KR102308261B1 - Method for providing chatbot service based on task transfer of chatbot and device for performing the method - Google Patents

Abstract

The present invention relates to a chatbot service providing method based on business transfer between chatbots and an apparatus for performing the method. The service providing method based on business transfer between chatbots comprises the following steps of: receiving, by a chatbot service apparatus, target sub-conversation information from a user apparatus; determining, by the chatbot service apparatus, a response chatbot based on the target sub-conversation information; and transmitting, by the chatbot service apparatus, a chatbot message to the user apparatus in response to the target sub-conversation information via the response chatbot.

Description

Method for providing chatbot service based on task transfer of chatbot and device for performing the method

The present invention relates to a method of providing a chatbot service based on business transfer between chatbots, and an apparatus for performing such a method.

The existing bot operation method is a method in which one bot focuses on providing a specific service or a single bot provides all various services. Technically, it is possible to provide an all-in-one service with a single chatbot by building a knowledge base in all fields, but from the user's point of view, it may feel mechanical, so it may not be sympathetic and may rather resist the all-round chatbot. The goal of chatbots should be to form intimacy with people, to relieve the burden of work as a virtual assistant, and to be able to perform a role that can help in various fields in real time.

Therefore, it is necessary to study how to provide services through interaction between chatbots when a user inquires about a task that exceeds their original role by assigning an exclusive role and task to the chatbot.

An object of the present invention is to solve all of the above problems.

Another object of the present invention is to provide individual services to users based on the plurality of chatbots based on mutual business transfer between the plurality of chatbots.

Another object of the present invention is to actively deliver necessary information to a user based on a chatbot in consideration of the user's location and the user's current situation.

A representative configuration of the present invention for achieving the above object is as follows.

According to an embodiment of the present invention, in a service providing method based on business transfer between chatbots, the chatbot service device receives target sub-conversation information from a user device, and the chatbot service device responds based on the target sub-conversation information determining a chatbot; and sending, by the chatbot service device, a chatbot message to the user device in response to the target sub-conversation information through the response chatbot.

On the other hand, the service providing method based on business transfer between chatbots may further include, by the chatbot service device, performing a chatbot business transfer from the current chatbot to the responding chatbot when the current chatbot is not the responsive chatbot. .

In addition, the determining of the response chatbot may include: determining, by the chatbot service device, a primary chatbot relevance probability based on word information and context information of the target sub-conversation information; determining a chatbot relevance probability based on a primary chatbot relevance probability; and determining, by the chatbot service device, the response chatbot based on the chatbot relevance probability.

According to an embodiment of the present invention, a chatbot service apparatus providing a chatbot service may include a communication unit configured to communicate with a user device and a processor operatively connected to the communication unit, wherein the processor is configured to receive a target from the user device. Receive sub-conversation information, determine a response chatbot based on the target sub-conversation information, and transmit a chatbot message to the user device in response to the target sub-conversation information through the response chatbot.

Meanwhile, when the current chatbot is not the response chatbot, the processor may be implemented to transfer the chatbot business from the current chatbot to the response chatbot.

In addition, the processor determines a primary chatbot relevance probability based on word information and context information of the target sub-conversation information, and determines a chatbot relevance probability based on previous sub-conversation information and the primary chatbot relevance probability, and It may be implemented to determine the responsive chatbot based on a chatbot relevance probability.

According to the present invention, it is possible to provide individual services to users based on the plurality of chatbots based on mutual business transfer between the plurality of chatbots.

In addition, according to the present invention, necessary information can be actively delivered to the user based on the chatbot in consideration of the user's location and the user's current situation.

