CN114661882A - Robot chat management method, equipment and medium - Google Patents
Robot chat management method, equipment and medium Download PDFInfo
- Publication number
- CN114661882A CN114661882A CN202210327187.0A CN202210327187A CN114661882A CN 114661882 A CN114661882 A CN 114661882A CN 202210327187 A CN202210327187 A CN 202210327187A CN 114661882 A CN114661882 A CN 114661882A
- Authority
- CN
- China
- Prior art keywords
- information
- parameter
- user
- chat
- parameters
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000007726 management method Methods 0.000 title claims abstract description 12
- 230000015654 memory Effects 0.000 claims abstract description 15
- 238000000034 method Methods 0.000 claims description 33
- 230000004044 response Effects 0.000 claims description 33
- 230000008859 change Effects 0.000 claims description 27
- 230000007774 longterm Effects 0.000 claims description 23
- 238000003058 natural language processing Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 238000004891 communication Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 abstract description 2
- 238000012163 sequencing technique Methods 0.000 abstract description 2
- 241000219109 Citrullus Species 0.000 description 10
- 235000012828 Citrullus lanatus var citroides Nutrition 0.000 description 10
- 230000008569 process Effects 0.000 description 5
- 230000001149 cognitive effect Effects 0.000 description 4
- 235000013305 food Nutrition 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 230000037213 diet Effects 0.000 description 2
- 235000005911 diet Nutrition 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 206010010144 Completed suicide Diseases 0.000 description 1
- 241001465754 Metazoa Species 0.000 description 1
- 210000000988 bone and bone Anatomy 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000002996 emotional effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 235000012054 meals Nutrition 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
- G06F40/35—Discourse or dialogue representation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Machine Translation (AREA)
Abstract
The invention provides a robot chat management method, which comprises the following steps: obtaining user chat information, and analyzing information related to a target intention in the user chat information; constructing a plurality of parameter lists according to the information related to the target intention, the user chat information and the user information; generating a context responsive to the user chat information based on a plurality of parameter lists. By the robot chat management method, a robot chat system based on a human forgetting rule can be effectively optimized, the machine can perform weight sequencing on the current chat session background information by simulating the human memory forgetting rule, so that the chat context corresponding to the recent session background information is preferentially judged, a natural communication effect closer to the conversation between people is formed, and the intelligence level and the communication efficiency of the robot are improved.
Description
Technical Field
The invention relates to the field of machine learning of computers, in particular to a robot chat management method, equipment and a medium.
Background
Current deep machine learning based natural language understanding techniques have enabled semantic understanding of individual questions. In order to achieve responses that are more in line with the habits of human natural language, it is necessary to obtain more information in conjunction with the context of the current conversation to determine which of a plurality of possible answers is more in line with the current context. On the other hand, the memory of the computer software is different from the human brain, so that most of the situations can be recorded permanently. Excessive historical memory can instead interfere with the analytical understanding of the current context, making an unexpected decision.
Disclosure of Invention
In order to solve the above problems, the present invention provides a robot chat management method, including:
obtaining user chat information, and analyzing information related to a target intention in the user chat information;
constructing a plurality of parameter lists according to the information related to the target intention, the user chatting information and the user information;
generating a response context responsive to the user chat information based on the plurality of parameter lists.
In some embodiments of the present invention, constructing a plurality of parameter lists according to the information related to the target intention, the user chat information, and the user information comprises:
analyzing the information related to the target intention based on natural language processing, and extracting information parameters in the information;
establishing a short-term context information parameter list and a background information parameter list based on the information parameters; and
and constructing a target intention list through the information parameter and the short-term context information parameter list.
In some embodiments of the invention, the method further comprises:
recording the occurrence time of the information parameters in the short-term context information parameter list, setting a weight for each information parameter, and setting the change rate of the weight along with the change of time;
and updating the weight of the information parameter in the short-term context information parameter list in response to the information parameter corresponding to the occurrence in the user chat information.
In some embodiments of the invention, the method further comprises:
establishing a long-term context information parameter list, combining the short-term context information parameter list into the long-term context information parameter list when the user finishes chatting, and adding 1 to the reference count of the repeated information parameters;
in response to a reference count change of the information parameter, concurrently modifying a rate of change of the weight of the information parameter.
