information. The route information generating unit 43 outputs the route information represented by the generated map representation to the map information generating unit 42.

[0057] Next, the map representation of the route information is converted into an intermediate representation (step S25). In step S25, the map information generation unit 42 converts the map representation of the route information generated by the route information generation unit 43 into an intermediate representation, and generates route information represented in the intermediate representation. The map information generation unit 42 outputs the generated route information represented in the intermediate representation to the language information conversion unit 41.

[0058] Next, the intermediate representation of the route information is converted into a natural language representation (step S26). In step S26, the language information conversion unit 41 converts the intermediate representation of the route information generated by the map information generation unit 42 into a natural language representation, and generates route information expressed in a natural language representation.

[0059] Next, the route information expressed in a natural language expression is output (step S27). In step S27, the language information output unit 13 outputs the route information expressed in a natural language expression generated by the language information conversion unit 41. This allows the user of the map generating device 2 to know the travel route from the departure point to the destination as an output in response to the input of information on the departure point and the destination. In addition, the route information expressed in a natural language expression output by the language information output unit 13 can be input to a moving body such as a robot and used for navigation of the moving body.

[0060] In step S23, if the route information generating unit 43 determines that the generation of the travel route is impossible, it outputs a notification prompting the input of information necessary for generating the route information (step S28). In step S28, the route information generating unit 43 outputs the content of the input information necessary for generating the travel route from the departure point to the destination to the language information converting unit 41. The information necessary for generating the travel route includes, for example, the position information of the departure point or destination that is not included in the stored map information stored in the storage unit 30, and information of a discontinuous part when the departure point and the destination are not continuously connected even in the stored map information. The language information converting unit 41 to which the content of the information necessary for generating the travel route is input converts the content of the information necessary for generating the travel route into a natural language expression. The language information output unit 13 outputs the content of the information necessary for generating the travel route expressed in the natural language expression to the outside, and notifies the user of the map generating device 2. As a result, the user inputs the information necessary for generating the travel route again, and the stored map information can be used to generate the travel route from the departure point to the destination.

[0061] According to the map generating device 2 of the present embodiment described above, the calculation unit 40 generates intermediate information expressed in an intermediate expression from natural language information that expresses the position information of a plurality of points in a natural language expression, and then generates map information expressed in a map expression from the intermediate information. This makes it possible to make the position information of a plurality of points that is in the human brain and not specifically expressed into concrete map information. Therefore, it is possible to reduce the cost of generating a map.

[0062] Furthermore, according to the map generating device 2 of this embodiment, the language information output unit 13 can output the memorized map information stored in the storage unit 30 as natural language information expressed in a natural language expression that can be understood by humans. This allows the memorized map information to be effectively used in fields where it is used by humans to understand, such as route guidance for pedestrians and vehicle drivers.

[0063] Furthermore, according to the map generating device 2 of this embodiment, the route information generating unit 43 generates a route between at least two of the multiple locations, using the stored map information stored in the storage unit 30. This makes it possible to use the stored map information for generating a travel route.

[0064] Furthermore, according to the map generating device 2 of this embodiment, the route information generating unit 43 uses the input information on the departure point and the destination to determine whether or not a travel route from the departure point to the destination can be generated. When the route information generating unit 43 determines that generation of the travel route is impossible, it causes the language information output unit 13 to output a notification prompting input of natural language information necessary for generating the travel route. As a result, even if the map generating device 2 cannot generate a travel route using the map information stored in the storage unit 30, the map generating device 2 can generate a travel route by inputting the natural language information necessary for generating the travel route.

[0065] <Third embodiment> 6 is a schematic diagram showing a schematic configuration of a map generating device 3 of the third embodiment. The map generating device 3 of the third embodiment is different from the map generating device 1 (FIG. 1) of the first embodiment in that the map generating device 3 corrects map information using environmental information acquired by an external detection device.

[0066] The map generating device 3 of the third embodiment includes a language information input unit 11, an environmental information input unit 14, a calculation unit 50, and a storage unit 30. The calculation unit 50 includes a language information conversion unit 21, a map information generating unit 22, a spatial information grasping unit 51, and a map information correction unit 52. The map generating device 3 generates map information by receiving natural language information that expresses position information of a plurality of points in a natural language expression, as in the map generating device 1 of the first embodiment. The map generating device 3 further corrects the map information stored in the storage unit 30 by using a detection result by an external detection device. The calculation unit 50 corresponds to a "conversion unit" in the claims.