1 is a conceptual diagram illustrating a chatbot service apparatus according to an embodiment of the present invention.
2 is a conceptual diagram illustrating an operation of a chatbot determining unit according to an embodiment of the present invention.
3 is a conceptual diagram illustrating an operation of a chatbot determining unit according to an embodiment of the present invention.
4 is a conceptual diagram illustrating an operation of a chatbot decision unit according to an embodiment of the present invention.
5 is a conceptual diagram illustrating an operation of a chatbot task transfer unit according to an embodiment of the present invention.
6 is a conceptual diagram illustrating an operation of a chatbot decision unit according to an embodiment of the present invention.
7 is a conceptual diagram illustrating a chatbot service according to an embodiment of the present invention.
8 is a conceptual diagram illustrating a chatbot service according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0023] Reference is made to the accompanying drawings, which show by way of illustration specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail 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, certain shapes, structures, and characteristics described herein may be implemented with changes from one embodiment to another without departing from the spirit and scope of the present invention. In addition, it should be understood that the location or arrangement of individual components within each embodiment may be changed without departing from the spirit and scope of the present invention. Accordingly, the detailed description given below is not to be taken in a limiting sense, and the scope of the present invention should be taken to cover the scope of the claims and all equivalents thereto. In the drawings, like reference numerals refer to the same or similar elements throughout the various aspects.

Hereinafter, various preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings in order to enable those of ordinary skill in the art to easily practice the present invention.

Existing chatbot service is a method in which one chatbot focuses on providing a specific service or provides all various services through one bot. Technically, these chatbot services can feel mechanical from the user's point of view, making the chatbot service unfamiliar. The goal of chatbots should be to form intimacy with people, to relieve the burden of work as a virtual assistant, and to be able to perform a role that can help in various fields in real time. A service providing method based on business transfer between chatbots and a device for performing such a method provide a chatbot service, wherein a plurality of chatbots corresponding to each of a plurality of business areas are set, conversation contents with a user, and a user's location And/or a conversation between the chatbot and the user suitable for the work area may be conducted based on the user's situation.

As the chatbot service is provided through multiple chatbots, existing conversation information and information necessary for future efficient consultation are automatically transferred between the multiple chatbots, rather than simply transferring work through the chatbots, resulting in a smooth hand-off between the multiple chatbots. off) by providing a chatbot service based on multiple chatbots, an improved customer experience than the existing chatbot service can be provided to users.

In the method for providing a service based on business transfer between chatbots according to an embodiment of the present invention and a device for performing such a method, assuming that the chatbot service is a service for medical and tourism purposes, it is possible to provide local living convenience for medical tourists and at the same time It has a direct effect of revitalizing the local economy by recommending more diverse local services based on location-based information.

Hereinafter, in an embodiment of the present invention, a chatbot for medical treatment and a chatbot for tourism are disclosed as examples of business transfer between chatbots. However, the service providing method based on the business transfer between chatbots according to an embodiment of the present invention and the business transfer method between chatbots disclosed in the apparatus for performing such a method can be used in various chatbot service areas as well as medical and tourism. , these embodiments may also be included in the scope of the present invention.

1 is a conceptual diagram illustrating a chatbot service apparatus according to an embodiment of the present invention.

In FIG. 1, a chatbot service device that enables business transfer between a plurality of chatbots is disclosed.

Referring to FIG. 1 , the chatbot service apparatus may include a communication unit 100 , a chatbot determination unit 120 , a chatbot task transfer unit 140 , a chatbot service unit 160 , and a processor (not shown).

The communication unit 100 may be implemented to communicate with a user device that is a target of providing the chatbot service.

The chatbot determining unit 120 may be implemented to collect conversation information between the user and the chatbot based on communication with the user device, and to determine a chatbot related to the conversation information based on the conversation information. For example, if the chatbot service is a medical service and a tourism service, a determination may be made based on the conversation information as to whether the first chatbot for the medical service or the second chatbot for the tourism service will respond. . A specific chatbot task determination method of the chatbot determination unit 120 will be described later.

The chatbot task transfer unit 140 may be implemented to transfer tasks between chatbots that will conduct a conversation with a user based on the determination of the chatbot decision unit 120 . The chatbot task transfer unit 140 may perform a task transfer procedure to a chatbot that will finally perform an answer based on the determination of the chatbot decision unit 120 . The chatbot task transfer unit 140 may proceed with the chatbot task transfer procedure when a change in the chatbot currently communicating with the user is required. The chatbot business transfer procedure may be a procedure for converting a chatbot that has been talking with a user to another chatbot. This will be described in detail later.

