Intelligent office interaction method and device based on AI technology

Abstract

The embodiment of the invention provides an intelligent office interaction method, device and equipment based on an AI technology and a computer readable storage medium. The method comprises the following steps: receiving question information sent by a user; analyzing the question information through a trained recognition model to obtain picture content information; analyzing the picture content information, and determining a response reply mode; according to the response reply mode, matching a corresponding flow guide; and generating a response image based on the flow guide. In this way, the problems that in the prior art, the identification content is inaccurate, and the identification speed is too low are solved, and the user experience is greatly improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of data processing, and in particular to an intelligent office interaction method, device, equipment and computer-readable storage medium based on AI technology. Background Art

[0002] Mobile office platforms transfer traditional offline cumbersome processes to online for fast and efficient execution. Enterprises can create their own personalized office systems through procurement or self-construction. Like conventional online platforms, mobile office platforms also have online service functions, which guide users to complete specific operations based on their questions. Usually, such online service functions exist in the form of intelligent customer service, which is set up in the IM online instant messaging section. It realizes answer selection prompts based on the questions raised. Users select questions with similar content and obtain corresponding answers. In essence, it is similar to the "help" module function of early online systems.

[0003] In the existing technology, multi-modal questioning methods are generally supported, including text, voice, pictures, videos, etc., but the results returned are all standard answers in a set mode, among which questions and answers in picture mode are more obvious, and often there will be inaccurate recognition content, too slow recognition speed, and even irrelevant answers. On the other hand, the current AI application mode will present the content of the page in an "explanation" mode, which does not provide any positive help to the actual users, but instead causes complex and redundant interfaces. Summary of the invention

[0004] According to the embodiments of the present application, an intelligent office interaction solution based on AI technology is provided. Based on artificial intelligence technology and the characteristics of mobile office applications, the level of answer content is judged based on the content of the question, and a fast portal-style answer image is constructed according to the answer level. This solves the problems of inaccurate content recognition and slow recognition speed in the prior art, and greatly improves the user experience.

[0005] In a first aspect of the present application, a smart office interaction method based on AI technology is provided. The method comprises: Receive inquiry messages sent by users; Analyze the question information through the trained recognition model to obtain the picture content information; Analyze the picture content information to determine a response reply mode; According to the response reply mode, a corresponding process guide is matched; based on the process guide, a response image is generated.

[0006] Furthermore, the question information includes images and / or text.

[0007] Furthermore, the recognition model is trained using a stochastic gradient descent method and a back propagation algorithm.

[0008] Furthermore, it also includes: The question information is modified.

[0009] Furthermore, the modifying of the question information includes: The question information is modified in the following manner: ; Among them, u is the image pixel; v is the relative position value of the image pixel; σ is the standard deviation; π is the ratio of a circle to a circle; e is the base of natural logarithms.

[0010] Furthermore, the loss function of the recognition model is defined as follows: ; in, is the value of the cross entropy loss function; c is the image pixel; g(c) is the probability value of the true distribution; k(c) is the probability value of the predicted distribution; and is the penalty coefficient, which corrects the wrong prediction results to achieve the best prediction effect; is the corrected true distribution probability value; is the corrected predicted distribution probability value;

[0011] Furthermore, analyzing the picture content information to determine the response reply mode includes: The picture content information is analyzed in the following manner to determine the response content level score: ; Among them, G is the response content level score; For question sentences, if the question is an imperative sentence, then The value is greater than or equal to 1; otherwise, the value is 0; A score collection for the inquiry content in the question; t is the number of interrogation words that appear in the question; is the adjustment coefficient; A response reply mode is determined according to the response content level score.

[0012] In a second aspect of the present application, an intelligent office interaction device based on AI technology is provided. The device comprises: A receiving module, used for receiving the question information sent by the user; A recognition module, used to analyze the question information through a trained recognition model to obtain picture content information; An analysis module, used to analyze the image content information and determine a response reply mode; The response module is used to match the corresponding process guidance according to the response reply mode; and generate a response image based on the process guidance.

[0013] In a third aspect of the present application, an electronic device is provided, which includes a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the program, the method described above is implemented.

[0014] In a fourth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method according to the first aspect of the present application is implemented.

