Interactive response system based on artificial intelligence

Abstract

The present invention discloses an interactive response system based on artificial intelligence, which relates to the field of human-computer interaction technology and solves the technical problem that interactive responses are not combined with the actual situation of the interactive process and the fixed interactive mode reduces the user's interactive experience. The present invention identifies the interactive information of the interactive personnel and extracts the features of the interactive information through artificial intelligence, further searches and displays the corresponding results according to the features, analyzes the interactive results in combination with historical data, selects the most matching interactive results for display, and then displays and analyzes the results, comprehensively analyzes and processes the interactive personnel, the interactive environment and the interactive problems, and generates corresponding interactive response methods based on the interactive problems in different situations, so as to fit the actual situation of the interactive users, adapt to the interactions in different situations, and improve the user's interactive experience.

Description

Technical Field

[0001] The present invention relates to the technical field of human-computer interaction, and in particular to an interactive response system based on artificial intelligence. Background Art

[0002] With the continuous development of speech recognition technology, many companies have applied speech recognition technology to traditional IVR systems, but most of these applications are still at the level of intelligent voice navigation.

[0003] According to Chinese patent application number CN201610920316.1, an interactive voice response method is disclosed, including: performing voice recognition on a user's voice input; determining a target service from a voice recognition result; comparing the confidence of the voice recognition result with a confidence range required to execute the target service; and controlling the execution of the target service based on the confidence comparison result.

[0004] The above patent utilizes the confidence level of speech recognition results to give different scripts for subsequent business interactions, differentiate the businesses, and improve the accuracy of interactions between different businesses. Some very important business regulations must be confirmed by users, regardless of how high the confidence level of the previous speech recognition results is, in order to ensure security.

[0005] However, the above patent cannot make reasonable adjustments according to the actual situation of the interacting persons when conducting interactive responses, which will affect the interactive process and further reduce the user's interactive experience. Secondly, some existing interactive response systems respond to corresponding content by obtaining the interactive information of the interacting persons when in use. However, when responding, there is no comprehensive analysis of the environmental impact of the interactive process, and a fixed response method is used for interaction. Such a response method cannot interact in accordance with the actual situation of the user, thereby reducing the user's interactive experience. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides an interactive response system based on artificial intelligence, which solves the problem that interactive responses are not combined with the actual situation of the interactive process and the fixed interactive method reduces the user's interactive experience.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an interactive response system based on artificial intelligence, comprising:

[0008] An interaction result display unit is used to search for interaction results according to the acquired interaction information, obtain interaction features by performing feature recognition on the interaction information, obtain matching results according to the interaction features and matching with the database, calculate the comprehensive matching degree between the interaction features and the matching results by integrating historical data, and simultaneously select the matching result with the largest comprehensive matching degree value to generate the interaction result, and then transmit the interaction result to the display analysis unit;

[0009] A display analysis unit is used to display and analyze the obtained interaction results, obtain personnel classification information by identifying the interacting personnel, and perform display analysis according to the interacting personnel of different classifications, and make a comprehensive judgment based on the interaction environment and the interaction problem to obtain display information, and then transmit the voice display analysis signal in the display information to the secondary analysis processing unit;

[0010] The secondary analysis processing unit is used to analyze the acquired voice presentation signal, analyze the current environmental decibels, and determine whether the current environmental decibels have an impact on the voice presentation, and generate an impact analysis signal or a normal output signal at the same time, and analyze the impact analysis signal to generate adjustment output information.

[0011] As a further solution of the present invention: it also includes an interactive information acquisition unit and a processed information output unit;

[0012] An interaction information collection unit is used to collect interaction information corresponding to the interaction user, and transmit the collected interaction information to the interaction result display unit, and the interaction information includes voice interaction information and text interaction information;

[0013] The processing information output unit is used to display the interactive results based on the adjusted output information and the normal output signal.