The chatbot service unit 160 may be implemented to learn data for the operation of the chatbot. For example, the chatbot service unit 160 may perform learning for the operation of the chatbot based on the conversation information. Specifically, the chatbot service unit 160 may perform machine learning-based learning on word information and context information for determining the chatbot in the chatbot decision unit, and may determine the exclusivity of the work area between the chatbots. . That is, the chatbot service unit 160 may perform learning for the operations of the chatbot determination unit 120 and the chatbot task transfer unit 140 .

The processor may be implemented to control operations of the communication unit 100 , the chatbot determination unit 120 , the chatbot task transfer unit 140 , and the chatbot service unit 160 .

2 is a conceptual diagram illustrating an operation of a chatbot determining unit according to an embodiment of the present invention.

2 discloses a chatbot task determination method of a chatbot decision unit.

Referring to FIG. 2 , the chatbot determiner may determine the chatbot related to the conversation information based on conversation information between the user and the chatbot.

The chatbot determining unit may determine the task of the chatbot to respond to the user (hereinafter, the response chatbot) based on the conversation information. The chatbot determining unit may determine a chatbot related to the target sub-conversation information 220 by analyzing the plurality of sub-conversation information 210 and 220 included in the conversation information. The sub-conversation information may be a text unit input and transmitted by a user, and may be a text unit transmitted at once. For example, in the case of a general messenger, one conversation unit in which text information input by the user is transmitted by pressing a transmit button may correspond to sub-conversation information. The target sub-conversation information may be sub-conversation information to which the chatbot responds among the plurality of sub-conversation information. In other words, sub-conversation information that is a response target of the response chatbot 240 may be expressed in terms of target sub-conversation information.

The chatbot determining unit may determine which chatbot among a plurality of chatbots serviced by the chatbot service device will be the response chatbot 240 to respond to the target sub-conversation information 220 sent by the user.

The chatbot determining unit may determine a default chatbot at the start of a conversation, and start a conversation with the user based on the default chatbot. Thereafter, the chatbot determining unit may determine the chatbot 240 responding to the target sub-conversation information 220 based on the analysis result of the target sub-conversation information 220 and the previous sub-conversation information 210 input by the user.

The chatbot determination unit may determine the chatbot relevance probability 230, which is a probability value for the corresponding possibility of each chatbot to the target sub-conversation information 220, based on the information on the word and context included in the target sub-conversation information 220. have.

For example, it may be assumed that a first chatbot for providing medical information and a second chatbot for providing tourism information exist. The chatbot decision unit converts the target sub-conversation information 220 into the chatbot relevance probability 230 (first chatbot relevance probability, second chatbot relevance probability) through analysis of the target sub-conversation information 220 to be analyzed. can When the first chatbot to the n-th chatbot exists in the chatbot service, (the first chatbot relevance probability, ?, the n-th chatbot relevance probability) of the first chatbot to the n-th chatbot for the target sub-conversation information 220 is the chatbot relevance It can be expressed in terms of probability 230 . For example, when the first chatbot and the second chatbot exist, the chatbot relevance probability 230 of the target sub-conversation information 220 is (0.9, 0.05) (the first chatbot relevance probability, the second chatbot relevance probability) ) can be expressed as

When the first chatbot relevance probability of the target sub-conversation information 220 is greater than or equal to the first determination threshold and the second chatbot relevance probability is less than or equal to the second determination threshold, a response to the target sub-conversation information 220 is provided. It may be determined as the task of the first chatbot, and the first chatbot may be determined as the response chatbot. If the current chatbot is the first chatbot, the first chatbot continuously responds, and if the current chatbot is the second chatbot, the chatbot transfer procedure to the first chatbot through the chatbot task transfer unit may be performed.

The chatbot determination unit determines a response to the target sub-conversation information 220 when the first chatbot relevance probability of the target sub-conversation information 220 is less than or equal to the third determination threshold and the second chatbot relevance probability is greater than or equal to the fourth determination threshold. It may be determined as the task of the second chatbot, and the second chatbot may be determined as the response chatbot. If the current chatbot is the second chatbot, the second chatbot continuously responds, and if the current chatbot is the first chatbot, the chatbot transfer procedure to the second chatbot through the chatbot decision unit may be performed.

