BI data large screen display system based on voice control

Abstract

The present invention relates to the field of data interaction technology, and specifically discloses a BI data large-screen display system based on voice control, including: a voice input instruction extraction module extracts voice input instructions in real time; a supplementary input instruction generation module generates supplementary input voice feedback instructions; an instruction supplement issuance and reception module issues supplementary input voice feedback instructions, and at the same time, obtains the latest extracted voice input instructions in sound data within a preset decibel value range received in real time around the BI data large screen as the latest voice input instruction; a data display instruction parsing module parses the voice input instruction and the latest voice input instruction to obtain a data display instruction; a data operation and display module is used to operate and display data in an associated database of the BI data large screen based on the data display instruction to obtain a BI data large screen display result; it is used to enable users to quickly and accurately control the BI data large screen to display corresponding data through voice instructions, thereby improving its convenience.

Description

Technical Field

[0001] The present invention relates to the field of data interaction technology, and in particular to a BI data large-screen display system based on voice control. Background Art

[0002] At present, BI data (Business Intelligence Data) refers to data used to support corporate decision-making, usually including various information of the enterprise, such as sales, inventory levels, production quantities, employee performance, etc. These data can come from various systems and applications within the enterprise, or from external data providers or public data sources. BI data is usually collected, integrated, cleaned, stored and analyzed to help enterprise managers make better decisions. Through BI tools and techniques, data analysts can mine trends, correlations and patterns in data, discover hidden information and insights, and thus improve the operation, planning and strategy formulation of the enterprise. BI data can help enterprises achieve more efficient resource utilization, reduce costs, improve competitiveness and customer satisfaction, etc. BI data large screen display system is a technology that visualizes BI data in the form of charts, tables, etc. Its purpose is to make the data more intuitively presented to people and facilitate enterprise managers to make decisions. At present, there are many mature BI data large screen display system technologies and solutions on the market, which can be mainly divided into the following categories:

[0003] 1. Traditional BI large-screen display system: This type of system usually uses desktop application software, such as Power BI, Tableau, Sisense, etc., and connects to the database or other data sources through the API interface to quickly generate and display reports. This type of system has strong data processing capabilities and real-time performance, and is suitable for data query scenarios that require fast response.

[0004] 2. Web BI large screen display system: Web BI large screen display system is a technology that embeds BI functions into web applications. Users can access the system through a browser to analyze and view data. This type of system has the advantages of good cross-platform and simple deployment, and is suitable for access on mobile devices.

[0005] 3. Mobile BI large-screen display system: A BI large-screen display system for mobile devices such as smartphones and tablets. Its main features are simple interface and convenient operation. Common mobile BI large-screen display systems include Amap and Baidu Index.

[0006] 4. Large-screen display system: This type of system mainly includes hardware devices such as projectors and LED screens, as well as software content creation tools related to these devices. Users can display the generated BI data charts, reports and other content on the screen, supporting functions such as multi-screen splicing and large-screen scaling.

[0007] 5. Self-service BI large-screen display system: This type of system provides enterprises with the ability to self-service data exploration and report generation. Users can create their own BI reports through drag-and-drop operations and customize BI models and indicators at any time. Common self-service BI large-screen display systems include Microsoft Power BIDesktop, Alibaba Cloud Data Plus, etc.

[0008] In short, there are various BI data large screen display system technologies and solutions on the market, and you can choose the appropriate solution according to actual needs and corporate budget. In actual applications, you can also combine a variety of technical means, such as cloud computing, big data, artificial intelligence, etc., to achieve more efficient and intelligent data analysis and management.

[0009] Although speech recognition technology has been widely used in smart homes, smart cars and other fields, it is relatively less used in BI data large-screen display systems. Traditional BI data large-screen display systems are mainly operated through input devices such as keyboards and mice, and mainly rely on visual interfaces and manual operations, which to a certain extent limits the efficiency of data analysis and display. This not only limits the convenience of user operation, but also increases the learning cost. At the same time, with the continuous development of big data technology, enterprises' demand for data analysis is increasing, and traditional operation methods can no longer meet users' needs for efficient and convenient data analysis.

[0010] Therefore, the present invention proposes a BI data large-screen display system based on voice control. Summary of the invention

[0011] The present invention provides a BI data large-screen display system based on voice control, which is used to issue corresponding supplementary input voice feedback instructions through analysis of the current missing elements of the voice input instructions input by the controller, so as to realize continuous guided supplementation of the controller's voice input instructions until the data display instructions containing all necessary elements are obtained, so that users can quickly and accurately control the BI data large-screen to display corresponding data through voice instructions, thereby ensuring the accuracy and efficiency of BI data analysis and display while improving the user's operation convenience.

