WO2021169630A1 - 可配置化报告生成方法、装置、设备及可读存储介质 - Google Patents
可配置化报告生成方法、装置、设备及可读存储介质 Download PDFInfo
- Publication number
- WO2021169630A1 WO2021169630A1 PCT/CN2021/071057 CN2021071057W WO2021169630A1 WO 2021169630 A1 WO2021169630 A1 WO 2021169630A1 CN 2021071057 W CN2021071057 W CN 2021071057W WO 2021169630 A1 WO2021169630 A1 WO 2021169630A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- micro
- expression
- target
- risk
- data
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/904—Browsing; Visualisation therefor
Definitions
- This application relates to the technical field of data visualization, and in particular to a configurable report generation method, device, equipment, and computer-readable storage medium.
- Human-machine dialogue is a sub-direction of the field of artificial intelligence, which specifically refers to that humans can interact with machines through natural language.
- One of the problems solved by human-machine dialogue is task-driven multi-round dialogue.
- the purpose of task-driven multi-round dialogue is to complete tasks, such as helping users complete tasks such as ordering air tickets.
- the completion of such tasks requires multiple rounds of machines and users. Interaction, so as to collect user needs and related information, and finally complete the task of ordering air tickets.
- the inventor realizes that collecting user needs and related information requires relevant business personnel to collect it in the dialogue record after the end of the man-machine dialogue. There is no better way to flow the dialogue through the scene, the user’s intention, Slot information, conversation content, etc. are integrated into a unified report. It requires relevant salespersons to query the conversation records by themselves. Manually generate reports. The efficiency is relatively low. Moreover, the collected information is difficult to determine the truthfulness of the final report. Low.
- the main purpose of this application is to provide a configurable report generation method, device, equipment, and computer-readable storage medium, aiming to solve the technical problems of low efficiency and low accuracy of existing report generation methods.
- the present application provides a configurable report generation method.
- the configurable report generation method includes the following steps: when a configuration instruction is detected, a configuration interface corresponding to the configuration instruction is displayed, and a configuration interface based on the configuration is received. The configuration parameters entered on the interface; if a collection instruction triggered based on the configuration parameters is detected, the report element of the configuration parameter is determined, and the target data is extracted from the data source corresponding to the report element; and the target data is determined to correspond to Based on the data risk attribute, generate a target report from the target data.
- the present application also provides a configurable report generating device
- the configurable report generating device includes: a configuration module for displaying the configuration interface corresponding to the configuration instruction when the configuration instruction is detected, And receive the configuration parameters input based on the configuration interface; the extraction module is used to determine the report element of the configuration parameter if a collection instruction triggered based on the configuration parameter is detected, and list the data source corresponding to the report element
- the target data is extracted from the data;
- the risk module is used to determine the data risk attribute corresponding to the target data;
- the generating module is used to generate a target report based on the data risk attribute of the target data.
- the present application also provides a configurable report generating device, the configurable report generating device includes a processor, a memory, and a device that is stored in the memory and can be executed by the processor.
- a configurable report generation program wherein when the configurable report generation program is executed by the processor, the following method is implemented: when a configuration instruction is detected, a configuration interface corresponding to the configuration instruction is displayed, and a configuration interface based on the configuration interface is received Input configuration parameters; if a collection instruction triggered based on the configuration parameters is detected, the report element of the configuration parameter is determined, and target data is extracted from the data source corresponding to the report element; and the target data corresponding to the target data is determined Data risk attributes; based on the data risk attributes, a target report is generated from the target data.
- this application also provides a computer-readable storage medium that stores a configurable report generation program, where the configurable report generation program is executed by a processor.
- a configuration instruction when a configuration instruction is detected, the configuration interface corresponding to the configuration instruction is displayed, and the configuration parameter input based on the configuration interface is received; if a collection instruction triggered based on the configuration parameter is detected, the configuration is determined And extract target data from the data source corresponding to the report element; determine the data risk attribute corresponding to the target data; based on the data risk attribute, generate a target report on the target data.
- This application collects corresponding target data according to the configuration parameters configured by the user, and conducts risk assessment on the target data, and finally generates a target report from the risk assessment results and target data, which can intuitively reflect the authenticity of the data and improve the conversion of data into reports. Efficiency, to achieve data visualization.
- FIG. 1 is a schematic diagram of the hardware structure of the configurable report generating device involved in the solution of the embodiment of the application.
- Fig. 2 is a schematic flowchart of a first embodiment of a method for generating a configurable report according to this application.
- FIG. 3 is a schematic diagram of a configuration interface of a generating device in an embodiment of a method for generating a configurable report according to the present application.
- Fig. 4 is a schematic flowchart of a second embodiment of a configurable report generating apparatus according to the present application.
- Figure 5 is a schematic diagram of a target report generation interface of this application.
- FIG. 6 is a schematic diagram of the functional modules of the first embodiment of the configurable report generating apparatus according to the present application.
- the technical solution of this application can be applied to the fields of artificial intelligence, smart city, digital healthcare, blockchain and/or big data technology to realize data visualization.
- the data involved in this application such as data risk attributes, model output results, and/or reports, etc.
- the configurable report generation method involved in the embodiments of the present application is mainly applied to a configurable report generating device.
- the configurable report generating device may be a device with display and processing functions such as a PC, a portable computer, and a mobile terminal.
- FIG. 1 is a schematic diagram of the hardware structure of the configurable report generating device involved in the solution of the embodiment of the application.
- the configurable report generating device may include a processor 1001 (for example, a CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005.
- the communication bus 1002 is used to realize the connection and communication between these components;
- the user interface 1003 may include a display (Display), an input unit such as a keyboard (Keyboard);
- the network interface 1004 may optionally include a standard wired interface, a wireless interface (Such as WI-FI interface);
- the memory 1005 can be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory.
- the memory 1005 can optionally also be a storage device independent of the aforementioned processor 1001 .
- FIG. 1 does not constitute a limitation on the configurable report generating device, and may include more or less components than shown in the figure, or a combination of certain components, or different components. Component arrangement.
- the memory 1005 as a computer-readable storage medium in FIG. 1 may include an operating system, a network communication module, and a configurable report generation program.
- the network communication module is mainly used to connect to the server and perform data communication with the server; and the processor 1001 can call the configurable report generation program stored in the memory 1005, and execute the configurable report provided by the embodiment of this application Generation method.
- the embodiment of the present application provides a method for generating a configurable report.
- FIG. 2 is a schematic flowchart of a first embodiment of a method for generating a configurable report according to the present application.
- the method for generating a configurable report includes the following steps.
- Step S10 When a configuration instruction is detected, a configuration interface corresponding to the configuration instruction is displayed, and configuration parameters input based on the configuration interface are received.
- Step S20 If a collection instruction triggered based on the configuration parameter is detected, the report element of the configuration parameter is determined, and the target data is extracted from the data source corresponding to the report element.
- Step S30 Determine the data risk attribute corresponding to the target data.
- step S40 based on the data risk attribute, a target report is generated from the target data.
- the method for generating a configurable report in this embodiment is applied to a configurable report generating device, hereinafter referred to as a generating device.
- the generating device can be a fixed terminal such as a computer, or a mobile terminal such as a mobile phone or an ipad.
- the connection of the round dialogue device can obtain the human-machine dialogue record stored in the human-machine multi-round dialogue device according to the pre-configured configuration parameters according to the collection instructions, and generate a visual report according to a certain format, so that the relevant salesperson can intuitively Determine the needs of the current user and the authenticity of the user to facilitate subsequent business development; the generating device can also have the function of man-machine dialogue.
- the man-machine dialogue record is stored, and according to the collection instructions, Pre-configured configuration parameters, obtain stored man-machine conversation records to generate reports for subsequent business development.