[0067] The environmental information input unit 14 is electrically connected to an external detection device. The environmental information input unit 14 receives environmental information acquired by sensing a sensor included in the external detection device as additional information. The sensor included in the external detection device is a camera or LiDAR (Light Detection And Ranging), and the environmental information is input as the detection result by the sensor. In this embodiment, the environmental information is information including the position, distance, shape, size, etc. of an object around the moving body detected by the sensor when the moving body, which is an external detection device equipped with a sensor, moves within a certain range including at least one point among a plurality of points in the stored map information.

[0068] The spatial information grasping unit 51 generates map information expressed in a map format, using the environmental information input by the environmental information input unit 14. In this embodiment, the spatial information grasping unit 51 generates map information related to the surrounding environment in which the mobile object has moved.

[0069] The map information correction unit 52 compares the map information generated by the spatial information understanding unit 51 with the stored map information. By comparing the two pieces of map information, the map information correction unit 52 determines whether or not the stored map information needs to be corrected or added.

[0070] Fig. 7 is a flowchart showing a map correction method in the map generating device 3. Next, the map correction method of this embodiment will be described in detail. The map correction method of this embodiment is started when detection information from an external detection device is input to the map generating device 3, and a computer program stored in the storage unit 30 is loaded into the calculation unit 50. Note that the generation of map information in the map generating device 3 is performed in the same manner as in the map generating device 1 of the first embodiment.

[0071] In the map correction method of this embodiment, first, environmental information is input from an external detection device (step S31). In step S31, the environmental information is input from the external detection device via the environmental information input unit 14. The environmental information input unit 14 outputs the input environmental information to the spatial information grasping unit 51.

[0072] Next, map information expressed in a map representation is generated from the environmental information (step S32). In step S32, the spatial information grasping unit 51 uses the environmental information to recognize the movement path of a moving body as an external detection device, and generates map information. The map information generated by the spatial information grasping unit 51 is the same as the map information generated by the map information generating unit 22, and is expressed in a map representation. The map information generated by the spatial information grasping unit 51 is output to the map information correcting unit 52. Note that the method of generating map information in the spatial information grasping unit 51 is not limited to this.

[0073] Next, it is determined whether or not the map information needs to be corrected (step S33). In step S33, the map information correction unit 52 extracts map information corresponding to the map information generated by the spatial information understanding unit 51 from the memory map information stored in the memory unit 30. The map information correction unit 52 compares the map information generated by the spatial information understanding unit 51 with the extracted memory map information, and determines whether or not the memory map information needs to be corrected. If the map information correction unit 52 determines that the map information needs to be corrected (step S33: Yes), the process proceeds to step S34. If the map information correction unit 52 determines that the map information does not need to be corrected (step S33: No), the process proceeds to step S36. Note that if the map information correction unit 52 cannot extract map information corresponding to the map information generated from the environmental information from the memory unit 30, the map information generated from the environmental information is not stored in the memory unit 30, and the map information correction unit 52 determines that the map information needs to be corrected.

[0074] If the map information correction unit 52 determines in step S33 that the map information needs to be corrected, it corrects the map information based on the environmental information (step S34). In step S34, the map information correction unit 52 corrects the part of the stored map information stored in the storage unit 30 that is determined to need correction. Next, in step S34, the corrected map information is stored (step S35). If the map information correction unit 52 cannot extract map information corresponding to the map information generated from the environmental information from the storage unit 30 in step S33, the map information generated from the environmental information is stored in the storage unit 30 in step S35. If the map information correction unit 52 determines that the map information does not need to be corrected (step S33: No), the map information correction unit 52 discards the map information generated from the environmental information (step S36).

[0075] According to the map generating device 3 of this embodiment described above, the calculation unit 50 can output the memorized map information stored in the storage unit 30 as natural language information expressed in a natural language expression that can be understood by humans. This allows the memorized map information to be effectively used in fields where it is used by humans to understand, such as route guidance for pedestrians and vehicle drivers.

[0076] Furthermore, according to the map generating device 3 of this embodiment, the map information correction unit 52 corrects the stored map information stored in the storage unit 30 by using additional information related to the positions of a plurality of points. This makes it possible to improve the accuracy of the stored map information stored in the storage unit 30.

[0077] <Modifications of this embodiment> The present invention is not limited to the above-described embodiment, and can be embodied in various forms without departing from the spirit and scope of the invention. For example, the following modifications are also possible.

[0078] [Variation 1] In the above embodiment, the natural language information includes information about a movement action and a space from a first point to a second point among the multiple points. However, the information included in the natural language information is not limited to this. The information included in the natural language information does not include information about the movement action, and may be information about the positional relationship of the multiple points.