[0012] The present invention provides a BI data large-screen display system based on voice control, comprising:

[0013] The voice input command extraction module is used to extract the voice input command in real time from the sound data within a preset decibel value range received in real time around the BI data screen;

[0014] A supplementary input instruction generation module is used to analyze the current missing elements of the voice input instruction based on the voice input instruction when the completeness of the voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing elements;

[0015] The command supplement issuing and receiving module is used to issue supplementary input voice feedback commands and, at the same time, obtain the latest voice input command extracted from the sound data within a preset decibel value range received in real time around the BI data screen as the latest voice input command;

[0016] A data display instruction parsing module, used for parsing the voice input instruction and the latest voice input instruction to obtain the data display instruction;

[0017] The data operation and display module is used to operate and display the data in the associated database of the BI data screen based on the data display instructions to obtain the BI data screen display results.

[0018] Preferably, the supplementary input instruction generation module includes:

[0019] The completeness judgment submodule is used to judge whether the completeness of the voice input instruction reaches a preset completeness index and obtain a completeness judgment result;

[0020] The supplementary input instruction generation submodule is used to analyze the current missing elements of the voice input instruction if the completeness judgment result is that the completeness of the voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing elements.

[0021] Preferably, the completeness judgment submodule includes:

[0022] An instruction element analysis and prediction unit, used to analyze the complete necessary elements and currently covered elements of the voice input instruction;

[0023] A first completeness determination unit, configured to determine that the completeness of the voice input instruction reaches a preset completeness index when determining that the currently covered elements of the voice input instruction cover the corresponding complete necessary elements, as a completeness determination result;

[0024] The second completeness determination unit is used to determine that the completeness of the voice input instruction fails to meet a preset completeness index when the currently covered elements of the voice input instruction cannot cover the corresponding complete necessary elements, as a completeness determination result.

[0025] Preferably, the instruction element analysis and prediction unit comprises:

[0026] The speech escape subunit is used to perform text escape on the speech input instruction, obtain the escaped text of the speech input instruction, and perform element analysis on the escaped text to determine the current covered elements of the speech input instruction;

[0027] A reference element determination subunit, configured to determine reference necessary elements of the voice input instruction based on the currently covered elements and the instruction necessary elements of the voice input instruction;

[0028] The prefix and suffix element inference subunit is used to determine whether there is a prefix and suffix element category for the reference necessary elements, obtain the prefix and suffix element inference result, and determine the complete necessary elements of the voice input instruction based on the prefix and suffix element inference result and the reference necessary elements.

[0029] Preferably, the reference element determination subunit comprises:

[0030] a first reference element determination terminal, configured to, when the currently covered elements of the voice input instruction cover the instruction necessary elements, regard all the instruction necessary elements contained in the voice input instruction as reference necessary elements of the voice input instruction;

[0031] The second reference element determination end is used to make possible item inferences based on all the necessary elements of the instruction currently contained in the voice input instruction when the currently covered elements of the voice input instruction do not cover the necessary elements of the instruction, determine at least one possible item of the inferred necessary element, and summarize each possible item of the inferred necessary element with some target single elements of the voice input instruction, so as to obtain multiple element groups containing each possible item of the inferred necessary element as reference necessary elements of the voice input instruction.

[0032] Preferably, the second reference element determination terminal includes:

[0033] The single element determination sub-terminal is used to determine each single element in the current covered elements of the voice input instruction as a target single element when the current covered elements of the voice input instruction do not cover the necessary elements of the instruction;

[0034] The single element pre-selection sub-end is used to screen out all single elements whose single instruction co-occurrence degree with each target single element exceeds the single instruction co-occurrence degree threshold in the instruction parsing element library, and regard them as all pre-selected single elements of each target single element;

[0035] The single element reselection sub-terminal is used to treat all preselected single elements of each target single element, whose pre-assigned necessity degree is not less than the pre-assigned necessity degree threshold, as all reselected single elements of each target single element;

[0036] The single element final selection sub-terminal is used to perform cluster analysis on all re-selected single elements of all target single elements, obtain at least one possible item of inferred necessary elements, and summarize each possible item of inferred necessary elements with some target single elements of the voice input instruction respectively, to obtain multiple element groups containing each possible item of inferred necessary elements as reference necessary elements for the voice input instruction.

