An internet-based intelligent answering method and system

Abstract

The application relates to the technical field of intelligent response, and particularly discloses an intelligent response method and system based on the Internet, which comprises the following steps: receiving a query request of a user, obtaining target content, determining a display flow of the target content based on the operation behavior of the user; obtaining the face acquisition permission of the user, obtaining face data of the user based on the face acquisition permission, and generating an observation area sequence according to the face data; receiving problem information input by the user in real time, positioning a target area in the display flow based on the observation area sequence; determining an identification processing path of the problem information based on the target area, and querying response information in a response information database based on the identification processing path; and the identification processing path is used for representing the traversal query order of the response information database. The application fundamentally optimizes the response process, greatly improves the response speed, and is convenient to popularize and use.

Description

Technical Field

[0001] This invention relates to the field of intelligent response technology, specifically an intelligent response method and system based on the Internet. Background Technology

[0002] Intelligent response services are value-added voice / SMS services designed for consumers who are unable to make phone calls. With the increasing prevalence of AI technology, many service industries are designing their own unique AI chatbots to respond to customer inquiries.

[0003] Most existing AI robots' question-answering services are based on machine learning technology. They train a neural network model on a large training set. When a question is received, the robot queries the answer information in a preset answer information database and responds. With the improvement of computer equipment performance, this query process is very fast, but there is still some delay. A very important reason for the delay is that the query order of the answer information database is not good enough. How to quickly locate the answer information and improve the response speed is the technical problem that the present invention aims to solve. Summary of the Invention

[0004] The purpose of this invention is to provide an Internet-based intelligent response method and system to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An internet-based intelligent response method, the method comprising:

[0007] The system receives user query requests, obtains target content, and determines the display flow of the target content based on user operation behavior; the display flow is a sequence of subsets of target content arranged based on time information.

[0008] Obtain the user's facial recognition permission, acquire the user's facial data based on the facial recognition permission, and generate an observation region sequence based on the facial data; the observation region sequence shares the same timeline as the display process.

[0009] The system receives user-inputted question information in real time and locates the target area in the display process based on the observed area sequence.

[0010] The identification and processing path for problem information is determined based on the target area, and the response information is queried in the response information database based on the identification and processing path; the identification and processing path is used to characterize the traversal and query order of the response information database.

[0011] As a further aspect of the present invention: the steps of receiving a user's query request, obtaining target content, and determining the display process of the target content based on the user's operation behavior include:

[0012] Receive a query request sent by a user containing account information and content index, obtain the user level based on the account information, and obtain the content level based on the content index;

[0013] When the user level reaches the content level, the target content is retrieved based on the content index;

[0014] Establish a timeline, record user actions based on the timeline, and read a subset of the target content based on the actions.

[0015] The display flow is generated based on the position of the subset relative to the target content.

[0016] As a further aspect of the present invention: the steps of obtaining the user's facial recognition permission, obtaining the user's facial data based on the facial recognition permission, and generating an observation region sequence based on the facial data include:

[0017] Send a permission request to the user to obtain the user's facial recognition permission;

[0018] Based on the facial recognition permission control, the preset acquisition device acquires the user's facial data;

[0019] The facial data is used for contour recognition to obtain the eye contour;

[0020] The eye contour is input into a preset recognition model to obtain the projection area of ​​the eye contour in the display area; the projection area is a sub-region of the display area.

[0021] The projection areas are arranged according to the time information to generate an observation area sequence.

[0022] As a further aspect of the present invention: the step of inputting the eye contour into a preset recognition model to obtain the projection area of ​​the eye contour in the display area includes:

[0023] The eye contour is input into a preset recognition model to obtain a projection area containing time, which is a quadrilateral determined by four points.

[0024] Arrange the projection areas according to time, and determine the degree of overlap between each projection area and other projection areas within the preset time radius.

[0025] When there are other projection areas with regional overlap reaching a preset threshold within the time radius, the two projection areas are merged according to their time endpoints; the result is the union of the two projection areas.

[0026] As a further aspect of the present invention: the step of receiving user-inputted question information in real time and locating the target region in the display process based on the observation region sequence includes:

[0027] The system receives user input containing the input time in real time and queries the projection area in the observation area sequence based on the input time.