When the chatbot determination unit has a first chatbot relevance probability of the target sub-conversation information 220 exceeds a third determination threshold value and is less than a first determination threshold value, and a second chatbot relevance probability exceeds a fourth determination threshold value and is less than a second determination threshold value , instead of determining a response to the target sub-conversation information 220 as a task of a specific chatbot, the existing chatbot may continuously respond.

In this way, a response chatbot to respond to the target sub-conversation information 220 may be determined through analysis of the target sub-conversation information 220 .

A method of determining the chatbot relevance probability 230 and the judgment threshold for determining the chatbot's task will be described in detail below.

3 is a conceptual diagram illustrating an operation of a chatbot determining unit according to an embodiment of the present invention.

3 discloses a method for determining a chatbot relevance probability of a chatbot decision unit.

Referring to FIG. 3 , the chatbot determination unit separates words included in the target sub-conversation information 300 to determine the chatbot relevance probability 330 , and considers the context on the target sub-conversation information 300 to determine the primary chatbot relevance. A probability 310 may be determined. Thereafter, the chatbot relevance probability 330 of the target sub-conversation information 300 may be finally determined in consideration of the previous sub-conversation information 320 before the target sub-conversation information 300 .

The primary chatbot relevance probability may be determined through the chatbot relevance probability for each word determined through word learning. For example, when learning for each word related to medical service and each word related to tourism service is performed through supervised learning based on word labeling, chatbot relevance probability, which is a probability value indicating which chatbot a specific word is related to (word) 305 may be determined. When the chatbot service is provided based on the first chatbot and the second chatbot, the chatbot relevance probability (word) 305 of a specific word is (first chatbot relevance probability (word), second chatbot relevance probability (word)) It can be expressed to indicate that it is a word related to a specific chatbot.

A chatbot relevance probability (word) 305 for each of the plurality of words extracted from the target sub-conversation information is determined, and the context related to the target word among the plurality of words may be analyzed to determine whether it is affirmative or negative. The target word may be a word having relatively high relevance to a specific chatbot based on the word chatbot relevance probability.

For example, 'Isn't there a tourist attraction nearby?' The target sub-conversation information of 'near', 'tourist', and 'none?' can be separated into plural words 'near', 'tourist' As the chatbot relevance probability (word) 305 for each (first chatbot relevance probability (word), second chatbot relevance probability (word)) may be determined.

Chatbot relevance probability (word) of 'near' (305) is (0.49, 0.51), chatbot relevance probability (word) of 'tourist' (305) is (0.09, 0.91), and chatbot relevance probability (word) of 'not present'. (305) can be assumed to be (0.49, 0.51). In this case, a word greater than or equal to a threshold value set for relevance to the second chatbot, such as 'tourist site', may be determined as the target word 307 .

After the target word 307 is determined, it may be determined whether a context related to the target word 307, 'no?', is a positive expression or a negative expression for the target word. In the case of 'No?', it may be a 'positive expression' as an inquiry. In this case, the primary chatbot relevance probability 310 may be determined based on the chatbot relevance probability (word) 305 of the target word 307 . For example, (0.09, 0.91), which is the chatbot relevance probability (word) 305 of 'tourist site', may be determined as the primary chatbot relevance probability 310 .

If the word 'dislike' is a context related to the target word 307, such as 'dislike tourism', 'dislike' may be recognized as a 'negative expression'. In this case, the difference between 1 and the chatbot relevance probability (word) 305 of the target word 307 may be determined as the primary chatbot relevance probability 310 . For example, if the chatbot relevance probability (word) (305) of 'tourist site' is (0.09, 0.91), and a negative expression such as 'I hate tourist destination' is connected, (0.91, (1-0.09, 1-0.91) 0.09) may be determined as the primary chatbot relevance probability 310 .