[0037] Preferably, the single element final selection sub-end performs cluster analysis on all reselected single elements of all target single elements to obtain at least one possible item of the inferred necessary element, and summarizes each possible item of the inferred necessary element and part of the target single elements of the voice input instruction respectively, and obtains a plurality of element groups including each possible item of the inferred necessary element as reference necessary elements of the voice input instruction, including:

[0038] Performing a preset number of arbitrary cluster divisions on all target single elements and all reselected single elements of all target single elements to obtain multiple element clusters in each division process;

[0039] Count the single instruction co-occurrence of all single elements in each element cluster in each division process in the instruction parsing element library, and take the total number of all single elements contained in each element cluster as the co-occurrence quantified multiple of each element cluster;

[0040] The product of the single instruction co-occurrence degree and the quantized multiple of the co-occurrence of each element cluster in each division process is regarded as the co-occurrence multiple of each element cluster in each division process;

[0041] The average of the co-occurrence multiples of all element clusters in each division process whose co-occurrence multiples are not less than the co-occurrence multiple threshold is taken as the estimated effect value of each division process;

[0042] In each division process where the estimated effect value exceeds the estimated effect threshold, all single elements except all target single elements in all element clusters whose co-occurrence multiples are not less than the co-occurrence multiples threshold are summarized as a possible item of the estimated necessary elements;

[0043] Aggregate each possible item of the inferred necessary element and all the target single elements contained in all element clusters whose co-occurrence multiples are not less than the co-occurrence multiple threshold in the corresponding division process as a feature group;

[0044] All element groups are considered as reference necessary elements for voice input commands.

[0045] Preferably, the supplementary input instruction generation submodule includes:

[0046] a missing element analysis unit, configured to, if the completeness judgment result is that the completeness of the voice input instruction does not meet a preset completeness index, take some elements included in the complete necessary elements of the voice input instruction and not included in the currently covered elements as the current missing elements of the voice input instruction;

[0047] The feedback instruction generating unit is used to generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing element.

[0048] Preferably, it also includes:

[0049] The data display instruction parsing submodule is used to parse the voice input instruction to obtain the data display instruction when the completeness of the voice input instruction reaches a preset completeness index.

[0050] Preferably, the data display instruction parsing module includes:

[0051] A data display instruction direct parsing submodule is used to parse the voice input instruction and the latest voice input instruction to obtain the data display instruction if it is determined based on the voice input instruction and the latest voice input instruction that the latest voice input instruction reaches a preset completeness index;

[0052] The data display instruction indirect parsing submodule is used to analyze the latest missing elements of the latest voice input instruction if it is determined based on the voice input instruction and the latest voice input instruction that the latest voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the latest voice input instruction and the corresponding latest missing elements, and issue the supplementary input voice feedback instruction until the latest received voice input instruction is determined to meet the preset completeness index, then the data display instruction is parsed from all voice control instructions received in this data display control process.

[0053] The beneficial effects of the present invention compared to the prior art are as follows: by analyzing the currently missing elements of the voice input instructions input by the controller, corresponding supplementary input voice feedback instructions are issued to achieve continuous guided supplementation of the controller's voice input instructions until a data display instruction containing all necessary elements is obtained, so that users can quickly and accurately control the BI data large-screen display of corresponding data through voice instructions, thereby ensuring the accuracy and efficiency of BI data analysis and display while improving the user's operation convenience.

[0054] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in this application document.

[0055] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0057] Figure 1 Schematic diagram of internal functional modules of a BI data large-screen display system based on voice control in an embodiment of the present invention;

[0058] Figure 2 Schematic diagram of internal functional submodules of a supplementary input instruction generation module in an embodiment of the present invention;

[0059] Figure 3 Schematic diagram of internal functional submodules of a data display instruction parsing module in an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0061] Embodiment 1:

[0062] The present invention provides a BI data large screen display system based on voice control, referring to Figure 1 ,include:

[0063] The voice input command extraction module is used to extract the voice input command in real time from the sound data within a preset decibel value range received in real time around the BI data screen;

[0064] A supplementary input instruction generation module is used to analyze the current missing elements of the voice input instruction based on the voice input instruction when the completeness of the voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing elements;

[0065] The command supplement issuing and receiving module is used to issue supplementary input voice feedback commands and, at the same time, obtain the latest voice input command extracted from the sound data within a preset decibel value range received in real time around the BI data screen as the latest voice input command;

[0066] A data display instruction parsing module, used for parsing the voice input instruction and the latest voice input instruction to obtain the data display instruction;

[0067] The data operation and display module is used to operate and display the data in the associated database of the BI data screen based on the data display instructions to obtain the BI data screen display results.

[0068] In this embodiment, the BI data large screen is a display screen for displaying BI data.

[0069] In this embodiment, the sound data within the preset decibel value range is the sound data whose decibel value is within the preset decibel value range, wherein the preset decibel value range is a preset decibel value range within which sound data that can obtain voice input instructions is allowed.

[0070] In this embodiment, the voice input instruction is extracted from the sound data and represents an instruction obtained after escaping the sound data, that is, a control instruction input by the controller based on the voice input method.

[0071] In this embodiment, voice input instructions are extracted in real time from sound data within a preset decibel range received in real time around the BI data screen, which can be achieved through existing voice recognition models such as hidden Markov models.

[0072] In this embodiment, the completeness of the voice input instruction indicates the completeness of the types of elements contained in the voice input instruction.