- this embodiment collects corresponding target data, and conducts risk assessment on the target data. Finally, the risk assessment result and target data are generated into a report, which can intuitively reflect the authenticity of the data and improve the conversion of data into reports. Efficiency, to achieve the visualization of target data.
- Step S10 When a configuration instruction is detected, a configuration interface corresponding to the configuration instruction is displayed, and configuration parameters input based on the configuration interface are received.
- the user can configure the parameters of the generating device.
- the specific generating device detects the configuration instruction issued by the user, it displays the configuration interface corresponding to the configuration instruction for the user to configure independently. After detecting that the user configuration is complete After that, you can accept the configuration parameters input by the user.
- the configuration interface includes the comprehensive element configuration area, the parameter element configuration area, and the intention element configuration area.
- the user can configure the parameters in these areas.
- each configuration area includes the report element name, data source, whether to output, etc., for example, in the comprehensive element configuration area, the report element name is micro-expression risk and text information abnormality, and the corresponding data source is micro-expression risk. Compare the risk with the text, and the user chooses whether to output.
- the parameter configuration area is set with a new button, as well as an edit and delete button. The user can click the new button to trigger the new report element, and the edit button can modify the configuration in the configured report element, and delete The button can delete the configured report elements, etc.
- the user does not need to configure the parameter-raising element area, that is, the parameter-raising element area is empty, while the integrated element can only be configured to output.
- the configuration instruction can be triggered by the generating device, that is, in addition to being issued by the user, the configuration instruction can also be triggered in a specific scenario, that is, in another implementation scenario, the generating device can monitor people in real time. Whether the machine dialogue is over, when the end of the man-machine dialogue is detected, the configuration instruction is triggered to display the configuration interface corresponding to the configuration instruction for the user to configure the configuration parameters.
- Step S20 If a collection instruction triggered based on the configuration parameter is detected, the report element of the configuration parameter is determined, and the target data is extracted from the data source corresponding to the report element.
- the generating device detects whether there is a collection instruction triggered based on the configuration parameters.
- the collection instruction can be issued by the user. For example, after the user configures the configuration parameters, click the relevant OK button to trigger Collect instructions.
- the collection instruction can also be triggered by the generating device, that is, in addition to being issued by the user, the collection instruction can also be triggered in a specific scenario, that is, in another implementation scenario, the generating device can monitor in real time Whether the man-machine dialogue is over, when the end of the man-machine dialogue is detected, a collection instruction is triggered to generate a device to collect target data.
- the generating device When the generating device detects the acquisition instruction triggered based on the configuration parameter, it determines the report element of the configuration parameter, and uses the report element to extract the target data from the corresponding data source. Specifically, after the report element of the configuration parameter is determined, the report is determined The parameter name of the element and the data source of the report element. Then, in the data source, the target data corresponding to the parameter name is collected. That is, in this step, the user establishes the association relationship between the report element and the data source in advance, and Establish the association relationship between the parameter name and the target data, such as the mapping relationship. Therefore, the generating device can determine the corresponding data source according to the report element configured by the user, and then collect the target data from the data source according to the parameter name of the report element.
- the user has previously configured a report element named name in the parameter element area, and the generating device can determine the corresponding data source according to the "name” as parameter extraction, and the "name” in the data source
- the parameter name is "name”. Therefore, the generating device collects target information related to "name” from the data source for parameter extraction, such as Zhang San and others.
- the data source is the data recorded during the man-machine dialogue, that is, in the case of multiple rounds of man-machine dialogue, the content of the man-machine dialogue will be recorded and saved as the data source.
- Step S30 Determine the data risk attribute corresponding to the target data.
- step S30 includes the following step.
- Step a Obtain micro-expression information of the target data, and determine whether the micro-expression information is a risk expression.
- the target data and the micro expressions are also stored in a one-to-one correspondence. For example, when the user’s name is asked, in addition to recording the user’s answer, the user’s micro expression information when answering the question is also collected. Therefore, the generating device is collecting the target After the data is obtained, the micro-expression information of the target data can be correspondingly obtained, and then, by judging the micro-expression information, it is determined whether the target data has a micro-expression risk.
- the micro-expression information corresponding to the current target data is compared with the micro-expression that has a risk flag set in advance to determine whether the current micro-expression information is a risk expression.
- step a includes the following steps.
- the generating device when it obtains the micro-expression information of the target data, it determines the number of times the micro-expression information appears and the number of frames, that is, how many times the micro-expression information appears in the scene corresponding to the target data, and the frames of the micro-expression number. For example, when the user answers the name, it lasts for 10 frames. In these 10 frames, there are 3 blinks, then when the target data is a name, the corresponding number of micro expressions is 3, and the number of frames is 10 frames.
- the number of frames can be the number of frames of the scene corresponding to the target data, or the number of frames from the appearance of the micro expression to the end of the micro expression. For example, when the user answers the name, it lasts for 10 frames, but the micro expression It starts to appear in the 3rd frame and does not appear after the 7th frame. Then, the number of frames of the micro-expression information is 4 frames at this time.
- micro expression information is a risk expression.
- the generating device calculates the frame-to-time ratio based on the number of times the micro-expression appears and the number of frames, that is, divides the number of times the micro-expression information appears by the number of frames to obtain the frame-to-time ratio, and determines whether the frame-to-time ratio is greater than a preset threshold. If it is, it is determined that the micro-expression information is a risk expression.
- step b if yes, it is determined that there is a risk of micro-expression in the target data.
- the target data has a micro-expression risk, indicating that the current target data has low credibility.
- step a includes: acquiring the micro-expression information of the target data, and obtaining the micro-expression information from the micro-expression information Select the target micro-expression information; it is understandable that there may be multiple situations in the micro-expression information of the target data. For example, when the user answers the address question, there are also micro-expression information such as blinking, pursing, and smiling. Expression information is used as the basis for judging whether there is a micro-expression risk in the target data.
- the specific selection method can choose the longest micro-expression as the target micro-expression, such as blinking for 10 frames, pursing the mouth for 5 frames, smiling for 2 frames, etc., then select Blink as a target micro expression.
- the target micro-expression information is input into the micro-expression model to obtain the model output result; according to the model output result, it is judged whether the target micro-expression information information is a risk expression.
- the generating device inputs the acquired target micro-expression information into the pre-trained micro-expression model, and judges whether the current target micro-expression information is a risk expression through the model output result output by the micro-expression model.
- the micro-expression model is trained in advance based on historical dialogue records. Specifically, through prior knowledge, the known historical micro-expression information is used as the input of the model, and the known judgment result is used as the output of the micro-expression. Through model training, Get the micro-expression model.
- the number of frames that the micro-expression information appears can be used as the input of the micro-expression model, such as the number of frames for pursing the mouth; or the number of times the micro-expression information appears as the input of the model, such as the number of blinks; or the number of micro-expression information
- the number of times the information appears and the number of frames are used as input to the model, such as the number of frowns and the number of frowning frames. Since the historical dialogue records are known, the corresponding judgment result can be used as the output of the model to train the micro-expression model.
- micro-expression information there are 54 types of micro-expression information. Generally, the top five types with the highest frequency in answering the current question are selected as the target micro-expression information, and then it is judged whether there is micro-expression information that meets the risk conditions in the five types. The micro-expression information this time is determined to be a risk expression.
- step S40 based on the data risk attribute, a target report is generated from the target data.
- the generating device generates a target report from the target data according to the data risk attribute. Specifically, the target data belonging to the micro-expression risk is marked, and the target data is generated according to a preset format to generate a target report.
- the generating device needs to determine the risk attributes of the target data in turn, and finally, generating the corresponding target report, the generating device will have the risk of micro-expression
- the target data is summarized as abnormal points and displayed on the display interface according to the preset format. For example, the target data with micro-expression risk and the target data without micro-expression risk are displayed together. When displayed together, they are displayed in different colors to distinguish them, such as Those with micro-expression risk are marked in red, and those with no micro-expression risk are marked in green, etc.; or the target data with micro-expression risk and target data without micro-expression risk are displayed in multi-window switching, etc.