In some embodiments of the invention, the method further comprises:
judging whether the reference count of the information parameters in the long-term context information parameter list is greater than a preset value or not;
in response to a reference count of an information parameter in the long-term context information parameter list being greater than a predetermined value, adding the information parameter to the context information parameter list and marking the information parameter.
In some embodiments of the invention, the method further comprises:
classifying the information parameters and adding classification information of the information parameters to the parameter lists;
and generating a response context responding to the chat information of the user according to the classification information of the information parameters.
In some embodiments of the invention, the method further comprises:
analyzing the classification information of the information parameters in the parameter lists, and associating the information parameters of the same classification;
and in response to the information parameter belonging to any one of the same classification incidence relations appearing in the new chat information, updating the weights of all the information parameters belonging to the same classification incidence relations in the parameter lists.
In some embodiments of the invention, generating an answer context in response to the user chat information based on the plurality of parameter lists comprises:
and generating a plurality of context information according to the information parameters in the target intention list, and correcting the target intention according to the subsequent chat information of the user.
Another aspect of the present invention further provides a computer device, including:
at least one processor; and
a memory storing computer instructions executable on the processor, the instructions when executed by the processor implementing the steps of the method of any one of the above embodiments.
Yet another aspect of the present invention further provides a computer-readable storage medium, which stores a computer program, and the computer program realizes the steps of the method of any one of the above embodiments when executed by a processor.
By the robot chat management method, a robot chat system based on a human forgetting rule can be effectively optimized, the machine can perform weight sequencing on the current chat session background information by simulating the human memory forgetting rule, so that the chat context corresponding to the recent session background information is preferentially judged, a natural communication effect closer to the conversation between people is formed, and the intelligence level and the communication efficiency of the robot are improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of an embodiment of a method for managing a robot chat according to an embodiment of the present invention;
FIG. 2 is a block diagram of a computer device according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a readable storage medium according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the following embodiments of the present invention are described in further detail with reference to the accompanying drawings.
It should be noted that all expressions using "first" and "second" in the embodiments of the present invention are used for distinguishing two entities with the same name but different names or different parameters, and it should be noted that "first" and "second" are merely for convenience of description and should not be construed as limitations of the embodiments of the present invention, and they are not described in any more detail in the following embodiments.
As shown in fig. 1, the present invention provides a robot chat management method, including:
step S1, obtaining user chat information, and analyzing information related to the target intention in the user chat information;
step S2, constructing a plurality of parameter lists according to the information related to the target intention, the user chatting information and the user information;
step S3, generating a response context responding to the user chat information based on the plurality of parameter lists.
In the embodiment of the present invention, in step S1, according to the form of establishing a chat with the user, the chat content text sent by the user is obtained, and the chat content is analyzed by the natural language processing model, and information capable of representing the purpose intended by the user is analyzed, where the information is mostly keywords in the chat content text given by the user in the chat process.
In some embodiments of the present invention, the intention information capable of representing the purpose that the user wants to express may not be in the text of the chat content given by the user, for example, some chat content with metaphoric property, and the emotional state of the user obtained by analyzing the chat content of the user, etc. are not the chat content given by the user in the chat process. Therefore, the information capable of representing the purpose intention which the user wants to express is not limited to the chat content given by the user in the chat process.
In step S2, the information analyzed and obtained in the above step to represent the intention of the user to express the intention is stored, and a plurality of parameter lists are created as necessary for the information analyzed and obtained by the natural language processing model to represent the intention of the user to express the intention, and the parameter lists are used as "memory" of the chat robot.
In step S3, a response context is generated in response to the chat content of the user by the natural language model based on the corresponding information in the parameter lists established in step S2 as "memory" of the chat robot.
In some embodiments of the present invention, constructing a plurality of parameter lists according to the information related to the target intention, the user chat information, and the user information comprises:
analyzing the information related to the target intention based on natural language processing, and extracting information parameters in the information;
establishing a short-term context information parameter list and a background information parameter list based on the information parameters; and
and constructing a target intention list through the information parameters and the short-term context information parameter list.