[0073] In this embodiment, the preset completeness index is a preset voice input instruction that can be used as a corresponding completeness value that needs to be achieved when parsing the data parsing instruction.

[0074] In this embodiment, the current missing element of the voice input instruction is the element currently missing in the voice input instruction. The elements included in the instruction include, for example: operation object, operation type, or the operation object and operation type of the operation step that must be performed before and / or after this operation step.

[0075] In this embodiment, the supplementary input voice feedback instruction is a prompt instruction including the voice input instruction and the current missing element, and is used to prompt the voice inputter to complete the current missing element.

[0076] In this embodiment, the data display instruction is parsed from the voice input instruction and the latest voice input instruction, including:

[0077] The voice input command and the latest voice input command are escaped and merged into text, and then the existing semantic recognition model is used (for example, a neural network is used to model the grammatical structure of natural language sentences and infer semantics from them. This model usually uses a sequence to sequence (Seq2Seq) framework to map the input text sequence to the output text sequence, where each word is encoded as a latent state vector. Commonly used models include LSTM, GRU, and Transformer, etc.) to parse the data display instructions in the merged text.

[0078] In this embodiment, the data display instruction is an instruction that represents the control target of the controller and is used to control the BI data screen to limit and clarify the operation and display method of the data in the associated database.

[0079] In this embodiment, the associated database of the BI data screen is a database that needs to be retrieved and integrated as required and then displayed on the BI data screen in a visual form.

[0080] In this embodiment, the data in the associated database of the BI data screen is operated and displayed based on the data display instruction to obtain the BI data screen display result, including:

[0081] Based on the data display instructions, the data in the associated database of the BI data screen is operated, and then displayed on the BI data screen in the required display format to obtain the BI data screen display result.

[0082] The beneficial effects of the above technology are: by analyzing the current missing elements of the voice input instructions input by the controller, corresponding supplementary input voice feedback instructions are issued to achieve continuous guided supplementation of the controller's voice input instructions until a data display instruction containing all necessary elements is obtained, enabling users to quickly and accurately control the BI data large-screen to display corresponding data through voice instructions, thereby ensuring the accuracy and efficiency of BI data analysis and display while improving user convenience.

[0083] Embodiment 2:

[0084] On the basis of Example 1, an input instruction generation module is supplemented, referring to Figure 2 ,include:

[0085] The completeness judgment submodule is used to judge whether the completeness of the voice input instruction reaches a preset completeness index and obtain a completeness judgment result;

[0086] The supplementary input instruction generation submodule is used to analyze the current missing elements of the voice input instruction if the completeness judgment result is that the completeness of the voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing elements.

[0087] In this embodiment, the completeness determination result includes two possibilities: the completeness of the voice input instruction reaches a preset completeness index, and the completeness of the voice input instruction does not reach the preset completeness index.

[0088] The beneficial effects of the above technology are: by judging the completeness of the voice input instruction, further judging the situation where a supplementary input voice feedback instruction needs to be generated, and generating the corresponding supplementary input voice feedback instruction by analyzing the current missing elements of the voice input instruction.

[0089] Embodiment 3:

[0090] Based on Example 2, the integrity judgment submodule includes:

[0091] An instruction element analysis and prediction unit, used to analyze the complete necessary elements and currently covered elements of the voice input instruction;

[0092] A first completeness determination unit, configured to determine that the completeness of the voice input instruction reaches a preset completeness index when determining that the currently covered elements of the voice input instruction cover the corresponding complete necessary elements, as a completeness determination result;

[0093] The second completeness determination unit is used to determine that the completeness of the voice input instruction fails to meet a preset completeness index when the currently covered elements of the voice input instruction cannot cover the corresponding complete necessary elements, as a completeness determination result.

[0094] In this embodiment, the complete necessary elements of the voice input instruction are the elements that the voice input instruction must contain.

[0095] In this embodiment, the currently covered elements of the voice input instruction are the elements currently included in the voice input instruction.

[0096] In this embodiment, the currently covered elements covering the corresponding complete necessary elements means that the elements currently included in the voice input instruction include the elements that the voice input instruction must include.

[0097] The beneficial effect of the above technology is: by analyzing and comparing the complete necessary elements and currently covered elements of the voice input instruction, it is determined whether the completeness of the voice input instruction can reach the preset completeness index, that is, the completeness judgment result is obtained.

[0098] Embodiment 4:

[0099] Based on the third embodiment, the instruction element analysis and prediction unit includes:

[0100] The speech escape subunit is used to perform text escape on the speech input instruction, obtain the escaped text of the speech input instruction, and perform element analysis on the escaped text to determine the current covered elements of the speech input instruction;

[0101] A reference element determination subunit, configured to determine reference necessary elements of the voice input instruction based on the currently covered elements and the instruction necessary elements of the voice input instruction;

[0102] The prefix and suffix element inference subunit is used to determine whether there is a prefix and suffix element category for the reference necessary elements, obtain the prefix and suffix element inference result, and determine the complete necessary elements of the voice input instruction based on the prefix and suffix element inference result and the reference necessary elements.