[0028] Based on the input time, a subset of the target content is queried in the display process; the subset corresponds to the display area.

[0029] The target region is located within a subset of the target content based on the queried projection area.

[0030] As a further aspect of the present invention: the step of determining the identification and processing path of problem information based on the target area, and querying response information in the response information database based on the identification and processing path includes:

[0031] Based on the preset mapping relationship, query the data segment corresponding to the target area in the response information database, and build the recognition and processing path with the data segment as the center;

[0032] Extract keywords from the question information, and query response information in the response information database based on the keywords according to the identification and processing path;

[0033] The steps for establishing the response information database are as follows:

[0034] Establish a question database and its response information for each sub-content of the target content, extract keywords from the questions in the question database, and establish a response information database based on the keywords and their response information.

[0035] When generating a target region, query the relative position of the content contained in the target region within the target content, and determine the data segment based on the relative position.

[0036] The present invention also provides an Internet-based intelligent response system, the system comprising:

[0037] The display process determination module is used to receive user query requests, obtain target content, and determine the display process of the target content based on user operation behavior; the display process is a sequence of a subset of target content arranged based on time information.

[0038] An observation sequence generation module is used to obtain the user's facial recognition permission, obtain the user's facial data based on the facial recognition permission, and generate an observation region sequence based on the facial data; the observation region sequence shares the same timeline as the display process.

[0039] The target area positioning module is used to receive the question information input by the user in real time and locate the target area in the display process based on the observation area sequence;

[0040] The response information query module is used to determine the identification and processing path of the problem information based on the target area, and to query the response information database based on the identification and processing path; the identification and processing path is used to characterize the traversal and query order of the response information database.

[0041] As a further aspect of the present invention: the display process determination module includes:

[0042] The level acquisition unit is used to receive a query request sent by a user containing account information and content index, acquire the user level based on the account information, and acquire the content level based on the content index;

[0043] The content acquisition unit is used to acquire target content based on the content index when the user level reaches the content level;

[0044] A data reading unit is used to establish a timeline, record user operations based on the timeline, and read a subset of the target content based on the operations.

[0045] The process generation unit is used to generate a display process based on the position of the subset relative to the target content.

[0046] As a further aspect of the present invention: the observation sequence generation module includes:

[0047] The permission acquisition unit is used to send a permission acquisition request to the user and obtain the user's facial recognition permission;

[0048] A facial acquisition unit is used to acquire the user's facial data using a preset acquisition device based on the facial acquisition permission control.

[0049] A contour recognition unit is used to perform contour recognition on the facial data to obtain the eye contour;

[0050] The projection area determination unit is used to input the eye contour into a preset recognition model and obtain the projection area of ​​the eye contour in the display area; the projection area is a sub-region of the display area;

[0051] The projection area arrangement unit is used to arrange the projection areas according to time information to generate an observation area sequence.

[0052] As a further aspect of the present invention: the target area positioning module includes:

[0053] The projection area query unit is used to receive question information containing the input time from the user in real time, and query the projection area in the observation area sequence according to the input time.

[0054] A subset query unit is used to query a subset of the target content in the display process based on the input time; the subset corresponds to the display area;

[0055] The positioning execution unit is used to locate the target region within a subset of the target content based on the queried projection area.

[0056] Compared with the prior art, the beneficial effects of the present invention are as follows: During the content display process, the present invention obtains the user's facial data in real time based on the acquired permissions, obtains the user's focus based on the facial data, determines which part of the displayed content the user is observing based on the focus, and determines the traversal order of the response information database based on this focus, fundamentally optimizing the response process, greatly improving the response speed, and facilitating widespread use. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0058] Figure 1 This is a flowchart of an internet-based intelligent response method.

[0059] Figure 2 This is the first sub-flowchart of the Internet-based intelligent response method.

[0060] Figure 3 This is the second sub-flowchart of the Internet-based intelligent response method.

[0061] Figure 4 This is the third sub-flow diagram of the Internet-based intelligent response method.