- the configuration interface corresponding to the configuration instruction is displayed, and the configuration parameter input based on the configuration interface is received; if a collection instruction triggered based on the configuration parameter is detected, the configuration parameter is determined Report element, and extract target data from the data source corresponding to the report element; determine the data risk attribute corresponding to the target data; based on the data risk attribute, generate a target report on the target data.
- this application collects corresponding target data and conducts risk assessment on the target data.
- the risk assessment result and target data are generated into a report, which can intuitively reflect the authenticity of the data and improve the efficiency of data conversion into reports. , To achieve the visualization of target data.
- step S30 includes: step S31: For the text content of the target data, the text content is compared with the pre-stored input information to determine whether the text content is true; in step S32, if it is false, it is determined that the target data has a text information risk.
- the data risk attribute includes text information risk. Therefore, when determining whether the target data has text information risk, it is to determine whether the text content of the target data is true, so that the subsequently generated target report is true and reliable.
- Step S31 Extract the text content of the target data, compare the text content with pre-stored input information, and determine whether the text content is authentic.
- the generating device extracts the text content of the target data, compares the text content with pre-stored input information, and determines whether the text content is authentic.
- the incoming information refers to the authenticity information entered in advance, that is, the generating device has the incoming information in advance, after obtaining the target data, extracts the text content of the target data, and compares the extracted text content with the incoming information Yes, to determine whether the current text content is true.
- the user’s identity is generally determined first, and then the user is called to verify and record the relevant information.
- the user is submitting a loan application
- the submitted personal information and other information is the incoming information
- the generating device knows it in advance, and then collects the user’s personal information, that is, the text content mentioned above, through multiple rounds of man-machine dialogue during the remote face-to-face review. Then compare the two.
- the user submits the information at the time of application and shows that the age is 30 years old.
- the human-computer dialogue if the user answers that his age is 35 years old, it is determined that the two comparison results do not match.
- the corresponding text information risk parameter is true, that is, the text content is false.
- step S32 if it is false, it is determined that the target data has a text information risk.
- the target data if it is determined that the text content of the current target data is false, it is determined that the target data has a text information risk; if it is determined that the text content of the current target data is true, it is determined that the target data does not have a text information risk.
- the risk attribute of the target data can also be determined from the text content.
- the risk attribute of the target data is determined by judging whether the text content is true. At the same time, corresponding records are also needed to make the target report true and reliable.
- a third embodiment of the method for generating a configurable report of this application is proposed based on the first and second embodiments.
- the difference between the third embodiment of the configurable report generation method and the first and second embodiments of the configurable report generation method is that the configuration parameters include intent element information.
- the configurable report is generated
- the method further includes: step c, determining the associated scene in the intent element information, and the parameter value of the associated scene; step d, if a collection instruction triggered based on the intent element is detected, then corresponding from the intent element information
- the scene link corresponding to the associated scene is determined in the data source, and the target value of the scene link is determined, and the parameter value includes the target value; step e, according to the associated scene and the Target value and generate corresponding target report.
- the configuration parameters configured by the user include intent element information, it is necessary to obtain the corresponding associated scene and the target value corresponding to the associated scene, so that the subsequently generated target report is more detailed.
- the user's intention can be seen intuitively, which is convenient for follow-up business development.
- Step c Determine the associated scene in the intent element information and the parameter value of the associated scene.
- the configuration interface includes an intent element configuration area, as shown in Figure 3.
- the intent element configuration area includes report element name, data source, associated scene, parameter value, whether to output micro-expression risk, etc., which are collectively referred to as intent element information .
- the user can configure the parameters in the intent element configuration area.
- the generating device can determine the associated scene in the intent element information, and the parameter value in the associated scene, for example, ask the customer whether there is social insurance, and its corresponding The associated scenes of is with and without social insurance.
- the parameter value of the associated scene with social insurance is Yes, and the value of the associated scene without social insurance is No.
- Step d if a collection instruction triggered based on the intent element is detected, determine the scene link corresponding to the associated scene from the data source corresponding to the intent element information, and determine the target value of the scene link ,
- the parameter value includes the target value.
- the scene link corresponding to the associated scene is determined from the data source of the intent element information, so as to determine the target value of the associated scene.
- the data source of the intent element information It is intention recognition, which is different from the parameter extraction in the above-mentioned embodiment.
- the intention recognition is the complete scene link in the same scene. For example, in a human-machine multi-round dialogue, asking customers whether they have social insurance, there are two answering scenarios for this question. 1. There is social security; 2. There is no social security. In this scenario, there are two scenario links. In the final report, the feedback is whether the scenario link is hit, that is, "Yes" or "No".
- the target value of the scene link is further determined.
- the target value is one or more of the parameter values.
- the associated scene is social insurance and no social insurance.
- the parameter values are yes and no.
- Step e Generate a corresponding target report according to the associated scene and the target value.
- generating the target includes specifically taking the associated scene and the corresponding target value as the content of the target report, so that the generated target report can intuitively display the user's intention.
- step e includes: determining a value risk attribute corresponding to the target value; and generating a corresponding target report based on the associated scenario, the target value, and the value risk attribute.
- the configuration parameters configured by the user include intent element information
- the corresponding associated scene and the target value corresponding to the associated scene need to be obtained, so that the subsequently generated target report is more detailed.
- the user's intention can be seen intuitively, and the risk attribute needs to be judged on the associated scene and target value, so that the generated target report is true and reliable, and it is convenient for subsequent business development.
- Figure 5 is a schematic diagram of a target report generation interface, which summarizes target data with micro-expression risks and/or text information risks as abnormal points, and displays them in a preset format, such as micro-expression risks and text information risks. Or micro-expression risk and text information risk multi-window switch display, etc., displayed on the display interface.
- normal target data and abnormal target data can also be displayed together and displayed in different colors. For example, abnormal ones are marked in red, normal Are marked in green, etc.
- the embodiment of the present application also provides a configurable report generating device.
- FIG. 6 is a schematic diagram of the functional modules of the first embodiment of the configurable report generating apparatus according to the present application.
- the configurable report generating device includes: a configuration module 10, configured to display a configuration interface corresponding to the configuration instruction when a configuration instruction is detected, and receive configuration parameters input based on the configuration interface; extraction module 20. If a collection instruction triggered based on the configuration parameter is detected, determine the report element of the configuration parameter, and extract target data from the data source corresponding to the report element; the risk module 30 is used to determine The data risk attribute corresponding to the target data; the generating module 40 is configured to generate a target report for the target data based on the data risk attribute.
- the data risk attribute includes micro-expression risk
- the risk module is further used to: obtain micro-expression information of the target data, and determine whether the micro-expression information is a risk expression; if so, determine the target There is a risk of micro expressions in the data.
- the risk module is further configured to: obtain the micro-expression information of the target data, and determine the number of times the micro-expression information appears and the number of frames; based on the number of times and the number of frames, determine whether the micro-expression information is For the risk expression.
- the risk module is also used to: obtain micro-expression information of the target data, and select target micro-expression information from the micro-expression information; input the target micro-expression information into a micro-expression model to obtain Model output result; according to the model output result, determine whether the target micro-expression information is a risk expression.
- the data risk attribute includes text information risk
- the risk module is used to: extract the text content of the target data, compare the text content with pre-stored input information, and determine whether the text content is true ; If it is false, it is determined that the target data has text information risk.
- the generating module is further configured to: determine the associated scene in the intent element information, and the parameter value of the associated scene;
- the scene link corresponding to the associated scene is determined from the corresponding data source, and the target value of the scene link is determined, and the parameter value includes the target value; and the value is selected according to the associated scene and the target. Value to generate the corresponding target report.