[0014] As a further solution of the present invention: the specific manner in which the interaction result display unit generates the interaction result is:

[0015] The system obtains the interaction information of the corresponding interacting personnel, and automatically extracts the features of the interaction information to obtain the interaction features, and at the same time matches the interaction features with the pre-stored database to obtain the matching results of the interaction features through matching;

[0016] Then obtain historical data and filter the historical records that match the current interaction features. At the same time, obtain the repeated interaction results in the historical records and record them as repeated information i, where i=1, 2, ..., j, where j represents the number of repeated information. Then obtain the proportion of interaction features in repeated information i and record it as Zi. At the same time, obtain the number of impressions Ci corresponding to repeated information i and substitute the obtained parameters into the formula The comprehensive matching degree Qi corresponding to the repeated information i is calculated, and the repeated information i with the largest comprehensive matching degree Qi value is selected to generate the interaction result.

[0017] As a further solution of the present invention: the specific manner in which the display analysis unit performs display analysis on the interaction result is:

[0018] Image information of interacting persons is acquired through corresponding image acquisition devices, and the interacting persons are classified according to the image information to obtain person classification information, and the person classification information includes young persons and elderly persons. At the same time, the current interaction environment is analyzed, and the interaction environment is specifically classified into public environment and non-public environment. The interaction problems corresponding to the interacting persons are classified to obtain problem classification information, wherein the problem classification information includes privacy problems and non-privacy problems. At the same time, the interaction environment and interaction problems are comprehensively analyzed and processed.

[0019] As a further solution of the present invention: the analysis of the interaction problem as a privacy problem by the display analysis unit is as follows:

[0020] Analyze the current interaction environment. If the interaction environment is a public place, obtain the classification information of the interacting persons. If the classification information of the interacting persons is a young person, display the corresponding interaction results in the form of text and generate display information. If the interacting persons are elderly people, display the corresponding interaction results in the form of voice playback and generate display information.

[0021] The current interaction environment is analyzed. If the interaction environment is a non-public place, the display mode of the corresponding interaction result is displayed to the interacting personnel, and display information is generated, where the display information includes a voice display signal or a text display signal, and the voice display signal is transmitted to the secondary analysis and processing unit.

[0022] As a further solution of the present invention: the display analysis unit analyzes the interaction problem as a non-privacy problem in the following manner:

[0023] The corresponding interaction results are displayed to the interacting personnel, and display information is generated according to the interaction feedback of the interacting personnel, and the display information includes a voice display signal or a text display signal, and the voice display signal is transmitted to the secondary analysis processing unit at the same time.

[0024] As a further solution of the present invention: the specific manner in which the secondary analysis processing unit analyzes the voice presentation signal is:

[0025] Get the current environment decibel and the display decibel corresponding to the voice display, and compare the current environment decibel with the display decibel. If the current environment decibel is greater than the display decibel, it means that the current environment decibel has an impact on the voice display, and generate an impact analysis signal. At the same time, analyze the impact analysis signal. If the current environment decibel is less than the display decibel, it means that the current environment decibel has no impact on the voice display, and generate a normal output signal and transmit it to the processing information output unit.

[0026] As a further solution of the present invention: the specific method in which the secondary analysis processing unit analyzes the impact analysis signal is:

[0027] The difference between the current environment decibel and the display decibel is calculated, and the difference and the display decibel are summed to obtain the display volume, and adjustment output information is generated at the same time, and then the adjustment output information is transmitted to the processing information output unit.

[0028] The present invention provides an interactive response system based on artificial intelligence. Compared with the prior art, it has the following beneficial effects:

[0029] The present invention identifies the interaction information of the interacting persons, extracts features of the interaction information through artificial intelligence, further searches and displays corresponding results based on the features, analyzes the interaction results in combination with historical data, selects the most matching interaction results for display, and then displays and analyzes the results, performs comprehensive analysis and processing based on the interacting persons, interaction environment and interaction problems, and generates corresponding interaction response methods based on interaction problems in different situations, thereby being able to fit the actual situation of the interacting users, adapt to interactions in different situations, and improve the user's interaction experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a block diagram of the system principle of the present invention. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] For example, see Figure 1 The present application provides an interactive response system based on artificial intelligence, including: an interactive information collection unit, an interactive result display unit, a display analysis unit, a secondary analysis processing unit and a processing information output unit.