The method of determining the primary chatbot relevance probability 310 according to this context is an example, and other various methods may be used, and this embodiment may also be included in the scope of the present invention.

If there are a plurality of target words 307 , the primary chatbot relevance probability 310 may be determined based on each of the plurality of chatbot relevance probabilities (words) 305 of each of the plurality of target words 307 . For example, if the target sub-conversation information 300 is 'Tell me about tourist attractions and restaurants near here', 'tourist spots' and 'restaurants' may each be the target words 307, and 'tourism sites' and 'restaurants', respectively A primary chatbot relevance probability 310 may be determined based on the chatbot relevance probability (word) 305 of

After the determination of the primary chatbot relevance probability 310 of the target sub-conversation information 300, the chatbot relevance probability 330 of the target sub-conversation information 300 may be finally determined by reflecting the previous sub-conversation information 320. .

The previous sub-conversation information 320 may be n pieces of sub-conversation information that occurred previously in time based on the target sub-conversation information 300 . The previous chatbot relevance probability 325 is determined based on the respective chatbot relevance probability of the previous sub-conversation information 320, and the primary chatbot relevance probability 310 of the target sub-conversation information based on the previous chatbot relevance probability 325 is It may be calibrated to determine the chatbot relevance probability 330 .

For example, if n=3, the probability of relevance of the chatbot of the previous sub-conversation information 1 is (0.9, 0.1), the probability of the chatbot relevance of the previous sub-conversation information 2 is (0.7, 0.3), and the probability of the chatbot relevance of the previous sub-conversation information 3 is (0.7, 0.3). If this is (0.6, 0.2), the higher the weight is given to the more recently generated sub-conversation information based on the target sub-conversation information 300, and the previous chatbot relevance probability 325 can be determined based on Equation 1 below. have.

<Equation 1>

(Chatbot relevance probability of previous sub-conversation information 1 x 1st weight)/n+ (Chatbot relevance probability of previous sub-conversation information 2 x 2nd weight)/n+?+ (Chatbot relevance probability of previous sub-conversation information n x nth weight) )/n

Thereafter, the primary chatbot relevance probability 310 may be corrected based on the previous chatbot relevance probability 325 . For example, the chatbot relevance probability 330 of the target sub-conversation information 300 may be determined as an average value of the previous chatbot relevance probability 325 and the primary chatbot relevance probability 310 . In this way, the response chatbot 340 to respond to the target sub-conversation information 300 may be determined by reflecting the flow of the conversation.

4 is a conceptual diagram illustrating an operation of a chatbot decision unit according to an embodiment of the present invention.

4 discloses a method of determining the task of a chatbot by applying a judgment threshold value to the chatbot relevance probability of the chatbot determining unit.

Referring to FIG. 4 , when the chatbot relevance probability of the target sub-conversation information is determined, it may be determined whether the job of the chatbot is based on the chatbot relevance probability of the target sub-conversation information.

As described above, when the first chatbot for the first service and the second chatbot for the second service exist, the chatbot relevance probability is (the first chatbot relevance probability 410, the second chatbot relevance probability 420) can

According to an embodiment of the present invention, a judgment threshold value for determining whether the chatbot is a business of the chatbot based on the chatbot relevance probability may be set differently for each chatbot.

As described above, in the chatbot determination unit, the first chatbot relevance probability 410 of the target sub-conversation information is equal to or greater than the first determination threshold 411, and the second chatbot relevance probability 420 is the second determination threshold 412 or less. In this case, the response to the target sub-conversation information may be determined as the task of the first chatbot 416 , and the first chatbot 411 may be determined as the response chatbot 430 .

The chatbot determining unit is the target lower level when the first chatbot relevance probability 420 of the target sub-conversation information is below the third determination threshold 431 and the second chatbot relevance probability 420 is greater than or equal to the fourth determination threshold 414 A response to the conversation information may be determined as the task of the second chatbot 426 , and the second chatbot 426 may be determined as the response chatbot 430 .

At this time, the first judgment threshold 411 and the fourth judgment threshold 414 for determining whether the first chatbot 416 or the second chatbot 426 may be set differently, and the first judgment threshold The value 411 and the fourth judgment threshold 414 may be expressed in terms of a chatbot allocation threshold.