[0062] Figure 5 This is the fourth sub-flowchart of the Internet-based intelligent response method. Detailed Implementation

[0063] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0064] Example 1

[0065] Figure 1 This is a flowchart of an internet-based intelligent response method. In this embodiment of the invention, an internet-based intelligent response method includes:

[0066] Step S100: Receive the user's query request, obtain the target content, and determine the display process of the target content based on the user's operation behavior; the display process is a sequence of a subset of the target content arranged based on time information;

[0067] The target content can be an image or a document, both of which are displayed on an electronic device. The user enters a query request on the executing entity of this application, and the executing entity obtains and displays the target content. During the display process, the user will perform a series of operations on the target content, such as sliding, zooming, and folding, and adjust the display process according to the operation behavior. This process can be summarized as displaying different parts of the target content according to the user's operation behavior.

[0068] Step S200: Obtain the user's facial recognition permission, obtain the user's facial data based on the facial recognition permission, and generate an observation region sequence based on the facial data; the observation region sequence shares the same timeline as the display process;

[0069] There are many facial data collection devices. The prerequisite for collecting facial data is to have facial access permissions. Based on facial access permissions, the user's facial data is obtained and facial data is recognized. It is possible to determine where the user's focus is and, based on the focus, to find out which part of the target content the user is watching.

[0070] Step S300: Receive user-inputted question information in real time, and locate the target area in the display process based on the observed area sequence;

[0071] The system receives question information input by the user. Based on the generation time of the question information, it can query the user's observation status at that moment in the observation area sequence, and locate the target area in the target content based on the observation status.

[0072] Step S400: Determine the identification and processing path of the problem information based on the target area, and query the response information database based on the identification and processing path; the identification and processing path is used to characterize the traversal and query order of the response information database;

[0073] Different target areas correspond to different questions. During the reading process, users' questions are mostly immediate, meaning they think of something as they see it. Therefore, querying questions and their responses based on the areas observed by users can greatly improve the speed of response generation.

[0074] It should be noted that the process of querying response information is basically a database operation. Response information is queried in a preset response database based on the characteristics of the question. This application adjusts the query order, which can greatly improve the speed of response information generation.

[0075] Figure 2The first sub-flowchart of the Internet-based intelligent response method includes the following steps: receiving a user's query request, obtaining the target content, and determining the display process of the target content based on the user's operation behavior:

[0076] Step S101: Receive a query request sent by the user containing account information and content index, obtain the user level based on the account information, and obtain the content level based on the content index;

[0077] Different users can view different target content, and this process is distinguished by user level and content level;

[0078] Step S102: When the user level reaches the content level, obtain the target content according to the content index;

[0079] When the user level is sufficient, the target content is retrieved based on the content index.

[0080] Step S103: Establish a timeline, record user operations based on the timeline, and read a subset of the target content based on the operations.

[0081] The reason for reading and displaying a portion of the target content based on the operation behavior is that the display area is limited, and only a portion of the target content can be displayed each time.

[0082] Step S104: Generate a display flow based on the position of the subset relative to the target content;

[0083] The display process is the arrangement order of the positions of subsets displayed at different times relative to the target content.

[0084] Figure 3 The second sub-flowchart of the Internet-based intelligent response method includes the steps of obtaining the user's facial recognition permission, obtaining the user's facial data based on the facial recognition permission, and generating an observation region sequence based on the facial data.

[0085] Step S201: Send a permission request to the user to obtain the user's facial recognition permission;

[0086] The simplest way to obtain permissions is to use a pop-up window to request permissions from the user.

[0087] Step S202: Based on the facial recognition permission control, the preset acquisition device acquires the user's facial data;

[0088] The data acquisition device can be an existing camera.

[0089] Step S203: Perform contour recognition on the facial data to obtain the eye contour;

[0090] Locating the eye contour in facial data is not difficult, and there are many related technical solutions available.

[0091] Step S204: Input the eye contour into a preset recognition model to obtain the projection area of ​​the eye contour in the display area; the projection area is a sub-region of the display area;

[0092] The state of the eyes and the eyeballs can reflect where the user's focus is. The focus is on a certain area of ​​the display area, namely the projection area mentioned above.

[0093] Step S205: Arrange the projection areas according to the time information to generate an observation area sequence;

[0094] By arranging the projection areas according to time, the sequence of observation areas can be obtained.

[0095] In one embodiment of the technical solution of the present invention, the step of inputting the eye contour into a preset recognition model and obtaining the projection area of ​​the eye contour in the display area includes:

[0096] The eye contour is input into a preset recognition model to obtain a projection area containing time, which is a quadrilateral determined by four points.