[0079] [Variation 2] In the above embodiment, the intermediate representation refers to natural language information expressed in a natural language expression being divided into a plurality of fixed phrases. However, the intermediate representation is not limited to this. When natural language information is input, it is preferable that the intermediate representation be a representation that makes it easy to convert the natural language expression into a map representation in a map information generation unit located after the language information conversion unit.

[0080] [Variation 3] In the second embodiment, the language information output unit 13 of the map generating device 2 outputs to the user a travel route from a departure point to a destination in a natural language expression. However, the method of outputting the travel route by the map generating device 2 is not limited to this. For example, when outputting to a moving body such as a robot, the travel route may be output in an expression that is convenient for control input for the moving body.

[0081] [Variation 4] In the above embodiment, the map representation is represented by a mathematical formula including a descriptor, and the storage unit stores map information represented by a mathematical formula including a descriptor. However, the map representation is not limited to this. The map representation may be represented by a mathematical formula including symbols indicating a position or a direction.

[0082] [Variation 5] In the above embodiment, the storage unit 30 stores map information expressed by a mathematical formula including a descriptor representing a position or direction as a map expression, and as described in the second embodiment, the language information output unit 13 outputs the map information in a natural language expression. However, the method of outputting the stored map information is not limited to this. For example, the map generating device may further include a map generating unit that generates a map using map information expressed by a mathematical formula stored in the storage unit, and a map output unit that outputs the map generated by the map generating unit. In addition, when outputting the map information to a moving body such as a robot, the map information expressed by the map expression may be output as it is.

[0083] Although the present aspect has been described above based on the embodiment and modified examples, the above-mentioned embodiment of the aspect is intended to facilitate understanding of the present aspect and does not limit the present aspect. The present aspect may be modified or improved without departing from the spirit and scope of the claims, and equivalents are included in the present aspect. Furthermore, if a technical feature is not described as essential in this specification, it may be deleted as appropriate.

[0084] <Application example 1> A map generating device, comprising: an input unit to which natural language information expressing location information of a plurality of points in a natural language expression is input; a conversion unit that converts a natural language expression of the natural language information inputted to the input unit into a map expression to generate map information in which the location information is expressed in a map expression, The conversion unit is a language information conversion unit that converts a natural language expression of the natural language information into an intermediate representation to generate intermediate information in which the location information is expressed in the intermediate representation; a map information generating unit that generates the map information by converting an intermediate representation of the intermediate information into a map representation. Map generator. <Application example 2> The map generating device according to Application Example 1, the natural language information includes information about a movement action and a space from a first point to a second point among the plurality of points; the language information conversion unit generates the intermediate information by dividing the natural language information into a plurality of combinations of information related to movement actions and information related to space. Map generator. <Application example 3> The map generating device according to the first or second aspect of the present invention further comprises: a storage unit that stores the map information generated by the map information generation unit, The map information generating unit determining whether the newly generated map information is the same as the map information stored in the storage unit; when it is determined that the generated map information is not the same as the map information stored in the storage unit, the generated map information is added to and stored in the storage unit; Map generator. <Application Example 4> The map generating device according to any one of Application Examples 1 to 3 further comprises: a storage unit that stores the map information generated by the map information generation unit, the map information generating unit generates intermediate information from the map information stored in the storage unit; the language information conversion unit generates natural language information from the intermediate information generated by the map information generation unit, The map generating device further comprises: an output unit that outputs the natural language information generated by the language information conversion unit to an outside; Map generator. <Application example 5> The map generating device according to any one of Application Examples 1 to 4 further comprises: a route information generating unit that generates a route between at least two of the plurality of points by using map information stored in the storage unit; Map generator. <Application Example 6> The map generating device according to any one of Application Examples 1 to 5, The input unit is capable of inputting information regarding a departure point and information regarding a destination, The route information generation unit determining whether or not a travel route from the departure point to the destination can be generated using map information stored in the storage unit; when it is determined that generation of the travel route is impossible, outputting a notification from the output unit to prompt input of information necessary for generation of the travel route; Map generator. <Application Example 7> The map generating device according to any one of Application Examples 1 to 6, the input unit is capable of inputting additional information regarding the positions of the plurality of points acquired by an external detection device, The map generating device further comprises: a storage unit that stores the map information generated by the map information generating unit; a map information correction unit that corrects the map information stored in the storage unit by using the additional information, Map generator. <Application Example 8> A map generating method for generating a map using a map generating device, comprising: an input step of inputting natural language information expressing location information of a plurality of points in a natural language expression; a conversion step of converting the natural language expression of the natural language information input in the input step into a map expression to generate map information in which the location information is expressed in a map expression, The converting step comprises: a language information conversion step of converting a natural language expression of the natural language information into an intermediate representation to generate intermediate information in which the location information is expressed in the intermediate representation; a map information generating step of generating the map information by converting the intermediate representation of the intermediate information into a map representation. Map generation method. <Application Example 9> A computer program for causing a computer to generate a map, an input function for inputting natural language information expressing location information of a plurality of points in a natural language expression; a conversion function of converting a natural language expression of the natural language information input by the input function into a map expression, thereby generating map information in which the location information is expressed in a map expression; The conversion function is a language information conversion function that converts a natural language expression of the natural language information into an intermediate representation to generate intermediate information in which the location information is expressed in the intermediate representation; a map information generating function for generating the map information by converting the intermediate representation of the intermediate information into a map representation; Computer program. [Explanation of symbols]