[0015] The AI-based intelligent office interaction method provided in the embodiment of the present application receives question information sent by a user; analyzes the question information through a trained recognition model to obtain image content information; analyzes the image content information to determine an answer reply mode; matches the answer reply mode with a corresponding process guide; and generates a response image based on the process guide, thereby solving the problems of inaccurate content recognition and slow recognition speed in the prior art and greatly improving the user experience.

[0016] It should be understood that the contents described in the Summary of the Invention are not intended to limit the key or important features of the embodiments of the present application, nor are they intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein: Figure 1 is a flow chart of an intelligent office interaction method based on AI technology according to an embodiment of the present application; Figure 2 A schematic diagram of a neural network process structure according to an embodiment of the present application; Figure 3 A schematic diagram of a smart office service quick guidance technology base according to an embodiment of the present application; Figure 4 A schematic diagram for generating a response result according to an embodiment of the present application; Figure 5 is a block diagram of an intelligent office interaction device based on AI technology according to an embodiment of the present application; Figure 6 It is a schematic diagram of the structure of a terminal device or a server suitable for implementing the embodiments of the present application. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0019] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0020] Figure 1 A flow chart of an intelligent office interaction method based on AI technology according to an embodiment of the present disclosure is shown. The method includes: S110, receiving a question message sent by a user.

[0021] In some embodiments, the cultural information sent by the user can be received by wired or wireless means. The question information usually includes images and / or texts. For example, the question information includes images and texts.

[0022] S120, analyzing the question information through the trained recognition model to obtain picture content information.

[0023] In some embodiments, Figure 2 As shown in the figure, in order to find the best balance between lightweight and recognition accuracy, a 224*224 image is used as the image and cropped through blur processing to generate an image of the appropriate size to extract features and perform convolution multiple times. Starting from the second convolution, each time it needs to be calculated through the pooling layer, and finally the corresponding probabilities of all possible recognition results are output through softmax conversion.

[0024] Specifically, the question information (image) received in step S110 is blurred, the object features in the image are extracted, the image is encrypted using base64, the encrypted 224*224 image is used as the input of the model, and the redundant parts are removed; for images larger than 224, the image is cropped at a single scale and multiple scales, and the sub-image containing the object is retained as data. Further, the processed data is corrected in the following way. That is, the high-frequency signal is filtered by convolution wave, the low-frequency signal is retained, and the image is convolved with the normal distribution to reduce image noise and reduce the level of detail: ; where T(u,v) is the value of the function, i.e., a function of the variables u and v; u is the image pixel; v is the relative position value of the image pixel; is the standard deviation, which is used to control the width of the function curve, that is, the degree of "diffusion" of the distribution; is pi, which is approximately equal to 3.14159; is the base of natural logarithms, approximately equal to 2.71828.

[0025] In some embodiments, the stochastic gradient descent method and the back propagation algorithm may be used to train the recognition model. Taking a 6-layer convolutional neural network model as an example, the initial input is a 224*224 picture.

[0026] Specifically, taking the 6-layer convolutional neural network model as an example, the initial input is a 224*224 image: In a convolutional neural network, the larger the convolution kernel, the larger the receptive field, the more image information is seen, and the better the global features are obtained. However, a large convolution kernel will lead to a surge in the amount of calculation. Considering the comprehensive factors of performance optimization and image recognition accuracy, the neural network model in this disclosure consists of 11 convolutional layers and 3 fully connected layers (14 layers). Among the 11 convolutional layers, 8 layers use 3*3 separable convolution kernels, and 3 layers use 5*5 separable convolution kernels. The construction and operation process is as follows: Input Layer: Input: I, of size H×W×C, where C is the number of color channels (e.g., for RGB images, C=3); First convolution: Convolution kernel: K1, size 3*3; The output is C1, with dimensions ,in is the number of output channels.

[0027] The specific steps are: Calculate the standard deviation: ; Calculate the variance: ; Normalize the matrix to get the value within the limited range : ; in, It is a very small constant used for numerical stability and can be pre-set according to the actual application scenario.

[0028] Entering the first deformable convolutional layer: Convolution kernel: K2, size is 3*3.

[0029] Offset generation: An additional convolutional layer is used to generate the offset δK2.

[0030] The output of this layer is C2, with a size of ; Refer to the previous step to calculate the standard deviation, variance and batch normalization to get the matrix ; Second convolutional layer: Convolution kernel: K3, size 3*3.