[0033] The interaction information collection unit is used to collect interaction information corresponding to the interactive users and transmit the collected interaction information to the interaction result display unit. The specific collected interaction information includes voice interaction information and text interaction information.

[0034] The interactive result display unit is used to search for interactive results according to the acquired interactive information, obtain interactive features by identifying the interactive information, and obtain matching results according to the interactive features and the database, and calculate the comprehensive matching degree between the interactive features and the matching results by integrating historical data, and at the same time screen the matching results with the largest comprehensive matching degree value to generate interactive results, and then transmit the interactive results to the display analysis unit. The specific method of generating the interactive results is:

[0035] The system obtains the interaction information of the corresponding interacting personnel, and automatically extracts the features of the interaction information to obtain the interaction features. The feature extraction here is a prior art, which analyzes the input interaction information by using an artificial intelligence model, extracts the features, and matches the interaction features with the pre-stored database to obtain the matching results of the interaction features.

[0036] For example, in the application of customer service robots, when a user asks the robot for information about a product, the system will immediately capture the user's voice or text input and automatically identify key information, such as product name, model, or specific function inquiries. Subsequently, the system understands this key information through a built-in large language model and retrieves the corresponding product details or answers from the database.

[0037] Then, historical data is obtained, and historical records matching the current interaction features are screened in the historical data. The screened historical records here represent situations where the historical interaction features in the historical data are similar to the current interaction features. At the same time, repeated interaction results in the historical records are obtained, and the repeated interaction results are recorded as repeated information i, and i=1, 2, ..., j, where j represents the number of repeated information. Then, the proportion of interaction features in repeated information i is obtained and recorded as Zi. At the same time, the number of displays Ci corresponding to repeated information i is obtained, and the proportion of interaction features is represented by the number of occurrences of the interaction features in repeated information i. The specific number of occurrences is represented by the sum of the total number of occurrences of the corresponding word. At the same time, the number of displays is represented by the total number of times the repeated information is displayed to the interacting personnel, and the obtained parameters are substituted into the formula The comprehensive matching degree Qi corresponding to the repeated information i is calculated, and the repeated information i with the largest comprehensive matching degree Qi value is selected to generate the interaction result.

[0038] The display analysis unit is used to display and analyze the obtained interaction results, obtain personnel classification information by identifying the interacting personnel, and perform display analysis according to the interacting personnel of different classifications. At the same time, a comprehensive judgment is made based on the interaction environment and the interaction problem to obtain display information, and then the voice display analysis signal in the display information is transmitted to the secondary analysis processing unit. The specific method of generating the display information is:

[0039] The image information of the interacting persons is obtained through the corresponding image acquisition equipment, and the interacting persons are classified according to the image information to obtain the personnel classification information, and the personnel classification information includes young persons and elderly persons. The specific method of classification here is to conduct comprehensive recognition and analysis through artificial intelligence combined with computer vision and machine learning technologies, and the analysis methods include facial feature analysis, hair feature analysis, body language and posture, and machine learning methods. At the same time, the current interaction environment is analyzed, and the interaction environment is specifically classified into public environment and non-public environment, and the interaction problems corresponding to the interacting persons are classified to obtain problem classification information, wherein the problem classification information includes privacy problems and non-privacy problems, and the interaction environment and interaction problems are comprehensively analyzed and processed; and the method of classifying the interaction problems here is also to analyze and classify through the combination of machine learning and artificial intelligence technology. This is the existing technology and will not be elaborated on here.

[0040] If the interaction question obtained is a privacy question, the specific analysis method is as follows:

[0041] The current interaction environment is analyzed. If the interaction environment is a public place, the classification information of the interacting persons is obtained. If the classification information of the interacting persons is young persons, the corresponding interaction results are displayed in the form of text, and display information is generated. If the interacting persons are elderly persons, the corresponding interaction results are displayed in the form of voice playback, and display information is generated. The voice playback here is played through the headphones on the interaction machine, and the corresponding interacting persons put on the headphones as prompted by the machine to obtain the corresponding interaction results.