- the generating module is further configured to: determine a value risk attribute corresponding to the target value; and generate a corresponding target report based on the associated scenario, the target value, and the value risk attribute.
- each module and unit in the above-mentioned configurable report generating device corresponds to each step in the above-mentioned embodiment of the above-mentioned configurable report generating method, and their functions and implementation processes will not be repeated here.
- the embodiment of the present application also provides a computer-readable storage medium.
- a configurable report generation program is stored on the computer-readable storage medium of the present application, and when the configurable report generation program is executed by a processor, the steps of the above-mentioned configurable report generation method are realized.
- the method implemented when the configurable report generation program is executed can refer to the various embodiments of the configurable report generation method of this application, which will not be repeated here.
- the storage medium involved in this application such as a computer-readable storage medium, may be non-volatile or volatile.
- the technical solution of this application essentially or the part that contributes to the existing technology can be embodied in the form of a software product, and the computer software product is stored in a storage medium (such as ROM/RAM) as described above. , Magnetic disks, optical disks), including several instructions to make a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) execute the methods described in the various embodiments of the present application.
- a terminal device which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
一种可配置化报告生成方法、装置、设备及可读存储介质，所述方法包括以下步骤：在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数（S10）；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据（S20）；确定所述目标数据对应的数据风险属性（S30）；基于所述数据风险属性，将所述目标数据生成目标报告（S40）。所述方法根据用户配置的配置参数，采集对应的目标数据，并对目标数据进行风险评估，最后将风险评估结果和目标数据生成报告，能直观的反应数据的真实性，提高数据转换成报告的效率，实现目标数据的可视化。
Description
本申请要求于2020年2月29日提交中国专利局、申请号为202010134978.2，发明名称为“可配置化报告生成方法、装置、设备及可读存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
本申请涉及数据可视化技术领域，尤其涉及一种可配置化报告生成方法、装置、设备及计算机可读存储介质。
人机对话是人工智能领域的一个子方向，具体指人可以通过自然语言与机器进行交互。人机对话解决的问题之一是任务驱动型多轮对话，任务驱动型多轮对话以完成任务为目的，例如帮助用户完成订购机票等任务，此类任务的完成需要通过机器和用户的多轮交互，从而收集用户的需求及相关信息，最终完成订购机票的任务。
然而，发明人意识到，收集用户的需求及相关信息，需要相关业务人员在人机对话结束后，在对话记录中进行收集，也没有较好的方法将对话流经的场景，用户的意图，槽位信息，对话内容等集成到统一的报告中，需要相关业务员自行查询对话记录，人工生成报告，效率较为低下，并且，收集到的信息很难确定真实，导致最后生成的报告真实性较低。
显然，现有技术中，还没有能帮助相关业务人员快速提取对话中的有效信息，并且将提取的有效信息生成较为真实的可视化报告，从而提高效率的方法。
本申请的主要目的在于提供一种可配置化报告生成方法、装置、设备及计算机可读存储介质，旨在解决现有报告生成方法效率较低且准确率较低的技术问题。
为实现上述目的，本申请提供一种可配置化报告生成方法，所述可配置化报告生成方法包括以下步骤：在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；确定所述目标数据对应的数据风险属性；基于所述数据风险属性，将所述目标数据生成目标报告。
此外，为实现上述目的，本申请还提供一种可配置化报告生成装置，所述可配置化报告生成装置包括：配置模块，用于在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；提取模块，用于若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；风险模块，用于确定所述目标数据对应的数据风险属性；生成模块，用于基于所述数据风险属性，将所述目标数据生成目标报告。
此外，为实现上述目的，本申请还提供一种可配置化报告生成设备，所述可配置化报告生成设备包括处理器、存储器、以及存储在所述存储器上并可被所述处理器执行的可配置化报告生成程序，其中所述可配置化报告生成程序被所述处理器执行时，实现以下方法：在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；确定所述目标数据对应的数据风险属性；基于所述数据风险属性，将所述目标数据生成目标报告。
此外，为实现上述目的，本申请还提供一种计算机可读存储介质，所述计算机可读存储介质上存储有可配置化报告生成程序，其中所述可配置化报告生成程序被处理器执行时，实现以下方法：在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；确定所述目标数据对应的数据风险属性；基于所述数据风险属性，将所述目标数据生成目标报告。
本申请根据用户配置的配置参数，采集对应的目标数据，并对目标数据进行风险评估，最后将风险评估结果和目标数据生成目标报告，能直观的反应数据的真实性，提高数据转换成报告的效率，实现数据的可视化。
图1为本申请实施例方案中涉及的可配置化报告生成设备的硬件结构示意图。
图2为本申请可配置化报告生成方法第一实施例的流程示意图。
图3为本申请可配置化报告生成方法一实施例中生成设备的配置界面示意图。
图4为本申请可配置化报告生成装置第二实施例的流程示意图。
图5为本申请一种目标报告生成界面示意图。
图6为本申请可配置化报告生成装置第一实施例的功能模块示意图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
本申请的技术方案可应用于人工智能、智慧城市、数字医疗、区块链和/或大数据技术领域，以实现数据可视化。可选的，本申请涉及的数据如数据风险属性、模型输出结果和/或报告等可存储于数据库中，或者可以存储于区块链中，比如通过区块链分布式存储，本申请不做限定。
本申请实施例涉及的可配置化报告生成方法主要应用于可配置化报告生成设备，该可配置化报告生成设备可以是PC、便携计算机、移动终端等具有显示和处理功能的设备。