[0031] Output: C3, size ; Batch Normalization: ; Residual Connection: enter: and

[0032] Output: , size is ; Pooling layer: Type: max pooling, window size is 2*2; Output: , size is .

[0033] Second deformable convolutional layer: Convolution kernel: , size is 3*3.

[0034] Offset generation: An additional convolutional layer is used to generate the offset ; Output: C2, size is ; Batch Normalization: .

[0035] The third convolutional layer Convolution kernel: , size is 3*3.

[0036] Output: , size is ; Batch Normalization: .

[0037] Residual Connection: enter: and

[0038] Output: , size is ; Depthwise Separable Convolutional Layer: Depthwise Convolution: Use A 3*3 convolution kernel; Point-wise convolution: Use A 1*1 convolution kernel; Output: , size is .

[0039] Fully connected layer: enter: ; Output: F, size is 1*1* ;in, is the number of output nodes of the fully connected layer.

[0040] Output layer: Type: Softmax layer (classification tasks).

[0041] Output: O, size is 1*1* ,in is the number of categories.

[0042] Furthermore, the generated The dimensional vector features are processed by the softmax prediction layer to obtain the probability value for image classification. The calculation formula is as follows: ; in, is the number of categories.

[0043] Furthermore, the loss function can be defined as follows: ; in, is the value of the cross entropy loss function, that is, a function of two probability distributions g and k.

[0044] c is the image pixel; g(c) is the probability value of the true distribution; k(c) is the probability value of the predicted distribution; and is the penalty coefficient, which corrects the wrong prediction results to achieve the best prediction effect; is the corrected true distribution probability value; is the predicted distribution probability value after correction. Further, the recognition model trained in the above manner is used to analyze the question information to obtain the picture content information.

[0045] S130, analyzing the picture content information to determine a response reply mode.

[0046] In some embodiments, the image content information is analyzed to determine the response content level score, which indicates the singleness of the response content from high to low. When the score is too high, the response image is only marked with one key information. As the score decreases, the marked content increases and the correlation with the key information becomes lower. The specific calculation method is as follows: ; Among them, G is the response content level score; For question sentences, if the question is an imperative sentence, then The value is greater than or equal to 1; otherwise, the value is 0; The score set for the inquiry content in the question is extracted, such as "want to know", "how", "how", "have", "where" and other question words, or clear platform business name words, such as "address book", "process approval", "audit results" and score according to the importance; t is the number of interrogation words that appear in the question; is the adjustment coefficient; If G<1, it is defined as the content average mode, and all responses that meet the question will be marked. The closer it is to 0, the more marked content will be; When 1≤G<3, it is defined as the focus marking mode, focusing on the problems that users need to solve urgently. The closer it is to 1, the more other contents are marked; When G≥3, it is defined as the only annotation mode, which only annotates the problem-solving information determined in the question; Finalize the recovery mode of the response.

[0047] S140, matching a corresponding process guide according to the response reply mode; and generating a response image based on the process guide.

[0048] In some embodiments, Figure 3 As shown, according to the determined reply mode and the constructed business process quick guidance base (including mapping relationship), the relationship mapping between the question words and the annotation information is matched, and the answer image is generated based on the process guidance corresponding to the relationship mapping.

[0049] The process guidance is not a link in essence, but a plug-in that follows the user's access process. It can remember the steps of the guidance process and prompt the user to complete the process before closing the page.

[0050] The user clicks on the process guide of the key business in the response image, jumps to the corresponding page, and completes the business according to the prompt information.

[0051] Specifically, Figure 4 As shown in the figure, after the input information of the user's question in the instant messaging dialog box is determined to be the intention of the smart office scene, the execution Figure 2 The image recognition neural network process recognizes the actual question information in the question image, matches it with the business words annotated in the process in the business process quick guidance technology base, or matches the best question words through semantic understanding, returns the annotated information and process guidance, and fills it into the question image. The main body of the answer image will automatically dim, and the recognized question information will be highlighted. After the user clicks, it will jump to the execution interface and attach the process guidance to help the user quickly complete the office process.

