Intelligent business data analysis system based on deep learning

Abstract

The invention relates to the technical field of business data analysis, and discloses a deep learning-based business data intelligent analysis system, which comprises a data acquisition module, a data integration module, an analysis processing module and an output module, and is characterized in that the data acquisition module comprises a text data acquisition unit, a picture data acquisition unit and a voice data acquisition unit; the text data acquisition unit performs text data set acquisition, the picture data acquisition unit performs picture data set acquisition, the voice data acquisition unit performs voice data set acquisition, and the data integration module integrates a picture text data set, a voice text data set and a text data set to form a comprehensive text data set. And the analysis processing module deletes and selects the comprehensive text data set and removes duplicate values, and the analysis processing module sorts and outputs related information from high to low corresponding data volumes according to text classification after text classification by the output module, so that a rich information basis is provided.

Description

Technical Field

[0001] The present invention relates to the technical field of business data analysis, and specifically to a business data intelligent analysis system based on deep learning. Background Art

[0002] Business data mainly refers to data generated during business processing or transaction processing. These data usually have high timeliness requirements and directly support business applications. Business data is generated in the daily operations and transaction activities of the enterprise, including sales records, customer interactions, inventory changes, financial transactions, etc. Business data can be divided into many types, including but not limited to customer data, transaction data, operational data, etc. Each type of data carries different business value and uses. Business data is crucial for corporate decision-making. Business data is an indispensable part of the daily operations of an enterprise. It not only records the transaction history of the enterprise, but is also an important basis for corporate decision-making and strategic planning. Through in-depth analysis of business data, enterprises can gain insight into market trends, optimize products and services, and improve operational efficiency.

[0003] With the advent of the big data era, the volume and complexity of business data have increased significantly. At the same time, business data exists in various forms. How to effectively manage and combine this data has become a major challenge facing enterprises. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In response to the shortcomings of the prior art, the present invention provides a business data intelligent analysis system based on deep learning, which integrates various types of data, provides a more comprehensive information perspective, helps to deeply understand all aspects of the business, and provides decision makers with a richer information basis, so as to make more accurate and comprehensive decisions, while maintaining the uniqueness of the data and reducing storage costs, etc.

[0006] (II) Technical solution

[0007] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a business data intelligent analysis system based on deep learning, comprising a data acquisition module, a data integration module, an analysis and processing module and an output module;

[0008] The data acquisition module includes a text data acquisition unit, a picture data acquisition unit and a voice data acquisition unit. The text data acquisition unit is connected to the network platform through the network to collect text data, and forms a text data set Wj after numbering. The picture data acquisition unit is connected to the network platform through the network to collect picture data, and forms a picture data set Tj after numbering. The voice data acquisition unit is connected to the network platform through the network to collect voice data, and forms a voice data set Y after numbering.j , the data acquisition module sends the collected data set to the data integration module;

[0009] The data integration module includes a picture information processing unit, a voice information processing unit and a combining unit. The picture information processing unit converts the picture data set Tj into a picture text data set TWj. The voice information processing unit converts the voice data set Yj into a voice text data set YWj. The combining unit integrates and calculates the picture text data set TWj, the voice text data set YWj, and the text data set Wj to form a comprehensive text data set Zj. The data integration module sends the comprehensive text data set Zj to the analysis and processing module.

[0010] The analysis and processing module selects the comprehensive text data set Zj, removes duplicate values, and obtains the final data set Wj. The analysis and processing module performs text classification WL on the final data in the final data set Wj and sends it to the output module. The output module sorts and outputs the corresponding data volume of the relevant information from high to low according to the text classification WL.

[0011] Preferably, the expression of the text dataset Wj is:

[0012] Wj={Wb1, Wb2, Wb3,···,Wb n}

[0013] In the above expression, Wb1~Wb n They represent the first and last text data in the text dataset respectively, and n represents that there are n text data collected in the text dataset.

[0014] Preferably, the expression of the image dataset Tj is:

[0015] Tj={Tp1, Tp2, Tp3,···,Tp n}

[0016] In the above expression, Tp1~Tp n They represent the first and last picture data in the picture data set respectively, and n represents that there are n picture data collected in the picture data set.

[0017] Preferably, the expression of the speech data set Yj is:

[0018] Yj={Yy1, Yy2, Yy3,···,Yy n}

[0019] In the above expression, Yy1~Yy n They represent the first and last speech data in the speech data set respectively, and n represents that there are n speech data collected in the speech data set.