[0103] In this embodiment, text translation is to convert the sound data into text with the same meaning, which can be accomplished by using the automatic speech recognition technology in the prior art, for example.

[0104] In this embodiment, element analysis is a process of determining elements included in the escaped text, by comparing the escaped text with a preset element library, and screening out elements in the element library contained in the escaped text as the current covered elements of the voice input instruction, wherein the element library is a pre-prepared library covering all elements included in all possible voice input instructions.

[0105] In this embodiment, the necessary elements of the instruction are the data operation object (which can be expressed as at least two single elements, such as: including the data table name or data source, and the data range to be operated in the corresponding data table or data source) and the data operation type.

[0106] In this embodiment, determining whether there is a prefix and suffix element category for the reference necessary element includes: searching a preset necessary prefix and suffix category library to determine whether there is a prefix and / or suffix element category for the reference necessary element recorded in the preset necessary prefix and suffix category library;

[0107] The necessary prefix and suffix category library records the prefix and / or suffix element category corresponding to each element, and may also record that the element has no corresponding prefix and / or suffix element category;

[0108] The prefix (suffix) element category is an element used to limit the operation object or operation type of the preceding (successor) step that must be executed for the operation step represented by a single element used to limit the operation object and operation type.

[0109] In this embodiment, the prefix and suffix element inference results include the following possibilities: the reference necessary element does not have a prefix element category and a suffix element category, the reference necessary element does have a prefix element category, and the reference necessary element does have a suffix element category.

[0110] In this embodiment, the complete necessary elements of the voice input instruction are determined based on the prefix and suffix element inference results and the reference necessary elements, namely:

[0111] The prefix and / or suffix elements and reference necessary elements that may be contained in the prefix and suffix element inference results are summarized and regarded as the complete necessary elements of the voice input instruction.

[0112] The beneficial effect of the above technology is: by analyzing the elements of the escaped text of the voice input instruction, its reference necessary elements are inferred, and they are summarized with the inferred prefix and / or suffix elements that may be included, thereby achieving accurate analysis of the complete necessary elements in the voice input instruction, wherein the current covered elements of the voice input instruction are also determined.

[0113] Embodiment 5:

[0114] On the basis of Example 4, the reference element determines the subunit, including:

[0115] a first reference element determination terminal, configured to, when the currently covered elements of the voice input instruction cover the instruction necessary elements, regard all the instruction necessary elements contained in the voice input instruction as reference necessary elements of the voice input instruction;

[0116] The second reference element determination end is used to make possible item inferences based on all the necessary elements of the instruction currently contained in the voice input instruction when the currently covered elements of the voice input instruction do not cover the necessary elements of the instruction, determine at least one possible item of the inferred necessary element, and summarize each possible item of the inferred necessary element with some target single elements of the voice input instruction, so as to obtain multiple element groups containing each possible item of the inferred necessary element as reference necessary elements of the voice input instruction.

[0117] The beneficial effects of the above technology are: based on whether the currently covered elements of the voice input instruction cover the necessary elements of the instruction, the reference necessary elements of the voice input instruction are analyzed according to different situations, and when the currently covered elements of the voice input instruction do not cover the necessary elements of the instruction, based on all the necessary elements of the instruction currently contained in the voice input instruction, reasonable and accurate inference of the unknown elements in the reference necessary elements of the voice input instruction is achieved.

[0118] Embodiment 6:

[0119] On the basis of Embodiment 5, the second reference element determining end includes:

[0120] The single element determination sub-terminal is used to determine each single element in the current covered elements of the voice input instruction as a target single element when the current covered elements of the voice input instruction do not cover the necessary elements of the instruction;

[0121] The single element pre-selection sub-end is used to screen out all single elements whose single instruction co-occurrence degree with each target single element exceeds the single instruction co-occurrence degree threshold in the instruction parsing element library, and regard them as all pre-selected single elements of each target single element;

[0122] The single element reselection sub-terminal is used to treat all preselected single elements of each target single element, whose pre-assigned necessity degree is not less than the pre-assigned necessity degree threshold, as all reselected single elements of each target single element;

[0123] The single element final selection sub-terminal is used to perform cluster analysis on all re-selected single elements of all target single elements, obtain at least one possible item of inferred necessary elements, and summarize each possible item of inferred necessary elements with some target single elements of the voice input instruction respectively, to obtain multiple element groups containing each possible item of inferred necessary elements as reference necessary elements for the voice input instruction.