[0042] For example, at a busy airport, a young traveler uses a self-service inquiry machine to ask about lost luggage (a privacy issue). After the system confirms that the environment is public and identifies the traveler as a young person, it decides to display how to handle such situations and contact information in text on the inquiry machine's screen, rather than through a voice broadcast that could be heard by others. At a public library in the city center, an elderly visitor uses the library's information desk to ask questions about his personal borrowing records (involving privacy). After the system recognizes that the environment is public and determines that the visitor is an elderly person, it informs the relevant information in detail in voice form through the private speaker of the information desk, rather than displaying it on the screen or other places where it may be peeped by others.

[0043] The current interaction environment is analyzed. If the current interaction environment is a non-public place, the display method of the corresponding interaction result is displayed to the interacting personnel, and display information is generated, wherein the display information includes a voice display signal or a text display signal. At the same time, the voice display signal is transmitted to the secondary analysis and processing unit, and the display method here includes text display or voice display. The interacting personnel select the corresponding display method by themselves and finally obtain the display information.

[0044] If the interaction question obtained is a non-privacy question, the specific analysis method is as follows:

[0045] The corresponding interaction result is displayed to the interacting personnel, and display information is generated according to the interaction feedback of the interacting personnel, and the display information here includes a voice display signal or a text display signal, and the voice display signal is transmitted to the secondary analysis processing unit at the same time.

[0046] Embodiment 2, as the embodiment 2 of the present invention, this embodiment is implemented on the basis of embodiment 1, and is different from embodiment 1 in that.

[0047] The secondary analysis processing unit is used to analyze the acquired voice display analysis signal, analyze the current environmental decibel, and determine whether the current environmental decibel has an impact on the voice display, and generate an impact analysis signal or a normal output signal, and analyze the impact analysis signal to generate adjustment output information, and the specific analysis method is as follows:

[0048] Obtain the current environment decibel, and obtain the display decibel corresponding to the voice display, and the display decibel here is the lowest sound decibel that a person can accept under normal circumstances, and compare the current environment decibel with the display decibel. If the current environment decibel is greater than the display decibel, it means that the current environment decibel has an impact on the voice display, and an impact analysis signal is generated. If the current environment decibel is less than the display decibel, it means that the current environment decibel has no impact on the voice display, and a normal output signal is generated and transmitted to the processing information output unit;

[0049] For example, in a busy street or public transportation, the environmental noise is high, which may drown out or distort the voice information, making it difficult to hear the message clearly, thereby generating an impact analysis signal. For example, in a library or a quiet office, the environmental noise is low and will not affect the clarity of the voice information, so a normal output signal is generated.

[0050] Then the generated impact analysis signal is processed, the difference between the current environment decibel and the display decibel is calculated, and the difference is summed with the display decibel to obtain the display volume, and the adjustment output information is generated at the same time, and then the adjustment output information is transmitted to the processing information output unit.

[0051] A processing information output unit is used to display the interactive results based on the adjusted output information and the normal output signal.

[0052] Embodiment 3, as the embodiment 3 of the present invention, focuses on combining the implementation processes of embodiment 1 and embodiment 2 for implementation.