[0020] Preferably, the image information processing unit is internally provided with an optical character recognition technology, and the image data set Tj is converted into an image text data set TWj through the optical character recognition technology. The speech information processing unit is internally provided with an automatic speech recognition technology, and the speech data set Yj is converted into a speech text data set YWj through the automatic speech recognition technology.

[0021] Preferably, the expression of the picture text dataset TWj is: {Tw1, Tw2, Tw3, . . . , Tw n}; The expression of the speech text data set YWj is: {Yw1, Yw2, Yw3, . . . , Yw n}, where Tw1-Tw n Represents Tp1-Tp in the image dataset Tj n Converted image text data, Yw1-Yw n Represents Y in the speech dataset Yj y1 -Yy n The converted speech-to-text data.

[0022] Preferably, the calculation formula of the comprehensive text dataset Zj is:

[0023]

[0024] In the above formula, Represents combining the data in the text dataset Wj, the image text dataset TWj, and the speech text data dataset YWj to form a comprehensive text dataset Zj, Zh1-Zh q They represent the first and last comprehensive text data in the comprehensive text dataset Zj respectively, and q represents that there are q comprehensive text data in the comprehensive text dataset Zj.

[0025] Preferably, the calculation formula of the final data set Wj is as follows:

[0026]

[0027] In the above formula, Zh i ∈Zj∧Zh i =1 represents any value Zh in the comprehensive text dataset Zj i , and it is the only value, Zh i =1 means that the first value in the comprehensive text data set Zj is extracted to the last data, and all Zh i =1, Ww1-Ww hIt represents the first to the last value in the final data set Wj formed after unique value extraction of all values ​​in the comprehensive text data set Zj, and h represents that there are h final data in the final data set Wj.

[0028] Preferably, the text classification WL is performed as follows:

[0029] Extract the vth final data in the final data set Wj and mark it as Ww v , Ww v The front end, middle end and end are divided into segments and marked as And Remove stop words and special symbols;

[0030]

[0031] in, Represents the front-end data after segmenting all the data in the final data set Wj. Represents the mid-range data after segmenting all the data in the final data set Wj. Represents the terminal data after segmenting all the data in the final data set Wj.

[0032] Preferably, the output module is arranged in order according to the amount of specific text information. The output is arranged in the order of .

[0033] Compared with the prior art, the present invention provides a business data intelligent analysis system based on deep learning, which has the following beneficial effects:

[0034] 1. The present invention collects text data, image data, and voice data to form corresponding data sets, which can comprehensively obtain various existing data, maintain the comprehensiveness of the data, enrich the data sources, and more clearly and accurately improve the quality of the data. At the same time, different types of data provide a more comprehensive information perspective, which helps to deeply understand all aspects of the business and provide decision makers with a richer information basis, so as to make more accurate and comprehensive decisions.

[0035] 2. The present invention converts the image data set Tj into the image text data set TWj through optical character recognition technology, converts the voice data set Yj into the voice text data set YWj through automatic speech recognition technology, and then integrates and calculates the image text data set TWj, the voice text data set YWj, and the text data set Wj, thereby maintaining the uniformity of the data and providing convenience for subsequent data analysis and processing. At the same time, the comprehensive text data set Zj is deleted and duplicate values ​​are removed, thereby maintaining the uniqueness of the data and improving the overall data quality. At the same time, a large amount of duplicate data will occupy unnecessary storage space. Removing duplicate values ​​can save storage resources and reduce storage costs. The data is then segmented and sorted to maintain the uniformity of the data output. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the structural system of the present invention; DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] See also Figure 1 , a business data intelligent analysis system based on deep learning, including a data acquisition module, a data integration module, an analysis and processing module and an output module;

[0039] The data collection module includes a text data collection unit, a picture data collection unit and a voice data collection unit. The text data collection unit collects text data through a network connection to a network platform, and forms a text data set Wj after numbering. The picture data collection unit collects picture data through a network connection to a network platform, and forms a picture data set Tj after numbering. The voice data collection unit collects voice data through a network connection to a network platform, and forms a voice data set Yj after numbering. The data collection module sends the collected data set to the data integration module.