In this embodiment, the information related to the target intent is further analyzed by the corresponding natural language model to obtain corresponding information parameters, where the information parameters may be corresponding keywords in the user chat context, as different chat environments or topics have different keywords, and the information parameters refer to one or more keywords that can replace the chat topic and the user purpose, and as mentioned above, some keywords may not be characters output by the user in the chat, but are categories or labels generated by the natural language model for the current state, emotion or purpose of the user according to the content of the user chat context.
Specifically, a short-term context information parameter list is established according to the analyzed information parameters, the short-term context information parameter list is a list used for recording main contents of the current conversation between the robot and the user, the short-term context information parameter list is established when the conversation starts, and all information parameters related to the conversation in the current conversation process are stored in the short-term context parameter list.
In addition, the analyzed information parameters and/or the information parameters in the short-term context information parameter list are added to the context information parameter list. The background information parameter list is a set of information parameters related to the user, and the information parameters in the background information parameter list may not be in the current chat content, i.e. not the topic in question by the current chat content. Is a set of lists of information parameters that are kept for different users for long. Specifically, the information includes parameter information closely related to the user identity, such as name, gender, age, phone, mailbox, and common address.
In particular, in some other embodiments, information parameters that appear at high frequency in the short-term context information parameter list are also added to the context information parameter list to be specially marked for the cognitive or knowledge context of the user.
Further, a target intention list is constructed according to the analyzed information parameters, wherein the target intention list is directly used for generating the conversation content of the robot responding to the user, namely the information parameters of the target intention list are input into the natural language model, and the corresponding utterance of the responding user is output by the natural language model.
In some embodiments of the invention, the method further comprises:
recording the occurrence time of the information parameters in the short-term context information parameter list, setting a weight for each information parameter, and setting the change rate of the weight along with the change of time;
and updating the weight of the information parameter in the short-term context information parameter list in response to the information parameter corresponding to the occurrence in the user chat information.
In this embodiment, when the short-term context information parameter list is in an initial state, or when the user initiates a corresponding dialog, the content of the short-term context information parameter list is empty, and the analyzed information parameters are added to the short-term context information parameter list after the session content of the user is analyzed. When the information parameter is added to the short-term context list, the information parameter to be added to the short-term context list needs to be weighted, and the weight is that the robot simulation shows the memory strength of the information parameter in the memory of the user.
Further, if the user mentions the same words or words again during the chat, the weights of the words or words are updated in the short-term context list.
It should be noted that, the robot chat management method provided by the present invention further performs abstract analysis on the chat content of the user through a natural language processing model to generate corresponding information parameters that do not exist in the chat content, for example, if the time when the user initiates the chat or the chat content relates to diet-related content, such as mentioning multiple dish names or food names, or according to the fact that the work and rest time of human being is at the meal time, when the user does not propose words such as explicit food ordering, information parameters such as food ordering are generated; or the user shows very negatively in the chat content, and if the chat content relates to buying dangerous drugs or inquiring the content with the tendency of liveness, a corresponding label with negative state and the tendency of suicide can be generated. When the robot generates the response context, a corresponding instructive utterance is generated or when the answer context is serious, a corresponding protective measure is executed according to the contact way and the address of the user in the background information parameter list and the contact way of the relatives, friends and the like of the user, for example, information is sent to the relatives and friends of the user to tell the relatives and friends of the user that the current thought state is dangerous. And if necessary, alarming according to the user address.
In some embodiments of the invention, the method further comprises:
establishing a long-term context information parameter list, merging the short-term context information parameter list into the long-term context information parameter list when the user finishes chatting each time, and adding 1 to the reference count of repeated information parameters;
concurrently modifying a rate of change of the weight of the information parameter in response to a reference count change of the information parameter.
In this embodiment, the robot chat management method provided by the present invention further creates a long-term context information parameter list for storing the cognitive range or cognitive hierarchy of the user. In particular, the data in the short-term context information parameter list is merged to the long-term context information parameter list each time a session with the user ends. And for the information parameter already existing in the long-term context information parameter list, adding a reference count item to the information parameter, and when the information parameter exists in both the long-term context information parameter list and the short-term context information parameter list, increasing the reference count item of the information parameter in the long-term context information parameter list by 1.