参照图1，图1为本申请实施例方案中涉及的可配置化报告生成设备的硬件结构示意图。本申请实施例中，可配置化报告生成设备可以包括处理器1001(例如CPU)，通信总线1002，用户接口1003，网络接口1004，存储器1005。其中，通信总线1002用于实现这些组件之间的连接通信；用户接口1003可以包括显示屏(Display)、输入单元比如键盘(Keyboard)；网络接口1004可选的可以包括标准的有线接口、无线接口(如WI-FI接口)；存储器1005可以是高速RAM存储器，也可以是稳定的存储器(non-volatile memory)，例如磁盘存储器，存储器1005可选的还可以是独立于前述处理器1001的存储装置。
本领域技术人员可以理解，图1中示出的硬件结构并不构成对可配置化报告生成设备的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。
继续参照图1，图1中作为一种计算机可读存储介质的存储器1005可以包括操作系统、网络通信模块以及可配置化报告生成程序。
在图1中，网络通信模块主要用于连接服务器，与服务器进行数据通信；而处理器1001可以调用存储器1005中存储的可配置化报告生成程序，并执行本申请实施例提供的可配置化报告生成方法。
本申请实施例提供了一种可配置化报告生成方法。
参照图2，图2为本申请可配置化报告生成方法第一实施例的流程示意图。
本实施例中，所述可配置化报告生成方法包括以下步骤。
步骤S10，在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数。
步骤S20，若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据。
步骤S30，确定所述目标数据对应的数据风险属性。
步骤S40，基于所述数据风险属性，将所述目标数据生成目标报告。
本实施例的可配置化报告生成方法应用于可配置化报告生成设备，以下简称生成设备，生成设备可以为电脑等固定终端，也可以是手机、ipad等移动终端，生成设备可以与人机多轮对话设备连接，可根据采集指令，按照事先配置好的配置参数，获取存储在人机多轮对话设备中的人机对话记录，并按照一定的格式生成可视化报告，以便相关业务员能够直观的确定当前用户的需求，以及用户的真实性，便于后续业务的开展；生成设备也可以具备人机对话功能，在进行人机多轮对话任务时，存储人机对话记录，并根据采集指令，按照事先配置好的配置参数，获取已储存的人机对话记录生成报告，以便后续业务的开展。
本实施例根据用户配置的配置参数，采集对应的目标数据，并对目标数据进行风险评估，最后将风险评估结果和目标数据生成报告，能直观的反应数据的真实性，提高数据转换成报告的效率，实现目标数据的可视化。
以下将对各个步骤进行详细的说明。
步骤S10，在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数。
在本实施例中，用户可在生成设备进行参数的配置，具体生成设备在检测到用户下达的配置指令时，显示配置指令对应的配置界面，以供用户进行自主配置，在检测到用户配置完成后，即可接受用户输入的配置参数，如图3所示，在配置界面中，包括综合元素配置区域，提参元素配置区域和意图元素配置区域等，用户可在这些区域中进行参数的配置，具体的，每一个配置区域中包括报告元素名，数据来源，是否输出等，如综合元素配置区域中，报告元素名为微表情风险和文字信息异常，其分别对应的数据来源为微表情风险和文字对比风险，用户选择是否输出。而提参元素配置区域设置有新增按钮，以及编辑和删除按钮，用户可通过点击新增按钮触发新增报告元素，而编辑按钮则可以在已配置的报告元素中进行配置的修改，而删除按钮则可以删除已配置的报告元素等。
可以理解的，用户可不对提参元素区域进行配置，也即提参元素区域为空，而综合元素则只能配置是否输出。
进一步地，所述配置指令可由生成设备触发，也即，配置指令除了是用户下达的之外，还可以是在特定场景下触发的，也即在另一实施场景下，生成设备可实时监测人机对话是否结束，在检测到人机对话结束时，触发配置指令，从而显示配置指令对应的配置界面，以供用户进行配置参数的配置。
步骤S20，若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据。
在本实施例中，在用户配置好配置参数后，生成设备检测是否存在基于配置参数触发的采集指令，该采集指令可由用户下达，如用户在配置好配置参数后，点击相关确定按钮，即触发采集指令。
进一步地，所述采集指令也可由生成设备触发，也即，采集指令除了是用户下达的之外，还可以是在特定场景下触发的，也即在另一实施场景下，生成设备可实时监测人机对话是否结束，在检测到人机对话结束时，触发采集指令，生成设备从而去采集目标数据。
生成设备在检测到基于配置参数触发的采集指令时，确定配置参数的报告元素，并通过报告元素，在对应的数据来源中提取目标数据，具体的，在确定配置参数的报告元素后，确定报告元素的参数名，以及报告元素的数据来源，然后，在数据来源中，采集参数名对应的目标数据，也即，在该步骤中，用户事先建立报告元素与数据来源之间的关联关系，以及建立参数名与目标数据的关联关系，如映射关系，因此，生成设备可根据用户配置的报告元素，确定对应的数据来源，然后，再根据报告元素的参数名，在数据来源中采集目标数据。
如图3所示，用户事先在提参元素区域配置了报告元素名为姓名的报告元素，生成设备即可根据“姓名”确定对应的数据来源为参数提取，并且“姓名”在数据来源中的参数名为“name”，因此，生成设备在参数提取的数据来源中，采集与“name”相关的目标信息，如张三等。
需要说明的是，数据来源为人机对话过程中记录下来的数据，也即，在人机多轮对话时，会记录保存人机对话的内容，作为数据来源。
步骤S30，确定所述目标数据对应的数据风险属性。
在本实施例中，在采集到目标数据后，确定目标数据对应的数据风险属性，其中，数据风险属性包括微表情风险，也即确定目标数据是否存在微表情风险，具体的，步骤S30包括以下步骤。
步骤a，获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情。
在该步骤中，需要说明的是，在人机对话的过程中，除了记录目标数据，如用户的姓名，之外，还采集用户在当前场景下的微表情信息，如眨眼睛，皱眉等，并且，目标数据与微表情也是一一对应储存，如在询问用户姓名时，除了记录用户回答的答案之外，还采集用户在回答该问题时的微表情信息，因此，生成设备在采集到目标数据后，即可对应获取目标数据的微表情信息，然后，通过判断微表情信息来确定目标数据是否存在微表情风险。
具体的，将当前目标数据对应的微表情信息与事先设置有风险标识的微表情进行比对，确定当前的微表情信息是否为风险表情。
进一步地，步骤a包括以下步骤。
获取所述目标数据的微表情信息，并确定所述微表情信息出现的次数以及帧数。
在该步骤中，生成设备在获取目标数据的微表情信息时，确定微表情信息出现的次数以及帧数，也即，在目标数据对应的场景下，出现多少次微表情，以及微表情的帧数。如用户在回答姓名时，持续10帧，在这10帧中，出现眨眼3次，那么目标数据为姓名时，对应的微表情次数即为3，帧数即为10帧。
需要说明的是，帧数可以是目标数据对应的场景的帧数，也可以是从出现微表情开始到微表情结束为止的帧数，如用户在回答姓名时，持续10帧，但是，微表情是在第3帧开始出现，在第7帧之后没在出现，那么，微表情信息的帧数此时为4帧。
基于所述次数和帧数，确定所述微表情信息是否为风险表情。
在该步骤中，生成设备根据微表情出现的次数以及帧数，计算帧次比，即用微表情信息出现的次数除以帧数，得到帧次比，确定帧次比是否大于预设阈值，若是，则确定微表情信息为风险表情。
步骤b，若是，则确定所述目标数据存在微表情风险。
在该步骤中，若是确定微表情信息为风险表情，则确定目标数据存在微表情风险，说明当前目标数据可信度较低。
进一步地，目标数据对应的微表情信息至少有两个时，也即目标数据有多个微表情信息时，步骤a包括：获取所述目标数据的微表情信息，并从所述微表情信息中选取目标微表情信息；可以理解的，目标数据的微表情信息可能存在多个的情况，如用户在回答地址问题时，同时存在眨眼，抿嘴，微笑等微表情信息，则需要从中选取目标微表情信息，作为判断目标数据是否存在微表情风险的基础，具体选取方式可选择出现时长最长的微表情作为目标微表情，如眨眼出现10帧，抿嘴5帧，微笑2帧等，则选取眨眼作为目标微表情。将所述目标微表情信息输入微表情模型中，以得到模型输出结果；根据模型输出结果，判断所述目标微表情信息信息是否为风险表情。
在该步骤中，生成设备将获取到的目标微表情信息输入事先训练好的微表情模型中，通过微表情模型输出的模型输出结果，判断当前的目标微表情信息是否为风险表情。
其中，微表情模型是事先根据历史对话记录训练出来的，具体通过先验知识，将已知的历史微表情信息作为模型的输入，将已知的判断结果作为微表情的输出，通过模型训练，得到微表情模型。
在具体实施时，可将微表情信息出现的帧数作为微表情模型的输入，如抿嘴的帧数；或者，将微表情信息出现的次数作为模型的输入，如眨眼次数；或者将微表情信息出现的次数和帧数作为模型的输入，如皱眉次数和皱眉帧数。由于历史对话记录可知，因此，可将相应的判断结果作为模型的输出，来训练得到微表情模型。
在具体实施时，微表情信息有54种，一般会选取回答当前问题时频率最高的前五种作为目标微表情信息，再判断五种内是否有符合风险条件的微表情信息，如紧张，则判定此次微表情信息为风险表情。
步骤S40，基于所述数据风险属性，将所述目标数据生成目标报告。
在本实施例中，生成设备根据数据风险属性，将目标数据生成目标报告，具体的，将属于微表情风险的目标数据进行标记，并按照预设格式，将目标数据生成目标报告。
可以理解的，在目标数据有多个时，如姓名，身份证号，现居住地址等，生成设备需要依次确定目标数据的风险属性，最后，生成对应的目标报告，生成设备将存在微表情风险的目标数据归纳为异常点，并按照预设格式显示在显示界面，如存在微表情风险的目标数据与不存在微表情风险的目标数据一块显示，在一块显示时，以不同颜色显示区分，如存在微表情风险的以红色标注，不存在微表情风险的以绿色标注等；或者存在微表情风险的目标数据与不存在微表情风险的目标数据多窗口切换显示等。