[0040] The expression of the text dataset Wi is:

[0041] Wj={Wb1, Wb2, Wb3,···,Wb n}

[0042] In the above expression, Wb1~Wb n Respectively represent the first and last text data in the text dataset, and n represents that there are n text data collected in the text dataset;

[0043] The expression of the image dataset Tj is:

[0044] Tj={Tp1, Tp2, Tp3,···,Tp n )

[0045] In the above expression, Tp1~Tp n They represent the first and last image data in the image dataset respectively, and n represents that there are n image data collected in the image dataset;

[0046] The expression of the speech data set Yj is:

[0047] Yj={Yy1, Yy2, Yy3,···,Yy n}

[0048] In the above expression, Yy1~Yy n They represent the first and last speech data in the speech data set respectively, and n represents that there are n speech data collected in the speech data set;

[0049] By collecting text data, image data, and voice data to form corresponding data sets, we can comprehensively obtain various existing data, maintain the comprehensiveness of the data, enrich the data sources, and improve the quality of the data more clearly and accurately. At the same time, different types of data provide a more comprehensive information perspective, which helps to deeply understand all aspects of the business and provide decision makers with a richer information basis, so as to make more accurate and comprehensive decisions.

[0050] The data integration module includes a picture information processing unit, a voice information processing unit and a combining unit. The picture information processing unit is internally provided with an optical character recognition technology, and the picture data set Tj is converted into a picture text data set TWj through the optical character recognition technology. The voice information processing unit is internally provided with an automatic voice recognition technology, and the voice data set Yj is converted into a voice text data set YWj through the automatic voice recognition technology. The combining unit integrates and calculates the picture text data set TWj, the voice text data set YWj, and the text data set Wj to form a comprehensive text data set Zj. The data integration module sends the comprehensive text data set Zj to the analysis and processing module.

[0051] The expression of the picture text dataset TWj is {Tw1, Tw2, Tw3, ···, Tw n ); the expression of the speech text data set YWj is {Yw1, Yw2, Yw3, . . . , Yw n}, where Tw1-Tw n Represents fp1-Tp in the image dataset Tj n Converted image text data, Yw1-Yw nRepresents Yy1-Yy in the speech dataset Yj n The converted speech-to-text data;

[0052] The calculation formula of the comprehensive text dataset Zj is:

[0053]

[0054] In the above formula, Represents combining the data in the text dataset Wj, the image text dataset TWj, and the speech text data dataset YWj to form a comprehensive text dataset Zj, Zh1-Zh q Respectively represent the first and last comprehensive text data in the comprehensive text dataset Zj, and q represents that there are q comprehensive text data in the comprehensive text dataset Zj;

[0055] The image dataset Tj is converted into the image text dataset TWj through optical character recognition technology, and the voice dataset Yj is converted into the voice text data dataset YWj through automatic speech recognition technology. Then, the image text dataset TWj, the voice text data dataset YWj, and the text dataset Wj are integrated and calculated, which maintains the uniformity of the data and provides convenience for subsequent data analysis and processing;

[0056] The analysis and processing module selects the comprehensive text data set Zj, removes duplicate values, and obtains the final data set Wj;

[0057] The calculation formula of the final data set Wj is as follows:

[0058]

[0059] In the above formula, Zh i ∈Zj∧Zh i =1 represents any value Zh in the comprehensive text dataset Zj i , and Zh i is a unique value, Zh i =1 means that the first value in the comprehensive text data set Zj is extracted to the last data, and all Zh i =1, Ww1-Ww h represents the first to the last value in the final data set Wj formed after unique value extraction of all values ​​in the comprehensive text data set Zj, and h represents that there are h final data in the final data set Wj;

[0060] The analysis and processing module performs text classification WL on the final data in the final data set Wj and sends it to the output module;

[0061] The text classification WL method is as follows:

[0062] Extract the vth final data in the final data set Wj and mark it as Ww v , Ww v The front end, middle end and end are divided into segments and marked as And Remove stop words and special symbols;

[0063]

[0064] in, Represents the front-end data after segmenting all the data in the final data set Wj. Represents the mid-range data after segmenting all the data in the final data set Wj. Represents the terminal data after segmenting all the data in the final data set Wj;

[0065] By deleting and selecting the comprehensive text data set Zj and removing duplicate values, the uniqueness of the data is maintained and the overall quality of the data is improved. At the same time, a large amount of duplicate data will take up unnecessary storage space. Removing duplicate values ​​can save storage resources and reduce storage costs. The data is then sorted in segments to maintain the uniformity of data output.

[0066] The output modules are arranged in order according to the amount of specific text information. The output is arranged in the order of .