[0097] The projection area is represented by four points (coordinates), and the projection area is a sub-region of the display area; the correspondence between the eye contour and the projection area is determined by a preset recognition model; the recognition model can use an existing neural network model.

[0098] Arrange the projection areas according to time, and determine the degree of overlap between each projection area and other projection areas within the preset time radius.

[0099] The human eye moves at extremely high speeds. Sometimes, the human eye changes rapidly, but the point of focus does not change. Only when the duration of focus on a certain projection area is within a certain range can it be considered a point of focus.

[0100] When there are other projection areas with regional overlap reaching a preset threshold within the time radius, the two projection areas are merged according to their time endpoints; the merged area is calculated as the union of the two projection areas.

[0101] Based on the attention time of each projection area, some projection areas with a time significantly less than the preset threshold can be removed, making the obtained observation area sequence more consistent with the focus.

[0102] Figure 4This is a flowchart of the third sub-process of the Internet-based intelligent response method. The step of receiving user-inputted question information in real time and locating the target area in the display process based on the observed area sequence includes:

[0103] Step S301: Receive the question information containing the input time from the user in real time, and query the projection area in the observation area sequence according to the input time;

[0104] The projection area is queried in the observation area sequence based on the input time.

[0105] Step S302: Based on the input time, query a subset of the target content in the display process; the subset corresponds to the display area;

[0106] Based on the input time, query the display process to determine which part of the target content the current display area corresponds to;

[0107] Step S303: Locate the target region within a subset of the target content based on the queried projection area;

[0108] The location of the projection area within the display area and the corresponding part of the display area within the target content indicate which part of the target content the user is currently focusing on.

[0109] Figure 5 This is the fourth sub-flowchart of the Internet-based intelligent response method. The steps of determining the identification and processing path of the problem information based on the target area and querying the response information database based on the identification and processing path include:

[0110] Step S401: Query the data segment corresponding to the target area in the response information database according to the preset mapping relationship, and build the recognition and processing path with the data segment as the center;

[0111] Step S402: Extract keywords from the question information, and query response information in the response information database based on the keywords according to the identification and processing path;

[0112] The purpose of steps S401 and S402 is to determine where and how to begin traversing the response information database; the response information database contains keyword items and response information items; response information can be retrieved by using keywords.

[0113] In one embodiment of the technical solution of the present invention, the steps for establishing the response information database are as follows:

[0114] Establish a question database and its response information for each sub-content of the target content, extract keywords from the questions in the question database, and establish a response information database based on the keywords and their response information.

[0115] When generating a target region, query the relative position of the content contained in the target region within the target content, and determine the data segment based on the relative position.

[0116] The process of establishing the response information database is completed in advance. Staff members use statistical data to pre-determine and answer the possible questions in each part of the target content; extract the keywords of the questions, store the corresponding response information in the keywords, and when a question is received, the response information can be obtained by direct matching.

[0117] Example 2

[0118] In this embodiment of the invention, an Internet-based intelligent response system is provided, the system comprising:

[0119] The display process determination module is used to receive user query requests, obtain target content, and determine the display process of the target content based on user operation behavior; the display process is a sequence of a subset of target content arranged based on time information.

[0120] An observation sequence generation module is used to obtain the user's facial recognition permission, obtain the user's facial data based on the facial recognition permission, and generate an observation region sequence based on the facial data; the observation region sequence shares the same timeline as the display process.

[0121] The target area positioning module is used to receive the question information input by the user in real time and locate the target area in the display process based on the observation area sequence;

[0122] The response information query module is used to determine the identification and processing path of the problem information based on the target area, and to query the response information database based on the identification and processing path; the identification and processing path is used to characterize the traversal and query order of the response information database.

[0123] The display process determination module includes:

[0124] The level acquisition unit is used to receive a query request sent by a user containing account information and content index, acquire the user level based on the account information, and acquire the content level based on the content index;

[0125] The content acquisition unit is used to acquire target content based on the content index when the user level reaches the content level;

[0126] A data reading unit is used to establish a timeline, record user operations based on the timeline, and read a subset of the target content based on the operations.