[0085] 1,2,3...Map Generator 11, 12…Language information input section 13...Language information output section 20,40,50...calculation section 21,41…Language information conversion section 22, 42…Map information generation section 30...Storage section 43...Route generation unit 52…Map information correction section

Claims

1. A map generating device, comprising: an input unit to which natural language information expressing location information of a plurality of points in a natural language expression is input; a conversion unit that converts a natural language expression of the natural language information inputted to the input unit into a map expression to generate map information in which the location information is expressed in a map expression, The conversion unit is a language information conversion unit that converts a natural language expression of the natural language information into an intermediate representation to generate intermediate information in which the location information is expressed in the intermediate representation; a map information generating unit that generates the map information by converting an intermediate representation of the intermediate information into a map representation. Map generator.

2. 2. A map generating device according to claim 1, The natural language information includes information regarding a movement action and a space from a first point to a second point among the plurality of points, the language information conversion unit generates the intermediate information by dividing the natural language information into a plurality of combinations of information related to movement actions and information related to space. Map generator.

3. The map generating device according to claim 1 or 2 further comprises: a storage unit that stores the map information generated by the map information generation unit, The map information generating unit determining whether the newly generated map information is the same as the map information stored in the storage unit; when it is determined that the generated map information is not the same as the map information stored in the storage unit, the generated map information is added to and stored in the storage unit; Map generator.

4. The map generating device according to claim 1 or 2 further comprises: a storage unit that stores the map information generated by the map information generation unit, the map information generating unit generates intermediate information from the map information stored in the storage unit; the language information conversion unit generates natural language information from the intermediate information generated by the map information generation unit, The map generating device further comprises: an output unit that outputs the natural language information generated by the language information conversion unit to an outside; Map generator.

5. The map generating device according to claim 4 further comprises: a route information generating unit that generates a route between at least two of the plurality of points by using map information stored in the storage unit; Map generator.

6. 6. A map generating device according to claim 5, The input unit is capable of inputting information regarding a departure point and information regarding a destination, The route information generation unit determining whether or not a travel route from the departure point to the destination can be generated using map information stored in the storage unit; when it is determined that generation of the travel route is impossible, outputting a notification from the output unit to prompt input of information necessary for generation of the travel route; Map generator.

7. 3. A map generating device according to claim 1 or 2, the input unit is capable of inputting additional information regarding the positions of the plurality of points acquired by an external detection device, The map generating device further comprises: a storage unit that stores the map information generated by the map information generating unit; a map information correction unit that corrects the map information stored in the storage unit by using the additional information, Map generator.

8. A map generating method for generating a map using a map generating device, comprising: an input step of inputting natural language information expressing location information of a plurality of points in a natural language expression; a conversion step of converting the natural language expression of the natural language information input in the input step into a map expression to generate map information in which the location information is expressed in a map expression, The converting step comprises: a language information conversion step of converting a natural language expression of the natural language information into an intermediate representation to generate intermediate information in which the location information is expressed in the intermediate representation; a map information generating step of generating the map information by converting the intermediate representation of the intermediate information into a map representation. Map generation method.

9. A computer program for causing a computer to generate a map, an input function for inputting natural language information expressing location information of a plurality of points in a natural language expression; a conversion function of converting a natural language expression of the natural language information input by the input function into a map expression, thereby generating map information in which the location information is expressed in a map expression; The conversion function is a language information conversion function that converts a natural language expression of the natural language information into an intermediate representation to generate intermediate information in which the location information is expressed in the intermediate representation; a map information generating function for generating the map information by converting the intermediate representation of the intermediate information into a map representation; Computer program.

Images