In addition, the invention optimizes the change rate of the weight of the information parameter, and adjusts the change rate of the weight of the information parameter according to the reference count of the information parameter. The change rate of the weight refers to the change situation of the corresponding information in the long-term memory of the user, such as watermelon, and the words with seasonal changes or periodic changes are deeply memorized in certain scenes in life. If in summer, the watermelon can be strengthened in the memory of people along with the fact that the watermelon is seen frequently. However, in winter, most people will lose memory of watermelon or taste. Therefore, when the weight becomes 0 or below a certain value, we assume that the user is not thinking, for example in winter, it is impossible for most people to think about watermelon if they do not see it or if they are given a hint by others or the environment.
To better fit the thinking habits of the person, the change of the reference count of the corresponding information parameter in the long-term context information parameter list is transformed or converted to the weight change rate of the corresponding information parameter. When the reference count is increased or the value of the reference count is high, the rate of change of the weight of the information parameter is conversely at a lower level. I.e. the user easily forgets what the information parameter represents.
In some embodiments of the present invention, the weight change rate of the information parameter is changed according to the value of the weight change rate itself. Assuming that the weight change rate of the watermelon is 0.1 and the weight value is 1, assuming that the weight value of the watermelon is equal to the product of the weight change rate and the current value subtracted from the current value every 1 day on the premise that the watermelon is not prompted, namely 1-0.1 × 1 is 0.9. The 10 days required if the thinking was that of linear changes is about the user forgetting the watermelon (not actively mentioned). However, in some embodiments of the present invention, instead of linear change, the weight change rate is increased when the weight value is lower than a certain value, for example, when the weight of watermelon is 0.5, the weight change rate is 0.15.
In some embodiments of the invention, the method further comprises:
judging whether the reference count of the information parameters in the long-term context information parameter list is greater than a preset value or not;
in response to a reference count of an information parameter in the long-term context information parameter list being greater than a predetermined value, adding the information parameter to the context information parameter list and marking the information parameter.
In this embodiment, the reference count of the information parameter in the long-term context information parameter list is determined, and if the reference count value of a certain information parameter reaches a threshold value set in advance, the information parameter is added to the context information parameter list. And marking the information parameters added into the background information parameter list in the background information parameter list to distinguish the original background information parameters which are set to represent the background information of the user.
In some embodiments of the invention, the method further comprises:
classifying the information parameters and adding classification information of the information parameters to the plurality of parameter lists;
and generating a response context responding to the chat information of the user according to the classification information of the information parameters.
In this embodiment, for the above-mentioned parameter lists: and performing additional classification on the information parameters in the short-term context parameter list, the long-term context parameter list, the background information parameter list and the target intention information list, and adding classification information to the corresponding information parameters in the parameter lists.
Furthermore, the classification may be in multiple dimensions, such as a domain or usage represented by the information parameter, which is a way of distinguishing topics or domains of the chat conversation, such as a "diet" if most of the information parameters in the chat context are food related, then some of the information parameters in the current chat are classified. If tourist attractions are present in the chat context, the classification of the information parameter extracted or analyzed based on the chat context is "tourist". The classification is based on the domain of the information parameter.
Further, the classification may also be a classification of security direction, for example, when some dangerous words appear in the chat context, or there is a danger hint in the potential intention of the chat context, for example, when the content output by the user is an effective way to ask how to treat the animal cuticle or dissolve the bone, and these questions may have corresponding personal safety problems of people, and the classification of the current user in the chat context from the security perspective of the corresponding information parameter is "dangerous". When the robot generates a corresponding response through the natural language model, the robot selects not to respond or provide a corresponding method answer, and does not execute corresponding auxiliary operation, which means to assist in purchasing a corresponding reagent.
In addition, the classification may be based on the personal awareness or personal privacy classification of the user, and when the personal awareness or personal privacy classification is performed, a range of the classification is preset, that is, a content mark related to the corresponding information parameter is designated as the personal awareness or personal privacy classification.
Further, for information parameters classified as personal privacy, information about chat context related to personal privacy will not be used to train the natural language model. This is because most chat robots have learning ability and learn a corresponding reply method from chat information of a user. For example, a company's intelligent robot model, learns an expletor when the network is open for testing. Therefore, the robot chat management method provided by the invention does not use information parameters related to personal privacy classification as training contents of the natural language model. And when generating corresponding privacy-related reply content, will also reply following content that is not learned from the relevant topics of other users.