[0127] The process generation unit is used to generate a display process based on the position of the subset relative to the target content.

[0128] The observation sequence generation module includes:

[0129] The permission acquisition unit is used to send a permission acquisition request to the user and obtain the user's facial recognition permission;

[0130] A facial acquisition unit is used to acquire the user's facial data using a preset acquisition device based on the facial acquisition permission control.

[0131] A contour recognition unit is used to perform contour recognition on the facial data to obtain the eye contour;

[0132] The projection area determination unit is used to input the eye contour into a preset recognition model and obtain the projection area of ​​the eye contour in the display area; the projection area is a sub-region of the display area;

[0133] The projection area arrangement unit is used to arrange the projection areas according to time information to generate an observation area sequence.

[0134] The target area positioning module includes:

[0135] The projection area query unit is used to receive question information containing the input time from the user in real time, and query the projection area in the observation area sequence according to the input time.

[0136] A subset query unit is used to query a subset of the target content in the display process based on the input time; the subset corresponds to the display area;

[0137] The positioning execution unit is used to locate the target region within a subset of the target content based on the queried projection area.

[0138] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An internet-based intelligent response method, characterized in that, The method includes: The system receives user query requests, obtains target content, and determines the display flow of the target content based on user operation behavior; the display flow is a sequence of subsets of target content arranged based on time information. Obtain the user's facial recognition permission, acquire the user's facial data based on the facial recognition permission, and generate an observation region sequence based on the facial data; the observation region sequence shares the same timeline as the display process. The system receives user-inputted question information in real time and locates the target area in the display process based on the observed area sequence. The identification and processing path for the problem information is determined based on the target area, and the response information is queried in the response information database based on the identification and processing path; the identification and processing path is used to characterize the traversal and query order of the response information database; The steps of obtaining the user's facial recognition permission, obtaining the user's facial data based on the facial recognition permission, and generating an observation region sequence based on the facial data include: Send a permission request to the user to obtain the user's facial recognition permission; Based on the facial recognition permission control, the preset acquisition device acquires the user's facial data; The facial data is used for contour recognition to obtain the eye contour; The eye contour is input into a preset recognition model to obtain the projection area of ​​the eye contour in the display area; the projection area is a sub-region of the display area. Arrange the projection areas according to time information to generate an observation area sequence; The step of receiving user-inputted question information in real time and locating the target area in the display process based on the observation area sequence includes: The system receives user input containing the input time in real time and queries the projection area in the observation area sequence based on the input time. Based on the input time, a subset of the target content is queried in the display process; the subset corresponds to the display area. Locate the target region within a subset of the target content based on the queried projection area; The steps of determining the identification and processing path for problem information based on the target region, and querying response information in the response information database based on the identification and processing path, include: Based on the preset mapping relationship, query the data segment corresponding to the target area in the response information database, and build the recognition and processing path with the data segment as the center; Extract keywords from the question information, and query response information in the response information database based on the keywords according to the identification and processing path; The steps for establishing the response information database are as follows: Establish a question database and its response information for each sub-content of the target content, extract keywords from the questions in the question database, and establish a response information database based on the keywords and their response information. When generating a target region, query the relative position of the content contained in the target region within the target content, and determine the data segment based on the relative position.

2. The Internet-based intelligent response method according to claim 1, characterized in that, The steps of receiving a user's query request, obtaining the target content, and determining the display process of the target content based on the user's operation behavior include: Receive a query request sent by a user containing account information and content index, obtain the user level based on the account information, and obtain the content level based on the content index; When the user level reaches the content level, the target content is retrieved based on the content index; Establish a timeline, record user actions based on the timeline, and read a subset of the target content based on the actions. The display flow is generated based on the position of the subset relative to the target content.

3. The Internet-based intelligent response method according to claim 1, characterized in that, The step of inputting the eye contour into a preset recognition model to obtain the projection area of ​​the eye contour in the display area includes: The eye contour is input into a preset recognition model to obtain a projection area containing time, which is a quadrilateral determined by four points. Arrange the projection areas according to time, and determine the degree of overlap between each projection area and other projection areas within the preset time radius. When there are other projection areas with regional overlap reaching a preset threshold within the time radius, the two projection areas are merged according to their time endpoints; the result is the union of the two projection areas.

Images