[0053] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0054] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An interactive response system based on artificial intelligence, characterized in that: include: The interactive result display unit is used to search for interactive results according to the acquired interactive information, obtain interactive features by identifying the interactive information, and obtain matching results according to the interactive features and the database, and calculate the comprehensive matching degree between the interactive features and the matching results by integrating historical data, and at the same time screen the matching results with the largest comprehensive matching degree value to generate interactive results, and then transmit the interactive results to the display analysis unit. The specific method of generating the interactive results is: Obtaining the interaction information of the corresponding interacting personnel, and automatically extracting the features of the interaction information to obtain the interaction features, and matching the interaction features with the pre-stored database to obtain the matching results of the interaction features; Then obtain historical data and filter the historical records that match the current interaction features. At the same time, obtain the interaction results that are repeated with the matching results in the historical records, and record the repeated interaction results as repeated information i, where i=1, 2, ..., j, where j represents the number of repeated information. Then obtain the proportion of interaction features in repeated information i and record it as Zi. At the same time, obtain the number of impressions Ci corresponding to repeated information i, and substitute the obtained parameters into the formula Calculate the comprehensive matching degree Qi corresponding to the repeated information i, and select the repeated information i with the largest comprehensive matching degree Qi value to generate the interaction result; A display analysis unit is used to display and analyze the obtained interaction results, obtain personnel classification information by identifying the interacting personnel, and perform display analysis according to the interacting personnel of different classifications, and make a comprehensive judgment based on the interaction environment and the interaction problem to obtain display information, and then transmit the voice display signal in the display information to the secondary analysis processing unit; The secondary analysis processing unit is used to analyze the acquired voice presentation signal, analyze the current environmental decibels, and determine whether the current environmental decibels have an impact on the voice presentation, and generate an impact analysis signal or a normal output signal at the same time, and analyze the impact analysis signal to generate adjustment output information.

2. The interactive response system based on artificial intelligence according to claim 1, characterized in that: It also includes an interactive information acquisition unit and a processed information output unit; An interaction information collection unit is used to collect interaction information corresponding to the interaction user, and transmit the collected interaction information to the interaction result display unit, and the interaction information includes voice interaction information and text interaction information; The processing information output unit is used to display the interactive results based on the adjusted output information and the normal output signal.

3. The interactive response system based on artificial intelligence according to claim 1 is characterized in that: The specific manner in which the display analysis unit performs display analysis on the interaction result is: Image information of interacting persons is acquired through corresponding image acquisition devices, and the interacting persons are classified according to the image information to obtain person classification information, and the person classification information includes young persons and elderly persons. At the same time, the current interaction environment is analyzed, and the interaction environment classification includes public environment and non-public environment. The interaction problems corresponding to the interacting persons are classified to obtain problem classification information, and the problem classification information includes privacy problems and non-privacy problems. At the same time, the interaction environment and interaction problems are comprehensively analyzed and processed.

4. The interactive response system based on artificial intelligence according to claim 3 is characterized in that: The analysis of the interactive problem as a privacy issue by the display analysis unit is as follows: Analyze the current interaction environment. If the interaction environment is a public place, obtain the classification information of the interacting persons. If the classification information of the interacting persons is a young person, display the corresponding interaction results in the form of text and generate display information. If the interacting persons are elderly people, display the corresponding interaction results in the form of voice playback and generate display information. The current interaction environment is analyzed. If the interaction environment is a non-public place, the display mode of the corresponding interaction result is displayed to the interacting personnel, and display information is generated, where the display information includes a voice display signal or a text display signal, and the voice display signal is transmitted to the secondary analysis and processing unit.

5. The interactive response system based on artificial intelligence according to claim 3 is characterized in that: The analysis method of the display analysis unit for the interaction problem as a non-privacy problem is: The corresponding interaction results are displayed to the interacting personnel, and display information is generated according to the interaction feedback of the interacting personnel, and the display information includes a voice display signal or a text display signal, and the voice display signal is transmitted to the secondary analysis processing unit at the same time.

6. The interactive response system based on artificial intelligence according to claim 1, characterized in that: The specific method in which the secondary analysis processing unit analyzes the voice presentation signal is as follows: Get the current environment decibel and the display decibel corresponding to the voice display, and compare the current environment decibel with the display decibel. If the current environment decibel is greater than the display decibel, it means that the current environment decibel has an impact on the voice display, and generate an impact analysis signal. At the same time, analyze the impact analysis signal. If the current environment decibel is less than the display decibel, it means that the current environment decibel has no impact on the voice display, and generate a normal output signal and transmit it to the processing information output unit.

7. The interactive response system based on artificial intelligence according to claim 1, characterized in that: The specific method in which the secondary analysis processing unit analyzes the impact analysis signal is as follows: The difference between the current environment decibel and the display decibel is calculated, and the difference and the display decibel are summed to obtain the display volume, and adjustment output information is generated at the same time, and then the adjustment output information is transmitted to the processing information output unit.

Images