[0124] In this embodiment, each single element in the elements currently covered by the voice input instruction is determined, including:

[0125] Treat each limiting condition in the element as a single element, for example, the data table name or data source in the data operation object (currently covering the element) is treated as a single element, and the data range being operated in the corresponding data table or data source is also treated as a single element.

[0126] In this embodiment, the instruction parsing element library is a library containing a large number of pre-prepared escaped texts of voice input instructions and the elements contained therein.

[0127] In this embodiment, the method for determining the single instruction co-occurrence degree of each single element in the instruction parsing element library and each target single element is as follows:

[0128] The ratio of the number of times each single element in the instruction parsing element library and each target single element appear simultaneously in the escaped text of a single voice input instruction in the instruction parsing element library to the total number of escaped texts of all voice input instructions in the instruction parsing element library is taken as the single instruction co-occurrence degree of each single element and each target single element.

[0129] In this embodiment, the single instruction co-occurrence threshold represents the frequency of co-occurrence of two single elements in the escaped text of a single voice input instruction in the instruction parsing element library.

[0130] In this embodiment, the pre-assigned necessity degree of the pre-selected list element is a numerical value assigned in advance to the pre-assigned list element, indicating its necessity degree in the data presentation instruction.

[0131] In this embodiment, the pre-assignment necessity threshold is a value that the pre-assignment necessity of the pre-selection list element must not be less than when the pre-assignment list element is regarded as a re-selection list element.

[0132] The beneficial effect of the above technology is: by analyzing the single instruction co-occurrence necessity of each single element in the current covered elements of the voice input instruction and the single element in the instruction parsing element library, comparing the pre-assigned necessity of its elements, and further clustering analysis, it is possible to screen and infer the possible items of the inferred necessary elements of the voice input instruction in the instruction parsing element library, and then accurately infer the reference necessary elements of the voice input instruction.

[0133] Embodiment 7:

[0134] On the basis of Example 6, the single element final selection sub-end performs cluster analysis on all reselected single elements of all target single elements to obtain at least one possible item of the inferred necessary element, and summarizes each possible item of the inferred necessary element and part of the target single elements of the voice input instruction respectively, and obtains a plurality of element groups including each possible item of the inferred necessary element as a reference necessary element of the voice input instruction, including:

[0135] Performing a preset number of arbitrary cluster divisions on all target single elements and all reselected single elements of all target single elements to obtain multiple element clusters in each division process;

[0136] Count the single instruction co-occurrence of all single elements in each element cluster in each division process in the instruction parsing element library, and take the total number of all single elements contained in each element cluster as the co-occurrence quantified multiple of each element cluster;

[0137] The product of the single instruction co-occurrence degree and the quantized multiple of the co-occurrence of each element cluster in each division process is regarded as the co-occurrence multiple of each element cluster in each division process;

[0138] The average of the co-occurrence multiples of all element clusters in each division process whose co-occurrence multiples are not less than the co-occurrence multiple threshold is taken as the estimated effect value of each division process;

[0139] In each division process where the estimated effect value exceeds the estimated effect threshold, all single elements except all target single elements in all element clusters whose co-occurrence multiples are not less than the co-occurrence multiples threshold are summarized as a possible item of the estimated necessary elements;

[0140] Aggregate each possible item of the inferred necessary element and all the target single elements contained in all element clusters whose co-occurrence multiples are not less than the co-occurrence multiple threshold in the corresponding division process as a feature group;

[0141] All element groups are considered as reference necessary elements for voice input commands.

[0142] In this embodiment, the element cluster is a cluster including a plurality of re-selection elements, and there are no overlapping re-selection elements among all element clusters obtained in each cluster division process.

[0143] In this embodiment, the method for determining the single instruction co-occurrence degree of all single elements in the instruction parsing element library is:

[0144] The ratio of the number of times all single elements appear simultaneously in the escaped text of a single voice input instruction in the instruction parsing element library to the total number of escaped texts of all voice input instructions in the instruction parsing element library is regarded as the single instruction co-occurrence degree of all single elements in the instruction parsing element library.

[0145] In this embodiment, the co-occurrence multiple of the element cluster represents the relative frequency of all single elements included in the element cluster appearing simultaneously in the escaped text of a single voice input instruction in the instruction parsing element library.

[0146] In this embodiment, the co-occurrence multiple threshold is a value that the co-occurrence multiple of the element cluster should not be less than when the co-occurrence multiple of the element cluster is used to determine the estimated effect value of the corresponding sub-division process.

[0147] In this embodiment, the estimated effect value of the division process indicates the degree to which all element clusters whose co-occurrence multiples are not less than the co-occurrence multiple threshold in the corresponding division process can be regarded as possible items of estimated necessary elements.