本实施例在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；确定所述目标数据对应的数据风险属性；基于所述数据风险属性，将所述目标数据生成目标报告。本申请根据用户配置的配置参数，采集对应的目标数据，并对目标数据进行风险评估，最后将风险评估结果和目标数据生成报告，能直观的反应数据的真实性，提高数据转换成报告的效率，实现目标数据的可视化。
进一步地，基于第一实施例提出本申请可配置化报告生成方法的第二实施例。可配置化报告生成方法的第二实施例与可配置化报告生成方法的第一实施例的区别在于，所述数据风险属性包括文本信息风险，参照图4，步骤S30包括：步骤S31，提取所述目标数据的文本内容，将所述文本内容与预存进件信息进行比对，确定所述文本内容是否真实；步骤S32，若虚假，则确定所述目标数据存在文本信息风险。
本实施例中，数据风险属性包括文本信息风险，因此，在确定目标数据是否存在文本信息风险时，也即是确定目标数据的文本内容是否真实，使得后续生成的目标报告真实可靠。
以下将对各个步骤进行详细说明。
步骤S31，提取所述目标数据的文本内容，将所述文本内容与预存进件信息进行比对，确定所述文本内容是否真实。
在本实施例中，生成设备提取所述目标数据的文本内容，将所述文本内容与预存进件信息进行比对，确定所述文本内容是否真实。
其中，进件信息是指事先输入的真实性信息，也即，生成设备事先拥有进件信息，在得到目标数据后，提取目标数据的文本内容，将提取到的文本内容与进件信息进行比对，从而判断当前文本内容是否真实。
可以理解的，在人机对话过程中，一般是先确定用户的身份，然后才给用户拨打电话，进行相关信息的核实与记录，为方便理解，以一例子作为解释：如用户在提交贷款申请时，所提交的个人信息等信息即为进件信息，生成设备事先知道，然后在远程面审时，通过人机多轮对话，再采集用户的个人信息，也即上述所述的文本内容，再将两者进行比对，比如，用户在申请时提交的资料中显示年龄为30岁，在人机对话中，如果用户回答自己年龄为35岁，则确定两者比对结果不匹配，确定对应的文本信息风险参数为true，即文本内容虚假。
步骤S32，若虚假，则确定所述目标数据存在文本信息风险。
在本实施例中，若确定当前目标数据的文本内容虚假，则确定目标数据存在文本信息风险，若确定当前目标数据的文本内容真实，则确定目标数据不存在文本信息风险。
本实施例在评判目标数据的风险属性时，还可以从文本内容进行，通过判断文本内容是否真实来确定目标数据的风险属性，若是虚假，则确定目标数据存在文本信息风险，后续在生成目标报告时，也需要对应记载，使得目标报告真实可靠。
进一步地，基于第一、第二实施例提出本申请可配置化报告生成方法的第三实施例。可配置化报告生成方法的第三实施例与可配置化报告生成方法的第一、第二实施例的区别在于，所述配置参数包括意图元素信息，步骤S10之后，所述可配置化报告生成方法还包括：步骤c，确定意图元素信息中的关联场景，以及所述关联场景的参数取值；步骤d，若检测到基于所述意图元素触发的采集指令，则从所述意图元素信息对应的数据来源中确定所述关联场景对应的场景链路，并确定所述场景链路的目标取值，所述参数取值包括所述目标取值；步骤e，根据所述关联场景和所述目标取值，生成对应的目标报告。
本实施例在生成目标报告的过程中，若用户配置的配置参数中包含有意图元素信息，则需要获取对应的关联场景，以及关联场景对应的目标取值，使得后续生成的目标报告更加详实，能直观的看出用户的意图，方便后续业务的开展。
以下将对各个步骤进行详细说明。
步骤c，确定意图元素信息中的关联场景，以及所述关联场景的参数取值。
在本实施例中，配置界面包括意图元素配置区域，如图3，意图元素配置区域包括报告元素名，数据来源，关联场景，参数取值，是否输出微表情风险等，这些统称为意图元素信息，用户可在意图元素配置区域进行参数的配置，在配置完成后，生成设备即可确定意图元素信息中的关联场景，以及关联场景下的参数取值，比如，询问客户是否有社保，其对应的关联场景为有社保和无社保，同时，有社保的关联场景的参数取值为是，无社保的关联场景的取值为否。
步骤d，若检测到基于所述意图元素触发的采集指令，则从所述意图元素信息对应的数据来源中确定所述关联场景对应的场景链路，并确定所述场景链路的目标取值，所述参数取值包括所述目标取值。
在本实施例中，在检测到采集指令时，从意图元素信息的数据来源中确定关联场景对应的场景链路，从而确定关联场景的目标取值，需要说明的是，意图元素信息的数据来源是意图识别，与上述实施例参数提取不同，意图识别即为同一场景下的完整场景链路，如在人机多轮对话中，询问客户有无社保，该问题下设置有两个回答场景，1，有社保；2，无社保，那么在该场景下，共有两条场景链路，而最后生成的报告中，需要反馈的是场景链路是否命中，也即“是”或者“否”，因此，在确定关联场景对应的场景链路之后，进一步确定场景链路的目标取值，目标取值为参数取值的一种或者多种，如上述例子，关联场景为有社保和无社保，参数取值为是和否，在采集时，若命中“有社保”的场景，则将“社保”这个意图提取的参数对应到“是”，也即“是”是“有社保”的目标取值。
步骤e，根据所述关联场景和所述目标取值，生成对应的目标报告。
在本实施例中，根据最终确定的关联场景和目标取值，生成目标包括，具体将关联场景以及对应的目标取值作为目标报告的内容，使得生成的目标报告能直观的展现用户的意图。
需要说明的是，在某些场景下，不一定只有是和否两者互斥的结果，如工作经历，在人机多轮对话中，询问客户每一份工作的年限，客户可能第一份工作做了1年，第二份工作做了半年，那么该场景下，并不互斥，因此，具体场景需要具体分析。
进一步地，若是在相同场景下，客户的回答不一致，则取后面的答案作为最终值，如客户先说有社保，后来又说没交社保，那按生成设备预设的链路配置“社保”这个参数的最终值为“否”。
进一步地，步骤e包括：确定所述目标取值对应的取值风险属性；基于所述关联场景、所述目标取值和所述取值风险属性，生成对应的目标报告。
在本实施例中，在生成目标报告的过程中，也需要对关联场景和目标取值进行风险判断，如客户在回答有社保时，眼神闪烁，符合微表情风险判定，则在生成报告时，将有社保标红等，具体与上述实施例的判定过程类似，在此不再赘述。
本实施例在生成目标报告的过程中，若用户配置的配置参数中包含有意图元素信息，则需要获取对应的关联场景，以及关联场景对应的目标取值，使得后续生成的目标报告更加详实，能直观的看出用户的意图，并且需要对关联场景和目标取值进行风险属性判断，使得生成的目标报告真实可靠，方便后续业务的开展。
如图5所示为一种目标报告生成界面示意图，将存在微表情风险和/或文本信息风险的目标数据归纳为异常点，并按照预设格式，如微表情风险和文本信息风险一块显示，或者微表情风险和文本信息风险多窗口切换显示等，显示在显示界面，同时，也可将正常的目标数据和异常的目标数据一块显示，以不同颜色显示区分，如异常的以红色标注，正常的以绿色标注等。
此外，本申请实施例还提供一种可配置化报告生成装置。
参照图6，图6为本申请可配置化报告生成装置第一实施例的功能模块示意图。
本实施例中，所述可配置化报告生成装置包括：配置模块10，用于在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；提取模块20，用于若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；风险模块30，用于确定所述目标数据对应的数据风险属性；生成模块40，用于基于所述数据风险属性，将所述目标数据生成目标报告。
进一步地，所述数据风险属性包括微表情风险，所述风险模块还用于：获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情；若是，则确定所述目标数据存在微表情风险。
进一步地，所述风险模块还用于：获取所述目标数据的微表情信息，并确定所述微表情信息出现的次数以及帧数；基于所述次数和帧数，确定所述微表情信息是否为风险表情。
进一步地，所述风险模块还用于：获取所述目标数据的微表情信息，并从所述微表情信息中选取目标微表情信息；将所述目标微表情信息输入微表情模型中，以得到模型输出结果；根据模型输出结果，判断所述目标微表情信息信息是否为风险表情。
进一步地，所述数据风险属性包括文本信息风险，所述风险模块用于：提取所述目标数据的文本内容，将所述文本内容与预存进件信息进行比对，确定所述文本内容是否真实；若虚假，则确定所述目标数据存在文本信息风险。
进一步地，所述生成模块还用于：确定意图元素信息中的关联场景，以及所述关联场景的参数取值；若检测到基于所述意图元素触发的采集指令，则从所述意图元素信息对应的数据来源中确定所述关联场景对应的场景链路，并确定所述场景链路的目标取值，所述参数取值包括所述目标取值；根据所述关联场景和所述目标取值，生成对应的目标报告。
进一步地，所述生成模块还用于：确定所述目标取值对应的取值风险属性；基于所述关联场景、所述目标取值和所述取值风险属性，生成对应的目标报告。
其中，上述可配置化报告生成装置中各个模块和单元与上述可配置化报告生成方法实施例中各步骤相对应，其功能和实现过程在此处不再一一赘述。
此外，本申请实施例还提供一种计算机可读存储介质。
本申请计算机可读存储介质上存储有可配置化报告生成程序，其中所述可配置化报告生成程序被处理器执行时，实现如上述的可配置化报告生成方法的步骤。
其中，可配置化报告生成程序被执行时所实现的方法可参照本申请可配置化报告生成方法的各个实施例，此处不再赘述。