[0052] According to the embodiments of the present disclosure, the following technical effects are achieved: Based on artificial intelligence technology and the characteristics of mobile office applications, the level of answer content is judged based on the content of the question, and a fast portal-style answer image is constructed according to the answer level. This solves the problems of inaccurate content recognition and slow recognition speed in existing technologies, and greatly improves the user experience.

[0053] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0054] The above is an introduction to the method embodiment. The following is a further explanation of the scheme described in this application through an apparatus embodiment.

[0055] Figure 5 A block diagram of an intelligent office interaction device based on AI technology according to an embodiment of the present application is shown, Figure 5 Shown include: The receiving module 510 is used to receive the question information sent by the user; The recognition module 520 is used to analyze the question information through the trained recognition model to obtain the picture content information; An analysis module 530 is used to analyze the picture content information and determine a response reply mode; The response module 540 is used to match the corresponding process guidance according to the response reply mode; and generate a response image based on the process guidance.

[0056] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0057] Figure 6 A schematic diagram of the structure of a terminal device or server suitable for implementing an embodiment of the present application is shown.

[0058] like Figure 6 As shown, the terminal device or server includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 602 or the program loaded from the storage part 608 to the random access memory (RAM) 603. In RAM 603, various programs and data required for the operation of the terminal device or server are also stored. CPU 601, ROM 602 and RAM 603 are connected to each other through a bus 604. Input / output (I / O) interface 605 is also connected to bus 704.

[0059] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed, so that a computer program read therefrom is installed into the storage section 608 as needed.

[0060] In particular, according to an embodiment of the present application, the above method flow steps can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a machine-readable medium, and the computer program includes a program code for executing the method shown in the flow chart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the central processing unit (CPU) 601, the above-mentioned functions defined in the system of the present application are executed.

[0061] It should be noted that the computer-readable medium shown in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0062] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of the code, and the aforementioned module, program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0063] The units or modules involved in the embodiments described in the present application may be implemented by software or hardware. The units or modules described may also be arranged in a processor. The names of these units or modules do not, in some cases, constitute limitations on the units or modules themselves.

[0064] As another aspect, the present application further provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable storage medium stores one or more programs, and when the above programs are used by one or more processors to execute the method described in the present application.

[0065] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of application involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the aforementioned application concept. For example, the above features are replaced with (but not limited to) technical features with similar functions applied in the present application.

Claims

1. An intelligent office interaction method based on AI technology, characterized in that: include: Receive inquiry messages sent by users; Analyze the question information through the trained recognition model to obtain the picture content information; Analyze the picture content information to determine a response reply mode; According to the response reply mode, matching the corresponding process guidance; Based on the process guidance, a response image is generated.

2. The method according to claim 1, characterized in that: The question information includes images and / or texts.

3. The method according to claim 2, characterized in that The recognition model is trained using a stochastic gradient descent method and a back-propagation algorithm.

4. The method according to claim 3, characterized in that Also includes: The question information is modified.

5. The method according to claim 4, characterized in that The modifying of the question information comprises: The question information is modified in the following manner: ； Among them, u is the image pixel; v is the relative position value of the image pixel; σ is the standard deviation; π is the ratio of a circle to a circle; e is the base of natural logarithms.

6. The method according to claim 5, characterized in that The loss function of the recognition model is defined as follows: ； in is the value of the cross entropy loss function; c is the image pixel; g(c) is the probability value of the true distribution; k(c) is the probability value of the predicted distribution; and is the penalty coefficient; is the corrected true distribution probability value; is the corrected predicted distribution probability value.

7. The method according to claim 6, characterized in that The analyzing the picture content information to determine the response reply mode includes: The picture content information is analyzed in the following manner to determine the response content level score: ； Among them, G is the response content level score; For question sentences, if the question is an imperative sentence, then The value is greater than or equal to 1; otherwise, the value is 0; A score collection for the inquiry content in the question; t is the number of interrogation words that appear in the question; is the adjustment coefficient; A response reply mode is determined according to the response content level score.

8. An intelligent office interactive device based on AI technology, characterized in that: include: A receiving module, used for receiving the question information sent by the user; A recognition module, used to analyze the question information through a trained recognition model to obtain picture content information; An analysis module, used to analyze the image content information and determine a response reply mode; A response module, used for matching the corresponding process guidance according to the response reply mode; Based on the process guidance, a response image is generated.

9. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images