[0148] In this embodiment, the inference effect threshold is a value that the corresponding inference effect value must exceed when all element clusters whose co-occurrence multiples are not less than the co-occurrence multiple threshold in the preset corresponding sub-division process can be regarded as possible items of inferred necessary elements.

[0149] The beneficial effect of the above technology is: through multiple clustering of all target single elements and all re-selected single elements of all target single elements, and using the calculated co-occurrence multiples of each element cluster in each division process as the screening basis, the element groups with a high degree of co-occurrence in the instruction parsing element library are reasonably screened out as the necessary reference elements for voice input instructions.

[0150] Embodiment 8:

[0151] On the basis of Example 3, the input instruction generation submodule is supplemented, including:

[0152] a missing element analysis unit, configured to, if the completeness judgment result is that the completeness of the voice input instruction does not meet a preset completeness index, take some elements included in the complete necessary elements of the voice input instruction and not included in the currently covered elements as the current missing elements of the voice input instruction;

[0153] The feedback instruction generating unit is used to generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing element.

[0154] The beneficial effect of the above technology is that when the completeness judgment result is that the completeness of the voice input instruction does not meet the preset completeness index, the voice input instruction and the current missing elements are determined, and the supplementary input voice feedback instruction is further generated.

[0155] Embodiment 9:

[0156] On the basis of Example 2, the invention further includes:

[0157] The data display instruction parsing submodule is used to parse the voice input instruction to obtain the data display instruction when the completeness of the voice input instruction reaches a preset completeness index.

[0158] In this embodiment, the data display instruction is parsed from the voice input instruction, including:

[0159] The escaped text of the voice input command is merged, and then the existing semantic recognition model is used (for example, a neural network is used to model the grammatical structure of natural language sentences and infer semantics from it. This model usually uses a sequence to sequence (Seq2Seq) framework to map the input text sequence to the output text sequence, where each word is encoded as a hidden state vector. Commonly used models include LSTM, GRU, and Transformer, etc.) to parse the data display instructions from the escaped text.

[0160] The beneficial effect of the above technology is: achieving parsing and acquisition of data display instructions.

[0161] Embodiment 10:

[0162] Based on Example 1, the data display instruction parsing module refers to Figure 3 ,include:

[0163] A data display instruction direct parsing submodule is used to parse the voice input instruction and the latest voice input instruction to obtain the data display instruction if it is determined based on the voice input instruction and the latest voice input instruction that the latest voice input instruction reaches a preset completeness index;

[0164] The data display instruction indirect parsing submodule is used to analyze the latest missing elements of the latest voice input instruction if it is determined based on the voice input instruction and the latest voice input instruction that the latest voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the latest voice input instruction and the corresponding latest missing elements, and issue the supplementary input voice feedback instruction until the latest received voice input instruction is determined to meet the preset completeness index, then the data display instruction is parsed from all voice control instructions received in this data display control process.

[0165] In this embodiment, the method steps for analyzing the latest missing element of the latest voice input instruction are the same as the method for analyzing the current missing element of the voice input instruction based on the voice input instruction.

[0166] In this embodiment, the method for generating a supplementary input voice feedback instruction based on the voice input instruction and the latest voice input instruction and the corresponding latest missing element is the same as the method for generating a supplementary input voice feedback instruction based on the voice input instruction and the current missing element.

[0167] In this embodiment, the method for determining whether the most recently received voice input instruction reaches the preset completeness index is the same as the method for determining whether the completeness of the voice input instruction reaches the preset completeness index.

[0168] In this embodiment, the method of parsing the data display instruction from all voice control instructions received in this data display control process is the same as the method used when parsing the data display instruction from the voice input instruction and the latest voice input instruction.

[0169] The beneficial effects of the above technology are: realizing the parsing of data display instructions and when the latest voice input instruction is determined to fail to meet the preset completeness index based on the voice input instruction and the latest voice input instruction, the cyclic generation and issuance of the instructions based on the voice input instruction and the latest voice input instruction ensures the integrity of the elements of the data display instructions finally parsed and the control accuracy of data analysis and display on the BI data screen based on the instructions.