In this case, the second judgment threshold 412 and the third judgment threshold 413 for determining that the first chatbot and the second chatbot are not the business may be set differently, and the second judgment threshold 412 may be set differently. and the third judgment threshold 413 may be expressed in terms of a chatbot unassigned threshold.

The chatbot allocation threshold/chatbot non-allocation threshold of the chatbot may be set lower as the exclusivity of the work area between the plurality of chatbots is higher, and may be set higher as the exclusivity of the work area between the plurality of chatbots is lower.

If the chatbot service has a relatively high exclusivity between the work areas of the service provided based on each of a plurality of chatbots, even if the chatbot allocation threshold/chatbot non-allocation threshold is set low, the accuracy of judgment based on the chatbot relevance probability can be relatively high.

Conversely, if the exclusivity between the work areas of the service provided based on each of the plurality of chatbots is relatively low in the chatbot service, and the chatbot allocation threshold/chatbot non-allocation threshold is set low, the There is a chance that the wrong chatbot will respond. Therefore, when the work area is not exclusive, the accuracy of the judgment based on the chatbot relevance probability can be improved by setting the chatbot assignment threshold/chatbot non-allocation threshold high.

The exclusivity of the work area may be determined by considering the degree of overlap of words overlapping each other in the work area. For example, if a first chatbot for medical service, a second chatbot for a tourism service, and a third chatbot for a restaurant service exist, the degree of overlap of words overlapping each other in the business area of the tourism service and the restaurant service is relatively high. In the case of medical service and tourism service, medical service and restaurant service, the overlapping degree of mutually overlapping words in the business area may be relatively low.

In consideration of the exclusivity of this work area, the judgment thresholds of the first chatbot to the nth chatbot may be adjusted to determine which chatbot responds to the target sub-conversation information.

In addition, after the chatbot decision unit determines the response chatbot 430, and the response chatbot 430 responds to the target sub-conversation information, the user's sub-target conversation information is collected and the determination accuracy of the response chatbot 430 is determined. You can also expand the range of chatbot allocation threshold/chatbot non-allocation threshold by receiving feedback.

The determination accuracy of the response chatbot 430 is a value related to the determination accuracy of the response chatbot 430 of the chatbot decision unit. As the determination accuracy is higher, the chatbot allocation threshold/chatbot non-allocation threshold may be extended to reset to determine the response chatbot 430 .

5 is a conceptual diagram illustrating an operation of a chatbot task transfer unit according to an embodiment of the present invention.

In FIG. 5 , the chatbot business transfer operation of the chatbot business transfer unit is started.

Referring to FIG. 5 , the chatbot business transfer unit may perform a chatbot transfer procedure when a change in the chatbot that currently communicates with the user is required. The chatbot migration procedure may be a procedure for converting a chatbot that has been talking with a user to another chatbot.

The chatbot task transfer unit may decide whether to keep the current chatbot or transfer the business to a new chatbot according to the chatbot decision result of the chatbot decision unit.

If the current chatbot and the chatbot according to the decision result of the chatbot decision unit are the same, the chatbot service can proceed without a separate transfer procedure.

If the current chatbot and the chatbot according to the determination result of the chatbot decision unit are not the same, business transfer may be performed between the chatbots. When task transfer is performed between chatbots, a response chatbot may be set as a chatbot according to the decision result of the chatbot decision unit rather than the current chatbot, and a response to the target sub-conversation information may be transmitted through the response chatbot.

During the chatbot transfer procedure, necessary information is transferred between the plurality of chatbots, so that a smooth handoff can occur between the plurality of chatbots.

For example, when it is necessary to transfer work from the first chatbot 510 to the second chatbot 520 , the second chatbot ( Transfer conversation information 550 necessary for the conversation with the 520 may be transmitted.

The transfer conversation information 550 is sub-conversation information necessary for a conversation with the second chatbot 520 , and may be conversation information having a high chatbot relevance probability with the second chatbot 520 .