可选的，本申请涉及的存储介质如计算机可读存储介质可以是非易失性的，也可以是易失性的。
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者系统不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者系统所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、方法、物品或者系统中还存在另外的相同要素。
上述本申请实施例序号仅仅为了描述，不代表实施例的优劣。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在如上所述的一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，空调器，或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。
Claims (20)
- 一种可配置化报告生成方法，其中，所述可配置化报告生成方法包括以下步骤：在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；确定所述目标数据对应的数据风险属性；基于所述数据风险属性，将所述目标数据生成目标报告。
- 如权利要求1所述的可配置化报告生成方法，其中，所述数据风险属性包括微表情风险，所述确定所述目标数据对应的数据风险属性的步骤包括：获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情；若是，则确定所述目标数据存在微表情风险。
- 如权利要求2所述的可配置化报告生成方法，其中，所述获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情的步骤包括：获取所述目标数据的微表情信息，并确定所述微表情信息出现的次数以及帧数；基于所述次数和帧数，确定所述微表情信息是否为风险表情。
- 如权利要求2所述的可配置化报告生成方法，其中，所述微表情信息至少包括两个时，所述获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情的步骤包括：获取所述目标数据的微表情信息，并从所述微表情信息中选取目标微表情信息；将所述目标微表情信息输入微表情模型中，以得到模型输出结果；根据模型输出结果，判断所述目标微表情信息信息是否为风险表情。
- 如权利要求1所述的可配置化报告生成方法，其中，所述数据风险属性包括文本信息风险，所述确定所述目标数据对应的数据风险属性的步骤包括：提取所述目标数据的文本内容，将所述文本内容与预存进件信息进行比对，确定所述文本内容是否真实；若虚假，则确定所述目标数据存在文本信息风险。
- 如权利要求1所述的可配置化报告生成方法，其中，所述配置参数包括意图元素信息，所述在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数的步骤之后，所述可配置化报告生成方法还包括：确定意图元素信息中的关联场景，以及所述关联场景的参数取值；若检测到基于所述意图元素触发的采集指令，则从所述意图元素信息对应的数据来源中确定所述关联场景对应的场景链路，并确定所述场景链路的目标取值，所述参数取值包括所述目标取值；根据所述关联场景和所述目标取值，生成对应的目标报告。
- 如权利要求6所述的可配置化报告生成方法，其中，所述根据所述关联场景和所述目标取值，生成对应的目标报告的步骤包括：确定所述目标取值对应的取值风险属性；基于所述关联场景、所述目标取值和所述取值风险属性，生成对应的目标报告。
- 一种可配置化报告生成装置，其中，所述可配置化报告生成装置包括：配置模块，用于在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；提取模块，用于若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；风险模块，用于确定所述目标数据对应的数据风险属性；生成模块，用于基于所述数据风险属性，将所述目标数据生成目标报告。
- 一种可配置化报告生成设备，其中，所述可配置化报告生成设备包括处理器、存储器、以及存储在所述存储器上并可被所述处理器执行的可配置化报告生成程序，其中所述可配置化报告生成程序被所述处理器执行时，实现以下方法：在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；确定所述目标数据对应的数据风险属性；基于所述数据风险属性，将所述目标数据生成目标报告。
- 如权利要求9所述的可配置化报告生成设备，其中，所述数据风险属性包括微表情风险，所述确定所述目标数据对应的数据风险属性时，具体实现：获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情；若是，则确定所述目标数据存在微表情风险。
- 如权利要求10所述的可配置化报告生成设备，其中，所述获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情时，具体实现：获取所述目标数据的微表情信息，并确定所述微表情信息出现的次数以及帧数；基于所述次数和帧数，确定所述微表情信息是否为风险表情。
- 如权利要求10所述的可配置化报告生成设备，其中，所述微表情信息至少包括两个时，所述获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情时，具体实现：获取所述目标数据的微表情信息，并从所述微表情信息中选取目标微表情信息；将所述目标微表情信息输入微表情模型中，以得到模型输出结果；根据模型输出结果，判断所述目标微表情信息信息是否为风险表情。
- 如权利要求9所述的可配置化报告生成设备，其中，所述数据风险属性包括文本信息风险，所述确定所述目标数据对应的数据风险属性时，具体实现：提取所述目标数据的文本内容，将所述文本内容与预存进件信息进行比对，确定所述文本内容是否真实；若虚假，则确定所述目标数据存在文本信息风险。
- 如权利要求9所述的可配置化报告生成设备，其中，所述配置参数包括意图元素信息，所述在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数的步骤之后，所述可配置化报告生成程序被所述处理器执行时还用于实现：确定意图元素信息中的关联场景，以及所述关联场景的参数取值；若检测到基于所述意图元素触发的采集指令，则从所述意图元素信息对应的数据来源中确定所述关联场景对应的场景链路，并确定所述场景链路的目标取值，所述参数取值包括所述目标取值；根据所述关联场景和所述目标取值，生成对应的目标报告。
- 一种计算机可读存储介质，其中，所述计算机可读存储介质上存储有可配置化报告生成程序，其中所述可配置化报告生成程序被处理器执行时，实现以下方法：在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数；若检测到基于所述配置参数触发的采集指令，则确定所述配置参数的报告元素，并在所述报告元素对应的数据来源中提取目标数据；确定所述目标数据对应的数据风险属性；基于所述数据风险属性，将所述目标数据生成目标报告。
- 如权利要求15所述的计算机可读存储介质，其中，所述数据风险属性包括微表情风险，所述确定所述目标数据对应的数据风险属性时，具体实现：获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情；若是，则确定所述目标数据存在微表情风险。
- 如权利要求16所述的计算机可读存储介质，其中，所述获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情时，具体实现：获取所述目标数据的微表情信息，并确定所述微表情信息出现的次数以及帧数；基于所述次数和帧数，确定所述微表情信息是否为风险表情。
- 如权利要求16所述的计算机可读存储介质，其中，所述微表情信息至少包括两个时，所述获取所述目标数据的微表情信息，并判断所述微表情信息是否为风险表情时，具体实现：获取所述目标数据的微表情信息，并从所述微表情信息中选取目标微表情信息；将所述目标微表情信息输入微表情模型中，以得到模型输出结果；根据模型输出结果，判断所述目标微表情信息信息是否为风险表情。
- 如权利要求15所述的计算机可读存储介质，其中，所述数据风险属性包括文本信息风险，所述确定所述目标数据对应的数据风险属性时，具体实现：提取所述目标数据的文本内容，将所述文本内容与预存进件信息进行比对，确定所述文本内容是否真实；若虚假，则确定所述目标数据存在文本信息风险。
- 如权利要求15所述的计算机可读存储介质，其中，所述配置参数包括意图元素信息，所述在检测到配置指令时，显示配置指令对应的配置界面，并接收基于所述配置界面输入的配置参数的步骤之后，所述可配置化报告生成程序被处理器执行时还用于实现：确定意图元素信息中的关联场景，以及所述关联场景的参数取值；若检测到基于所述意图元素触发的采集指令，则从所述意图元素信息对应的数据来源中确定所述关联场景对应的场景链路，并确定所述场景链路的目标取值，所述参数取值包括所述目标取值；根据所述关联场景和所述目标取值，生成对应的目标报告。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010134978.2A CN111428454A (zh) | 2020-02-29 | 2020-02-29 | 可配置化报告生成方法、装置、设备及可读存储介质 |