[0170] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. The BI data large-screen display system based on voice control is characterized by: include: The voice input command extraction module is used to extract the voice input command in real time from the sound data within a preset decibel value range received in real time around the BI data screen; A supplementary input instruction generation module is used to analyze the current missing elements of the voice input instruction based on the voice input instruction when the completeness of the voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing elements; The command supplement issuing and receiving module is used to issue supplementary input voice feedback commands and, at the same time, obtain the latest voice input command extracted from the sound data within a preset decibel value range received in real time around the BI data screen as the latest voice input command; A data display instruction parsing module, used for parsing the voice input instruction and the latest voice input instruction to obtain the data display instruction; The data operation and display module is used to operate and display the data in the associated database of the BI data screen based on the data display instructions to obtain the display results of the BI data screen; Among them, the supplementary input instruction generation module includes: The completeness judgment submodule is used to judge whether the completeness of the voice input instruction reaches a preset completeness index and obtain a completeness judgment result; A supplementary input instruction generation submodule is used to analyze the current missing elements of the voice input instruction if the completeness judgment result is that the completeness of the voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing elements; Among them, the completeness judgment submodule includes: An instruction element analysis and prediction unit, used to analyze the complete necessary elements and currently covered elements of the voice input instruction; A first completeness determination unit, configured to determine that the completeness of the voice input instruction reaches a preset completeness index when determining that the currently covered elements of the voice input instruction cover the corresponding complete necessary elements, as a completeness determination result; A second completeness determination unit is used to determine that the completeness of the voice input instruction fails to meet a preset completeness index when it is determined that the current covered elements of the voice input instruction cannot cover the corresponding complete necessary elements, as a completeness determination result; Among them, the instruction element analysis and prediction unit includes: The speech escape subunit is used to perform text escape on the speech input instruction, obtain the escaped text of the speech input instruction, and perform element analysis on the escaped text to determine the current covered elements of the speech input instruction; A reference element determination subunit, configured to determine reference necessary elements of the voice input instruction based on the currently covered elements and the instruction necessary elements of the voice input instruction; A prefix and suffix element inference subunit, used to determine whether there is a prefix and suffix element category for the reference necessary element, obtain a prefix and suffix element inference result, and determine the complete necessary elements of the voice input instruction based on the prefix and suffix element inference result and the reference necessary element; Among them, the reference elements determine the sub-units, including: a first reference element determination terminal, configured to, when the currently covered elements of the voice input instruction cover the instruction necessary elements, regard all the instruction necessary elements contained in the voice input instruction as reference necessary elements of the voice input instruction; A second reference element determination end is used for, when the currently covered elements of the voice input instruction do not cover the instruction necessary elements, to perform possible item inference based on all the instruction necessary elements currently contained in the voice input instruction, to determine at least one possible item of the inferred necessary element, and to summarize each possible item of the inferred necessary element with part of the target single elements of the voice input instruction, to obtain a plurality of element groups containing each possible item of the inferred necessary element as reference necessary elements of the voice input instruction; The second reference element determination end includes: The single element determination sub-terminal is used to determine each single element in the current covered elements of the voice input instruction as a target single element when the current covered elements of the voice input instruction do not cover the necessary elements of the instruction; The single element pre-selection sub-end is used to screen out all single elements whose single instruction co-occurrence degree with each target single element exceeds the single instruction co-occurrence degree threshold in the instruction parsing element library, and regard them as all pre-selected single elements of each target single element; The single element reselection sub-terminal is used to treat all preselected single elements of each target single element, whose pre-assigned necessity degree is not less than the pre-assigned necessity degree threshold, as all reselected single elements of each target single element; The single element final selection sub-terminal is used to perform cluster analysis on all re-selected single elements of all target single elements, obtain at least one possible item of inferred necessary elements, and summarize each possible item of inferred necessary elements with some target single elements of the voice input instruction respectively, to obtain multiple element groups containing each possible item of inferred necessary elements as reference necessary elements for the voice input instruction.

2. The BI data large-screen display system based on voice control according to claim 1 is characterized in that: Supplement the input instruction generation submodule, including: a missing element analysis unit, configured to, if the completeness judgment result is that the completeness of the voice input instruction does not meet a preset completeness index, take some elements included in the complete necessary elements of the voice input instruction and not included in the currently covered elements as the current missing elements of the voice input instruction; The feedback instruction generating unit is used to generate a supplementary input voice feedback instruction based on the voice input instruction and the current missing element.

3. The BI data large-screen display system based on voice control according to claim 1 is characterized in that: Also includes: The data display instruction parsing submodule is used to parse the voice input instruction to obtain the data display instruction when the completeness of the voice input instruction reaches a preset completeness index.

4. The BI data large-screen display system based on voice control according to claim 1 is characterized in that: Data display instruction parsing module, including: A data display instruction direct parsing submodule is used to parse the voice input instruction and the latest voice input instruction to obtain the data display instruction if it is determined based on the voice input instruction and the latest voice input instruction that the latest voice input instruction reaches a preset completeness index; The data display instruction indirect parsing submodule is used to analyze the latest missing elements of the latest voice input instruction if it is determined based on the voice input instruction and the latest voice input instruction that the latest voice input instruction does not meet the preset completeness index, and generate a supplementary input voice feedback instruction based on the voice input instruction and the latest voice input instruction and the corresponding latest missing elements, and issue the supplementary input voice feedback instruction until the latest received voice input instruction is determined to meet the preset completeness index, then the data display instruction is parsed from all voice control instructions received in this data display control process.

Images