In addition, according to the safety classification of the information parameters, a corresponding response mechanism is selectively generated according to the role information of the user, if the robot talks to a safety personnel such as police, and the like, the user conversation related to safety is allowed, and the user is allowed to be replied by generating corresponding response content based on the information parameters through a natural language model.
In some embodiments of the present invention, the classifying of the information parameters further includes classifying according to different industry fields and professional levels, or based on the cognitive level of the individual, and the classified information parameters can be selected by the user to be disclosed to the outside.
In some embodiments of the invention, the method further comprises:
analyzing the classification information of the information parameters in the parameter lists, and associating the information parameters of the same classification;
and in response to the information parameter belonging to any one of the same classification incidence relations appearing in the new chat information, updating the weights of all the information parameters belonging to the same classification incidence relations in the parameter lists.
In this embodiment, when analyzing the chat content of the user to obtain the corresponding information parameter, if there are other parameters of the same category in the chat process or in the chat content within a period of time under the category of the information parameter, when updating the weight of the information parameter, the weights of the information parameters of other categories are updated at the same time.
In some embodiments of the invention, the update of the rate of change of the weight of the information parameter is different from the update of the weight of other information parameters related to the information parameter. The updated value of the information parameter weight is greater than the updated value of the weight of the information parameter of the same category with which it is associated.
In some embodiments of the invention, generating an answer context in response to the user chat information based on the plurality of parameter lists comprises:
and generating a plurality of context information according to the information parameters in the target intention list, and correcting the target intention according to the subsequent chat information of the user.
In this embodiment, the information parameters in the target intention list also have weights, and the weights may be inherited by the information parameters in the long-term context information parameter list or the short-term context parameter list, that is, when the information parameters for realizing the target intention are insufficient and the information parameters need to be supplemented from the short-term context parameter list or the long-term context parameter list, the weights of the corresponding information parameters in the short-term context parameter list or the long-term context parameter list are inherited at the same time.
The information parameters in the target intention list are the key for directly generating the chat content of the reply user or executing the corresponding operation, so when the information parameters in the target intention list can generate a plurality of response contexts, the generated plurality of response contexts are sorted according to the weight of the information parameters in the target intention list, and the response context generated by the information parameters with the highest weight is sent to the user. And simultaneously, whether the sent response context is correct or not is determined according to the reply content of the subsequent chat content of the user.
In some embodiments of the present invention, when the corresponding response context is generated from a plurality of information parameters, the weights of the plurality of information parameters are added as a comprehensive weight value, and the generated response context with the highest comprehensive weight value is sent to the user or a corresponding operation is performed.
As shown in fig. 2, another aspect of the present invention further provides a computer device, including:
at least one processor 21; and
a memory 22, said memory 22 storing computer instructions 23 executable on said processor, said instructions 23 when executed by said processor implementing the steps of the method of any of the above embodiments.
As shown in fig. 3, a further aspect of the present invention also proposes a computer-readable storage medium 401, where the computer-readable storage medium 401 stores a computer program 402, and the computer program 402 implements the steps of the method according to any one of the above embodiments when being executed by a processor.
The foregoing is an exemplary embodiment of the present disclosure, but it should be noted that various changes and modifications could be made herein without departing from the scope of the present disclosure as defined by the appended claims. The functions, steps and/or actions of the method claims in accordance with the disclosed embodiments described herein need not be performed in any particular order. Furthermore, although elements of the disclosed embodiments of the invention may be described or claimed in the singular, the plural is contemplated unless limitation to the singular is explicitly stated.
It should be understood that, as used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly supports the exception. It should also be understood that "and/or" as used herein is meant to include any and all possible combinations of one or more of the associated listed items.
The numbers of the embodiments disclosed in the embodiments of the present invention are merely for description, and do not represent the merits of the embodiments.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
Those of ordinary skill in the art will understand that: the discussion of any embodiment above is meant to be exemplary only, and is not intended to intimate that the scope of the disclosure, including the claims, of embodiments of the invention is limited to these examples; within the framework of embodiments of the invention, also combinations between technical features of the above embodiments or different embodiments are possible, and there are many other variations of the different aspects of the embodiments of the invention described above, which are not provided in detail for the sake of brevity. Therefore, any omissions, modifications, substitutions, improvements, and the like that may be made without departing from the spirit and principles of the embodiments of the present invention are intended to be included within the scope of the embodiments of the present invention.