[0067] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A business data intelligent analysis system based on deep learning, characterized by: It includes data acquisition module, data integration module, analysis and processing module and output module; The data acquisition module includes a text data acquisition unit, a picture data acquisition unit and a voice data acquisition unit. The text data acquisition unit is connected to the network platform through the network to collect text data, and forms a text data set Wj after numbering. The picture data acquisition unit is connected to the network platform through the network to collect picture data, and forms a picture data set Tj after numbering. The voice data acquisition unit is connected to the network platform through the network to collect voice data, and forms a voice data set Yj after numbering. The data acquisition module sends the collected data set to the data integration module; The data integration module includes a picture information processing unit, a voice information processing unit and a combining unit. The picture information processing unit converts the picture data set Tj into a picture text data set TWj. The voice information processing unit converts the voice data set Yj into a voice text data set YWj. The combining unit integrates and calculates the picture text data set TWj, the voice text data set YWj, and the text data set Wj to form a comprehensive text data set Zj. The data integration module sends the comprehensive text data set Zj to the analysis and processing module. The analysis and processing module selects the comprehensive text data set Zj, removes duplicate values, and obtains the final data set Wj. The analysis and processing module performs text classification WL on the final data in the final data set Wj and sends it to the output module. The output module sorts and outputs the corresponding data volume of the relevant information from high to low according to the text classification WL.

2. The business data intelligent analysis system based on deep learning according to claim 1 is characterized by: The expression of the text dataset Wj is: Wj={Wb1、Wb2、Wb3、···、Wb n } In the above expression, Wb1~Wb n They represent the first and last text data in the text dataset respectively, and n represents that there are n text data collected in the text dataset.

3. The business data intelligent analysis system based on deep learning according to claim 2 is characterized by: The expression of the image dataset Tj is: Tj={Tp1、Tp2、Tp3、···、Tp n } In the above expression, Tp1~Tp n They represent the first and last picture data in the picture data set respectively, and n represents that there are n picture data collected in the picture data set.

4. The business data intelligent analysis system based on deep learning according to claim 3 is characterized by: The expression of the speech data set Yj is: Yj={Yy1、Yy2、Yy3、···、Yy n } In the above expression, Yy1~Yy n They represent the first and last speech data in the speech data set respectively, and n represents that there are n speech data collected in the speech data set.

5. The business data intelligent analysis system based on deep learning according to claim 4 is characterized by: The image information processing unit is internally provided with an optical character recognition technology, through which the image data set Tj is converted into an image text data set TWj. The voice information processing unit is internally provided with an automatic voice recognition technology, through which the voice data set Yj is converted into a voice text data set YWj.

6. The business data intelligent analysis system based on deep learning according to claim 5 is characterized by: The expression of the picture text dataset TWj is: {Tw1, Tw2, Tw3, . . . , Tw n }; The expression of the speech text data set YWj is: {Yw1, Yw2, Yw3, . . . , Yw n }, where Tw1-Tw n Represents Tp1-Tp in the image dataset Tj n Converted image text data, Yw1-Yw n Represents Yy1-Yy in the speech dataset Yj n The converted speech-to-text data.

7. The business data intelligent analysis system based on deep learning according to claim 6 is characterized by: The calculation formula of the comprehensive text dataset Zj is: In the above formula, It represents combining the data in the text dataset Wj, the picture text dataset TWj, and the speech text data dataset TWj to form a comprehensive text dataset Zj. Zh1-Zhq respectively represent the first and last comprehensive text data in the comprehensive text dataset Zj. q represents that there are q comprehensive text data in the comprehensive text dataset Zj.

8. The business data intelligent analysis system based on deep learning according to claim 7 is characterized by: The calculation formula of the final data set Wj is as follows: In the above formula, Zh i ∈Zj∧Zh i =1 represents any value Zh in the comprehensive text dataset Zj i , and Zh i is a unique value, Zh i =1 means that the first value in the comprehensive text data set Zj is extracted to the last data, and all Zh i = the only value of 1, Ww1-Ww h It represents the first to the last value in the final data set Wj formed after unique value extraction of all values ​​in the comprehensive text data set Zj, and h represents that there are h final data in the final data set Wj.

9. The business data intelligent analysis system based on deep learning according to claim 8, characterized in that: The text classification WL is as follows: Extract the vth final data in the final data set Wj and mark it as Ww v , Ww v The front end, middle end and end are segmented and marked as And Remove stop words and special symbols; in, Represents the front-end data after segmenting all the data in the final data set Wj. Represents the mid-range data after segmenting all the data in the final data set Wj. Represents the terminal data after segmenting all the data in the final data set Wj.

10. The business data intelligent analysis system based on deep learning according to claim 9, characterized in that: The output module is arranged in order according to the amount of specific text information. The output is arranged in the order of .

Images