The second chatbot 520 may respond to the user for the target sub-conversation information based on the received overall conversation information or the transferred conversation information 550 .

6 is a conceptual diagram illustrating an operation of a chatbot decision unit according to an embodiment of the present invention.

6 discloses a chatbot determination method of the chatbot determination unit. In particular, the chatbot decision operation of the chatbot decision unit is started in consideration of the user's current location and work progress.

Referring to FIG. 6 , the chatbot determiner may cause the chatbot to first deliver the conversation information to the user in response to the user's current location and/or the user's current situation rather than the target sub-conversation information 600 . That is, conversation information initiated by the chatbot rather than a response based on the target sub-conversation information 600 generated by the user may be generated by the chatbot.

That is, the chatbot message generated by the response based on the target sub-conversation information 600 generated by the user is passive conversation information (chatbot) 610, and the chatbot message initiated by the chatbot is active conversation information (chatbot) It can be expressed in terms of (620).

For example, if you want to receive Korean medical services in China, medical consultation, accommodation, and tourism are conducted in advance through the chatbot, and flight schedules, medical schedules, and tour schedules (e.g., accommodation schedules, meal schedules, etc.) ) may be set.

After that, when entering through an airport from a foreign country to Korea for medical service, a chatbot for guiding the user to the next place (eg, accommodation, hospital), etc. based on the location information that the user is located at the airport Services may be provided. In this case, the chatbot may transmit active conversation information (chatbot) 620 to the user device to deliver information on how to get from the airport to the next place.

In addition, the chatbot may transmit information that the user must comply with in advance (eg, abstinence from alcohol, fasting, etc.) in consideration of the hospital schedule, and may deliver information on the route to the hospital. For example, the chatbot may deliver the active conversation information (chatbot) 620 to the user device to continuously guide information on things to be observed before surgery and a method of moving from the accommodation to the hospital.

In addition, the chatbot may deliver information on a route to a pharmacy, a location of a pharmacy, a place where meals are available, etc. as active conversation information (chatbot) 620 when surgery is completed in a hospital.

Passive conversation information (chatbot) 610 based on target sub-conversation information 600 generated by a plurality of chatbots by the user, active conversation information based on the user's current location and/or user's situation (chatbot) By delivering 620 , a customized chatbot service may be provided to the user.

7 is a conceptual diagram illustrating a chatbot service according to an embodiment of the present invention.

In FIG. 7, a chatbot service for a medical service and a tourism service is disclosed.

Referring to FIG. 7 , the chatbot allows the user to have a conversation for reservations about medical care and tourism in advance.

7(a) and 7(b) are initial screens, which by default may be screens in which the user proceeds with settings for reservations related to medical care and tourism in advance. Frequently needed questions and items to be entered can be entered through the user interface without a separate conversation.

8 is a conceptual diagram illustrating a chatbot service according to an embodiment of the present invention.

In FIG. 8, a chatbot service for a medical service and a tourism service is disclosed.

In FIG. 8(a) , when a separate inquiry is made by the user, a process in which passive conversation information (chatbot) is provided by the chatbot as a process of having a conversation with the chatbot through the chatbot is started.

When the user asks about medical care, the first chatbot (medical) related to the medical service for medical service provides an answer, and when the user asks about tourism, the second chatbot (tourism) provides an answer This is initiated.

Referring to FIGS. 8B and 8C , a process of delivering active conversation information (chatbot) to the user based on the user's location and the user's situation is disclosed.

Specifically, in FIG. 8B , when the user arrives at the airport to receive medical service, active conversation information (chatbot) that guides the route from the airport to the accommodation is disclosed.

In (c) of FIG. 8 , active conversation information (chatbot) that provides information for a meal with a pharmacy after a user receives medical services is started.

The embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include hard disks, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floppy disks. medium), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. A hardware device may be converted into one or more software modules to perform processing in accordance with the present invention, and vice versa.

In the above, the present invention has been described with reference to specific matters, such as specific components, and limited embodiments and drawings, but these are only provided to help a more general understanding of the present invention, and the present invention is not limited to the above embodiments. Those of ordinary skill in the art to which the invention pertains can make various modifications and changes from these descriptions.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and the scope of the spirit of the present invention is not limited to the scope of the scope of the present invention. will be said to belong to

Claims (6)

How to provide chatbot service,
receiving, by the chatbot service device, target sub-conversation information from the user device;
determining, by the chatbot service device, a response chatbot based on the target sub-conversation information;
transmitting, by the chatbot service device, a chatbot message to the user device in response to the target sub-conversation information through the response chatbot; and
The method further comprising the step of, by the chatbot service device, transferring the chatbot business from the current chatbot to the responding chatbot when the current chatbot is not the response chatbot;
The step of determining the response chatbot includes:
determining, by the chatbot service device, a primary chatbot relevance probability based on word information and context information of the target sub-conversation information;
determining, by the chatbot service device, a chatbot relevance probability of the target sub-conversation information based on the previous sub-conversation information and the primary chatbot relevance probability; and
determining, by the chatbot service device, the response chatbot based on the chatbot relevance probability;
The response chatbot is determined based on an nth chatbot relevance probability for each of the n chatbots and a judgment threshold set for each of the n chatbots;
The previous sub-conversation information is n sub-conversation information that occurred previously in time based on the target sub-conversation information,
The previous chatbot relevance probability is determined based on the chatbot relevance probability of each of the previous sub-conversation information,
The previous chatbot relevance probability is determined by giving a higher weight to the more recently generated sub-conversation information based on the target sub-conversation information among the n sub-conversation information,
The chatbot relevance probability of the target sub-conversation information is determined by correcting the primary chatbot relevance probability based on the previous chatbot relevance probability. According to claim 1,
The judgment threshold includes a chatbot assignment threshold and a chatbot unassignment threshold,
The chatbot allocation threshold value and the chatbot non-allocation threshold value are set to be relatively low as the exclusivity of the work area between the n chatbots is relatively high, and relatively high as the exclusivity of the work area between the n chatbots is relatively low. is set,
The exclusivity method, characterized in that determined in consideration of the degree of overlap of mutually overlapping words in the work area. delete A chatbot service device that provides a chatbot service,
a communication unit implemented to communicate with the user device; and
Including a processor operatively connected to the communication unit,
The processor receives target sub-conversation information from the user device,
determine a response chatbot based on the target sub-conversation information,
is implemented to transmit a chatbot message to the user device in response to the target sub-conversation information through the response chatbot;
The method further comprising the step of, by the chatbot service device, transferring the chatbot business from the current chatbot to the responding chatbot when the current chatbot is not the response chatbot;
The processor determines the primary chatbot relevance probability based on the word information and context information of the target sub-conversation information to determine the response chatbot, and the target sub-conversation information based on the previous sub-conversation information and the primary chatbot relevance probability is implemented to determine a chatbot relevance probability of conversation information, and the chatbot service device determines the response chatbot based on the chatbot relevance probability;
The response chatbot is determined based on an nth chatbot relevance probability for each of the n chatbots and a judgment threshold set for each of the n chatbots;
The previous sub-conversation information is n sub-conversation information that occurred previously in time based on the target sub-conversation information,
The previous chatbot relevance probability is determined based on the chatbot relevance probability of each of the previous sub-conversation information,
The previous chatbot relevance probability is determined by giving a higher weight to the more recently generated sub-conversation information based on the target sub-conversation information among the n sub-conversation information,
The chatbot relevance probability of the target sub-conversation information is determined by correcting the primary chatbot relevance probability based on the previous chatbot relevance probability. 5. The method of claim 4,
The judgment threshold includes a chatbot assignment threshold and a chatbot unassignment threshold,
The chatbot allocation threshold value and the chatbot non-allocation threshold value are set to be relatively low as the exclusivity of the work area between the n chatbots is relatively high, and relatively high as the exclusivity of the work area between the n chatbots is relatively low. is set,
The chatbot service apparatus, characterized in that the exclusivity is determined in consideration of the degree of overlap of mutually overlapping words in the work area. delete

Images