Claims (10)
1. A robot chat management method, comprising:
obtaining user chat information, and analyzing information related to a target intention in the user chat information;
constructing a plurality of parameter lists according to the information related to the target intention, the user chatting information and the user information;
generating a response context responsive to the user chat information based on the plurality of parameter lists.
2. The method of claim 1, wherein the constructing a plurality of parameter lists according to the information related to the target intent, the user chat information and the user information comprises:
analyzing the information related to the target intention based on natural language processing, and extracting information parameters in the information;
establishing a short-term context information parameter list and a background information parameter list based on the information parameters; and
and constructing a target intention list through the information parameter and the short-term context information parameter list.
3. The method of claim 2, further comprising:
recording the occurrence time of the information parameters in the short-term context information parameter list, setting a weight for each information parameter, and setting the change rate of the weight along with the change of time;
and updating the weight of the information parameter in the short-term context information parameter list in response to the information parameter corresponding to the occurrence in the user chat information.
4. The method of claim 3, further comprising:
establishing a long-term context information parameter list, merging the short-term context information parameter list into the long-term context information parameter list when the user finishes chatting each time, and adding 1 to the reference count of repeated information parameters;
concurrently modifying a rate of change of the weight of the information parameter in response to a reference count change of the information parameter.
5. The method of claim 4, further comprising:
judging whether the reference count of the information parameters in the long-term context information parameter list is greater than a preset value or not;
in response to a reference count of an information parameter in the long-term context information parameter list being greater than a predetermined value, adding the information parameter to the context information parameter list and marking the information parameter.
6. The method of claim 2, further comprising:
classifying the information parameters and adding classification information of the information parameters to the parameter lists;
and generating a response context responding to the chat information of the user according to the classification information of the information parameters.
7. The method of claim 6, further comprising:
analyzing the classification information of the information parameters in the parameter lists, and associating the information parameters of the same classification;
and updating the weights of all the information parameters belonging to the same classification incidence relation in the plurality of parameter lists in response to the information parameter belonging to any one of the same classification incidence relation appearing in the new chat information.
8. The method of claim 2, wherein generating an answer context responsive to the user chat information based on the plurality of parameter lists comprises:
and generating a plurality of context information according to the information parameters in the target intention list, and correcting the target intention according to the subsequent chat information of the user.
9. A computer device, comprising:
at least one processor; and
a memory storing computer instructions executable on the processor, the instructions when executed by the processor implementing the steps of the method of any one of claims 1 to 8.
10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210327187.0A CN114661882A (en) | 2022-03-30 | 2022-03-30 | Robot chat management method, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210327187.0A CN114661882A (en) | 2022-03-30 | 2022-03-30 | Robot chat management method, equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114661882A true CN114661882A (en) | 2022-06-24 |
Family
ID=82032931
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210327187.0A Pending CN114661882A (en) | 2022-03-30 | 2022-03-30 | Robot chat management method, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114661882A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080147635A1 (en) * | 2006-12-13 | 2008-06-19 | Il Im | System, apparatus and method for providing weight to information gathering engine according to situation of user and computer readable medium processing the method |
| US20090055185A1 (en) * | 2007-04-16 | 2009-02-26 | Motoki Nakade | Voice chat system, information processing apparatus, speech recognition method, keyword data electrode detection method, and program |