| CN202010134978.2 | 2020-02-29 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2021169630A1 true WO2021169630A1 (zh) | 2021-09-02 |
Family
ID=71548010
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2021/071057 WO2021169630A1 (zh) | 2020-02-29 | 2021-01-11 | 可配置化报告生成方法、装置、设备及可读存储介质 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111428454A (zh) |
| WO (1) | WO2021169630A1 (zh) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113849381A (zh) * | 2021-09-10 | 2021-12-28 | 奇安信科技集团股份有限公司 | 通报预警方法、装置、计算机设备和存储介质 |
| CN114326423A (zh) * | 2021-12-21 | 2022-04-12 | 美的集团股份有限公司 | 一种场景化控制方案的自动生成方法及场景化控制设备 |
| CN117291165A (zh) * | 2023-11-24 | 2023-12-26 | 中国民航信息网络股份有限公司 | 一种数据报告生成方法、装置及设备 |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111428454A (zh) * | 2020-02-29 | 2020-07-17 | 深圳壹账通智能科技有限公司 | 可配置化报告生成方法、装置、设备及可读存储介质 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107845033A (zh) * | 2017-11-08 | 2018-03-27 | 上海壹账通金融科技有限公司 | 风控报告生成方法、装置、设备及计算机可读存储介质 |
| US20180322580A1 (en) * | 2017-05-03 | 2018-11-08 | Hui-Chi Hsiung | Intelligent Financial Management System |
| CN109785114A (zh) * | 2018-12-15 | 2019-05-21 | 深圳壹账通智能科技有限公司 | 用于贷款审核的信用数据展示方法、装置、设备及介质 |
| CN110727922A (zh) * | 2019-10-11 | 2020-01-24 | 集奥聚合(北京)人工智能科技有限公司 | 一种基于多维度数据流的反欺诈决策模型构建方法 |
| CN111428454A (zh) * | 2020-02-29 | 2020-07-17 | 深圳壹账通智能科技有限公司 | 可配置化报告生成方法、装置、设备及可读存储介质 |
-
2020
- 2020-02-29 CN CN202010134978.2A patent/CN111428454A/zh active Pending
-
2021
- 2021-01-11 WO PCT/CN2021/071057 patent/WO2021169630A1/zh active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180322580A1 (en) * | 2017-05-03 | 2018-11-08 | Hui-Chi Hsiung | Intelligent Financial Management System |
| CN107845033A (zh) * | 2017-11-08 | 2018-03-27 | 上海壹账通金融科技有限公司 | 风控报告生成方法、装置、设备及计算机可读存储介质 |
| CN109785114A (zh) * | 2018-12-15 | 2019-05-21 | 深圳壹账通智能科技有限公司 | 用于贷款审核的信用数据展示方法、装置、设备及介质 |
| CN110727922A (zh) * | 2019-10-11 | 2020-01-24 | 集奥聚合(北京)人工智能科技有限公司 | 一种基于多维度数据流的反欺诈决策模型构建方法 |
| CN111428454A (zh) * | 2020-02-29 | 2020-07-17 | 深圳壹账通智能科技有限公司 | 可配置化报告生成方法、装置、设备及可读存储介质 |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113849381A (zh) * | 2021-09-10 | 2021-12-28 | 奇安信科技集团股份有限公司 | 通报预警方法、装置、计算机设备和存储介质 |
| CN114326423A (zh) * | 2021-12-21 | 2022-04-12 | 美的集团股份有限公司 | 一种场景化控制方案的自动生成方法及场景化控制设备 |
| CN117291165A (zh) * | 2023-11-24 | 2023-12-26 | 中国民航信息网络股份有限公司 | 一种数据报告生成方法、装置及设备 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111428454A (zh) | 2020-07-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021169630A1 (zh) | 可配置化报告生成方法、装置、设备及可读存储介质 | |
| US10255484B2 (en) | Method and system for assessing facial skin health from a mobile selfie image | |
| CN104636501B (zh) | 一种多媒体网络人工即时翻译系统及方法 | |
| CN106131057B (zh) | 基于虚拟现实场景的认证和装置 | |
| CN111813399B (zh) | 基于机器学习的审核规则处理方法、装置及计算机设备 | |
| RU2018137829A (ru) | Способ, устройство и система отображения информации | |
| CN107123057A (zh) | 用户推荐方法及装置 | |
| CN107944378A (zh) | 一种自助服务的身份识别方法及自助服务系统 | |
| CN107256428A (zh) | 数据处理方法、数据处理装置、存储设备及网络设备 | |
| WO2020135334A1 (zh) | 电视应用主题切换方法、电视、可读存储介质及设备 | |
| CN104298983A (zh) | 具有分布式用户终端的舌苔图像获取与分析系统 | |
| CN106341380B (zh) | 一种用于对用户进行远程身份认证的方法、装置和系统 | |
| KR20200008838A (ko) | 가상 면접 제공 방법 및 이를 이용하는 장치 | |
| WO2020151170A1 (zh) | 一种岗位画像设置方法、岗位画像设置装置及终端设备 | |
| CN110874405A (zh) | 服务质检方法、装置、设备与计算机可读存储介质 | |
| JP6314579B2 (ja) | 通信装置、通信方法及びプログラム | |
| KR20200092166A (ko) | 감정을 인식하는 서버, 방법 및 컴퓨터 프로그램 | |
| CN106791237A (zh) | 用于电力公司的用户意见处理装置、系统及方法 | |
| JP2020160551A (ja) | 人事項目の分析支援装置、分析支援方法、プログラム、および記録媒体 | |
| CN115393077B (zh) | 基于贷款业务人机对话系统的数据处理方法及相关装置 | |
| US20220375468A1 (en) | System method and apparatus for combining words and behaviors | |
| US11449817B1 (en) | System and method for psychosocial technology protocol focused on the reduction for caregiver burnout and nursing home placement | |
| CN111147484B (zh) | 账号登录方法和装置 | |
| CN115906028A (zh) | 用户身份验证方法、装置和自助服务终端 | |
| JP2004295590A (ja) | 監査システム、監査方法、監査サーバ及び監査プログラム |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21760158 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 11/01/2023) |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 21760158 Country of ref document: EP Kind code of ref document: A1 |