| US20170324867A1 (en) * | 2016-05-06 | 2017-11-09 | Genesys Telecommunications Laboratories, Inc. | System and method for managing and transitioning automated chat conversations |
| US20180174020A1 (en) * | 2016-12-21 | 2018-06-21 | Microsoft Technology Licensing, Llc | Systems and methods for an emotionally intelligent chat bot |
| CN108566330A (en) * | 2018-03-21 | 2018-09-21 | 联想(北京)有限公司 | Information processing method and the first electronic equipment |
| CN109787881A (en) * | 2018-12-26 | 2019-05-21 | 广州灵聚信息科技有限公司 | A kind of dialogue method and device with forecast function |
| CN109829039A (en) * | 2018-12-13 | 2019-05-31 | 平安科技（深圳）有限公司 | Intelligent chat method, device, computer equipment and storage medium |
| CN110287297A (en) * | 2019-05-22 | 2019-09-27 | 深圳壹账通智能科技有限公司 | Dialogue replies method, apparatus, computer equipment and computer readable storage medium |
| CN114003702A (en) * | 2021-10-28 | 2022-02-01 | 上海嬉燃科技有限公司 | Data processing method for chatting and related device |
-
2022
- 2022-03-30 CN CN202210327187.0A patent/CN114661882A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080147635A1 (en) * | 2006-12-13 | 2008-06-19 | Il Im | System, apparatus and method for providing weight to information gathering engine according to situation of user and computer readable medium processing the method |
| US20090055185A1 (en) * | 2007-04-16 | 2009-02-26 | Motoki Nakade | Voice chat system, information processing apparatus, speech recognition method, keyword data electrode detection method, and program |
| US20170324867A1 (en) * | 2016-05-06 | 2017-11-09 | Genesys Telecommunications Laboratories, Inc. | System and method for managing and transitioning automated chat conversations |
| US20180174020A1 (en) * | 2016-12-21 | 2018-06-21 | Microsoft Technology Licensing, Llc | Systems and methods for an emotionally intelligent chat bot |
| CN108566330A (en) * | 2018-03-21 | 2018-09-21 | 联想(北京)有限公司 | Information processing method and the first electronic equipment |
| CN109829039A (en) * | 2018-12-13 | 2019-05-31 | 平安科技（深圳）有限公司 | Intelligent chat method, device, computer equipment and storage medium |
| CN109787881A (en) * | 2018-12-26 | 2019-05-21 | 广州灵聚信息科技有限公司 | A kind of dialogue method and device with forecast function |
| CN110287297A (en) * | 2019-05-22 | 2019-09-27 | 深圳壹账通智能科技有限公司 | Dialogue replies method, apparatus, computer equipment and computer readable storage medium |
| CN114003702A (en) * | 2021-10-28 | 2022-02-01 | 上海嬉燃科技有限公司 | Data processing method for chatting and related device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ore et al. | Opportunities and risks of artificial intelligence in recruitment and selection | |
| Feine et al. | Gender bias in chatbot design | |
| Adamson | Postfeminism, neoliberalism and a ‘successfully’balanced femininity in celebrity CEO autobiographies | |
| US20200394615A1 (en) | Artificial intelligence assisted hybrid enterprise/candidate employment assistance platform | |
| Chowdhury et al. | # YouToo? detection of personal recollections of sexual harassment on social media | |
| CN106777257B (en) | Intelligent dialogue model construction system and method based on dialect | |
| US10783447B2 (en) | Information appropriateness assessment tool | |
| EP2634712A2 (en) | Intelligent emotion-inferring apparatus, and inferring method therefor | |
| US11080468B2 (en) | Activity modeling in email or other forms of communication | |
| CN113360622B (en) | User dialogue information processing method and device and computer equipment | |
| Sawitri et al. | Keeping up with the cyberloafer: how do cyberloafing and creative self-efficacy bear with creativity? | |
| Craiut et al. | Is technology gender neutral? A systematic literature review on gender stereotypes attached to artificial intelligence | |
| Akbari et al. | E-SoRS: A personalized and social recommender service for E-learning environments | |
| Wahyono et al. | Text mining in chat room of online learning for detection emotion using artificial intelligence | |
| CN116955573B (en) | Question searching method, device, equipment and storage medium | |
| CN114661882A (en) | Robot chat management method, equipment and medium | |
| CN114546326A (en) | Virtual human sign language generation method and system | |
| Ahmad et al. | A review of effective communication and its impact on interpersonal relationships, conflict resolution, and decision-making | |
| JP2022153240A (en) | Information processing system and virtual human resource | |
| Williams et al. | The role of conscientiousness and cue utilisation in the detection of phishing emails in controlled and naturalistic settings | |
| Merzouki et al. | How to identify competence from interactions | |
| Amiri et al. | A link prediction strategy for personalized tweet recommendation through doc2vec approach | |
| Wang et al. | Learning from e-government: an agenda for social media research in is | |
| Russell et al. | Learning Relevant Explanations | |
| Kindbom | LSTM vs Random Forest for Binary Classification of Insurance Related Text |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |