CN114331226A - Intelligent enterprise demand diagnosis method and system and storage medium - Google Patents
Intelligent enterprise demand diagnosis method and system and storage medium Download PDFInfo
- Publication number
- CN114331226A CN114331226A CN202210218099.7A CN202210218099A CN114331226A CN 114331226 A CN114331226 A CN 114331226A CN 202210218099 A CN202210218099 A CN 202210218099A CN 114331226 A CN114331226 A CN 114331226A
- Authority
- CN
- China
- Prior art keywords
- enterprise
- data
- demand
- diagnosis
- fusion
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 56
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000003860 storage Methods 0.000 title claims abstract description 7
- 230000004927 fusion Effects 0.000 claims abstract description 57
- 238000004458 analytical method Methods 0.000 claims abstract description 19
- 238000001914 filtration Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 13
- 239000013598 vector Substances 0.000 claims description 50
- 230000006870 function Effects 0.000 claims description 16
- 238000000605 extraction Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 230000002860 competitive effect Effects 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 6
- 238000004519 manufacturing process Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims 2
- 238000013135 deep learning Methods 0.000 abstract description 5
- 239000000463 material Substances 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 11
- 230000004913 activation Effects 0.000 description 6
- 239000000284 extract Substances 0.000 description 5
- 230000008520 organization Effects 0.000 description 5
- 238000012549 training Methods 0.000 description 5
- 238000013501 data transformation Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 230000010354 integration Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000011534 incubation Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an enterprise demand intelligent diagnosis method, a system and a storage medium, wherein the method comprises the following steps: acquiring source data in a corresponding industry according to the industry category to which an enterprise belongs; filtering and preprocessing the acquired source data under the industry category, extracting key information effective to demand analysis, and eliminating noise and invalid information; extracting deep fusion characteristics of enterprise data; constructing a demand diagnosis identification model by using the obtained depth correlation characteristics; and carrying out actual enterprise demand diagnosis by using the constructed recognition model to obtain demand service matched with the current stage demand of the enterprise. The enterprise demand intelligent diagnosis method provided by the invention can greatly improve the data processing and analysis efficiency, reduce the cost of manpower and material resources, and the diagnosis model obtained by the deep learning method is more objective than the artificial analysis of experts, so that the enterprise can carry out intelligent diagnosis in different stages conveniently to obtain demand results.
Description
Technical Field
The invention belongs to the field of intelligent data mining and data analysis, and particularly relates to an enterprise demand intelligent diagnosis method, system and storage medium.
Background
At present, China is in the internet and intelligent information era, and the nation vigorously advocates to rush new opportunities in implementing digital economic strategies, accelerate the innovative development of industrial internet and promote the deep fusion of new-generation information technology and traditional enterprise service. At present, an enterprise management information system manages and queries various enterprise information, however, with the deepening of scientific and creative services, various entrepreneurship service organizations and incubators are dedicated to providing comprehensive incubation services such as scientific and technological achievement transformation, financing, project landing, policy service, industrial resource import and the like, and directly investing in high-quality early projects, different types of enterprises may need different demands in different stages of development, and the enterprises need to invest a large amount of manpower and material resources to perform data processing and perform demand analysis by experts, generally, in the prior art, in general, the experts in the field perform enterprise data research and survey, questionnaire survey, enterprise interview, seat interview and the like, and then make diagnosis reports and send the diagnosis reports to the enterprise side according to the data to provide decisions, however, in the process, the enterprise needs a special department to do the work, or needs to entrust a third-party service organization to perform diagnosis of demands, the process needs to process a large amount of data, a large amount of field workers need to participate and butt joint in the process of obtaining the conclusion, and the service efficiency is low and the period is long.
Disclosure of Invention
The present invention is directed to solve the above technical problems in the prior art, and provides an enterprise demand intelligent diagnosis method, system and storage medium.
The invention is realized by the following technical scheme: the method for intelligently diagnosing the enterprise requirements comprises the following steps:
step 1, acquiring source data in a corresponding industry according to the industry category to which an enterprise belongs, wherein the acquired source data can be acquired in the following various ways including but not limited to enterprise query APP, Web crawler, enterprise WeChat, small programs and the like. The acquired data at least comprises six-dimensional major data of enterprise basic data, enterprise financing data, enterprise operation data, enterprise production known data, enterprise label data and enterprise competitive product data, and specifically contained data subclass information, which are shown in the following table:
| data dimension | Subclass of data |
| Enterprise base data | Enterprise name, enterprise full name, basic information, time of establishment, region, registered capital, personnel size |
| Enterprise financing data | Financing rounds, financing time, financing amount, investment organization |
| Business operation data | Brief introduction of main business, product sale condition and enterprise product type |
| Enterprise informed product data | Patent, trademark, copyright and soft work |
| Enterprise tag data | Large category of industry, application scene, financing situation |
| Enterprise competitive product data | Competing pairsHand basic data, competitor financing data and competitor operation data |
The method collects 10 ten thousand large and medium-sized enterprises as source data samples, extracts service data containing the corresponding stages of the data, and sets a plurality of specific service subclasses under each specific service class, wherein the service classes comprise data services, commissioning consultation, entrepreneurial services and professional services, and the table is as follows:
step 2, filtering, preprocessing and extracting the acquired source data under the industry category, extracting key information useful for demand analysis, and eliminating noise and invalid information;
removing HTML source codes of the webpage according to the captured content related to the enterprise data, filtering and denoising the contents such as advertisements and comments in the webpage source codes, and extracting Chinese text contents in the captured data; the preprocessing of the chinese text may be to divide the sentence into a plurality of words, where the preprocessed data is a series of words, and includes one or more of data redundancy removal, data integration, data fusion, data transformation, and data reduction.
Step 3, extracting deep fusion characteristics of enterprise data;
and 2, after the enterprise text data preprocessed in the step 2 is obtained, because words in the text are discrete, each word is subjected to space mapping through the word embedding layer, and a low-dimensional dense word vector is obtained to represent each word. Thus, each word is a one-dimensional vector, a sentence can be represented by using a plurality of one-dimensional vectors, and a matrix is obtained to represent the sentence; the word vector can be pre-trained word vector of any one of word2vec, GloVe and fasttext, or fixed dimension vector which is randomly generated and obeys normal distribution can be used as the word vector.
In order to further extract depth features among enterprise data information and relevance among the information, in the process of extracting the depth convolution features, depth fusion is carried out on the obtained enterprise data word vector features of multiple dimensions, convolution operation of two branches is carried out on a first dimension feature, wherein the first branch comprises three convolution layers, the second branch convolution layer comprises one convolution layer, and the word vector fusion features under the dimension are obtained through convolution processing of the two branches; obtaining second dimension information and first dimension information to carry out correlation fusion characteristics, carrying out fusion add calculation on the word vector characteristics of the second dimension through characteristics obtained by convolution of a third branch and characteristics obtained by convolution of a second branch of the first dimension characteristics, finally carrying out concat calculation on the word vector fusion characteristics under the first dimension and the obtained correlation fusion characteristics among the dimensions, carrying out the fusion operation on the characteristics among the dimension characteristics, and finally obtaining the fusion characteristics of the dimension and the depth correlation characteristics among the dimensions;
in the application, the first branch of the first dimension characteristic is in a three-layer structure with convolution kernels of 1 × 1, 5 × 5 and 1 × 1, and the convolution kernel of the second branch is 1 × 1. The third branch is a three-layer structure with convolution kernels of 1 × 1, 5 × 5 and 1 × 1, so that not only can own dimensional data be fused through different channels to obtain fusion characteristics under different receptive fields, but also different dimensional characteristic fusion is performed through a plurality of channels with different dimensions, and the depth semantic characteristics with high word vector characteristic correlation can be obtained. Similar characteristics among other dimensions are subjected to fusion characteristic processing in the same process, so that the accuracy of model analysis can be greatly improved.
Step 4, building a demand diagnosis recognition model
After the input word vectors are fused through the feature fusion channel, as features with higher latitude are generated, subsequent pooling layer operation and full-connection layer calculation are carried out, and finally a demand result is output through a classification layer, wherein the activation function adopts a Relu activation function, and the loss function adopts a cross entropy (cross entropy) loss function:
wherein K is the number of label values of the required classes in model training, and y is a real label for a sample point i, and then y isi,k=1, otherwise equal to 0, i being the number of samples, the probability of a sample i belonging to a label k being pi,kN is the total number of samples, LlossTo train the loss value.
And 5, carrying out actual enterprise demand diagnosis by using the constructed recognition model to obtain demand service matched with the current stage demand of the enterprise.
In practice, the same word vector and depth feature may be extracted after the enterprise data is acquired according to the method in step 1, and if the acquired data is different, the user may manually correct the text content and input the corrected text content into the diagnosis and recognition model to obtain one or more service requirement tags that the enterprise needs to obtain at present.
Further, the application also provides an enterprise demand intelligent diagnosis system, which comprises the following functional modules:
the intelligent acquisition and source data acquisition module acquires source data in corresponding industries according to the industry types to which the enterprises belong, and the acquired source data can be acquired in various ways including but not limited to enterprise query APP, Web crawler, enterprise WeChat, small programs and the like. The acquired data at least comprises six-dimensional major data of enterprise basic data, enterprise financing data, enterprise operation data, enterprise production known data, enterprise label data and enterprise competitive product data, and specifically contained data subclass information, which are shown in the following table:
| data dimension | Subclass of data |
| Enterprise base data | Enterprise name, enterprise full name, basicInformation, time to establish, region, registered capital, size of personnel |
| Enterprise financing data | Financing rounds, financing time, financing amount, investment organization |
| Business operation data | Brief introduction of main business, product sale condition and enterprise product type |
| Enterprise informed product data | Patent, trademark, copyright and soft work |
| Enterprise tag data | Large category of industry, application scene, financing situation |
| Enterprise competitive product data | Competitor basic data, competitor financing data, competitor operation data |
The method collects 10 ten thousand large and medium-sized enterprises as source data samples, extracts service data containing the corresponding stages of the data, and sets a plurality of specific service subclasses under each specific service class, wherein the service classes comprise data services, commissioning consultation, entrepreneurial services and professional services, and the table is as follows:
the data filtering and data preprocessing module is used for filtering, preprocessing and extracting the acquired source data under the industry category, extracting key information which is useful for demand analysis, and eliminating noise and invalid information;
removing HTML source codes of the webpage according to the captured content related to the enterprise data, filtering and denoising the contents such as advertisements and comments in the webpage source codes, and extracting Chinese text contents in the captured data; the preprocessing of the chinese text may be to divide the sentence into a plurality of words, where the preprocessed data is a series of words, and includes one or more of data redundancy removal, data integration, data fusion, data transformation, and data reduction.
The enterprise data deep fusion feature extraction module is used for carrying out spatial mapping on each word of the preprocessed enterprise text data through a word embedding layer due to the fact that the words in the text are discrete, and low-dimensional and dense word vectors are obtained to represent each word. Therefore, each word is a one-dimensional vector, one sentence can be represented by a plurality of one-dimensional vectors, and a matrix can be obtained to represent one sentence; the word vector can be pre-trained word vector of any one of word2vec, GloVe and fasttext, or fixed dimension vector which is randomly generated and obeys normal distribution can be used as the word vector.
In order to further extract depth features among enterprise data information and the relevance among the information, in the process of extracting the depth convolution features, the depth features are merged layer by layer according to the sequence of enterprise basic data, different dimensionality features are fused through a plurality of channels, and then the depth semantic features with high word vector feature relevance are obtained. And the demand diagnosis and identification model building module is used for fusing input word vectors through a feature fusion channel and inputting the fused word vectors into a deep network for model training, and generating higher-dimension features, so that pooling layer operation and full-connection layer calculation are performed subsequently, and finally a demand result is output through a classification layer, wherein the activation function adopts a Relu activation function, and the loss function adopts a cross entropy (cross entropy) loss function.
Wherein K is the number of required class label values in model training, and for sample pointsi say y is a true tag, then yi,k=1, otherwise equal to 0, i being the number of samples, the probability of a sample i belonging to a label k being pi,kN is the total number of samples, LlossTo train the loss value. The specific training preset parameters and parameter adjustment are prior art and are not described herein.
And the diagnosis result output module is used for carrying out actual enterprise demand diagnosis by utilizing the constructed recognition model to obtain demand service matched with the current stage demand of the enterprise.
In practice, the same word vector and depth feature may be extracted after the enterprise data is acquired according to the method in step 1, and if the acquired data is different, the user may manually correct the text content and input the corrected text content into the diagnosis and recognition model to obtain one or more service requirement tags that the enterprise needs to obtain at present.
Besides, the application also provides a computing device and a computer-readable storage medium for intelligent diagnosis of enterprise needs, which are characterized by comprising a processor and a memory, wherein the memory stores computer-executable instructions capable of being executed by the processor, and the processor executes the computer-executable instructions to realize the intelligent diagnosis method of enterprise needs. When the computer-executable instructions are called and executed by the processor, the computer-executable instructions cause the processor to realize the intelligent enterprise demand diagnosis method.
Compared with the prior art, the invention has the beneficial effects that: the intelligent enterprise demand diagnosis and analysis method and system greatly improve the data processing and analysis efficiency, reduce the labor and material cost, enable a diagnosis model obtained by a deep learning method to be more objective than the manual analysis of experts, enable enterprise personnel to conveniently perform model analysis without mastering the analysis experience, and enable enterprises to perform intelligent diagnosis at different stages conveniently to obtain demand results. In addition, aiming at the characteristics that enterprise data is large and deep correlation exists among the data, a novel correlation characteristic convolution depth characteristic fusion extraction network is provided, the network can obtain multi-receptive-field depth fusion characteristics under own dimensionality, and further can perform fusion on different dimensionality word vector characteristics through multiple channels with different dimensionalities, so that depth semantic characteristics with high word vector characteristic correlation can be obtained, similar processing is performed among different dimensionality characteristics, and the accuracy of diagnostic model analysis is greatly improved.
Drawings
FIG. 1 is a data processing and service flow for existing enterprise demand diagnostics;
FIG. 2 illustrates a process of the intelligent diagnosis deep learning network architecture.
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings:
referring to fig. 1, a third party service organization is taken as an example to perform demand diagnosis, and generally includes the following steps: the demand diagnosis enterprise/organization puts demands: and carrying out advanced research on enterprises, determining the requirements of the enterprises and preliminarily determining, evaluating and diagnosing the intention of cooperation. Clear diagnosis plan: and deeply communicating with the enterprises, and making an evaluation diagnosis plan according to the characteristics of the enterprises. And (3) reaching a cooperation agreement: and (4) agreeing with the enterprise to evaluate the diagnostic rules and sign a cooperative agreement. And (4) performing online diagnosis and issuing questionnaires by the expert team, and performing on-site investigation on the enterprises. And performing questionnaire survey statistical data, performing enterprise diagnosis and analysis, processing enterprise and industry data, performing manual statistical analysis on the existing data, performing enterprise diagnosis and analysis by combining the current situation of enterprise capability, and finally obtaining a diagnosis report for the enterprise.
Referring to fig. 2: the intelligent diagnosis deep learning network architecture process comprises the following processing layers and technical contents:
the intelligent diagnosis deep learning network comprises a data filtering and preprocessing layer, a word vector feature extraction layer, a feature deep fusion layer, a pooling layer, a full connection layer and a result output layer, source data are obtained in corresponding industries according to the industry types to which enterprises belong, and the source data can be obtained in the following modes, including but not limited to enterprise query APP, Web crawler, enterprise WeChat, small programs and the like. Filtering and preprocessing the acquired source data under the industry category through a data filtering and preprocessing layer, extracting key information which is useful for demand analysis, and eliminating noise and invalid information; removing HTML source codes of the webpage according to the captured content related to the enterprise data, filtering and denoising the contents such as advertisements and comments in the webpage source codes, and extracting Chinese text contents in the captured data; the preprocessing of the chinese text may be to divide the sentence into a plurality of words, where the preprocessed data is a series of words, and includes one or more of data redundancy removal, data integration, data fusion, data transformation, and data reduction.
And expressing the preprocessed enterprise text data by using a low-dimensional and dense word vector through a word vector feature extraction layer. Each word is a one-dimensional vector, and one sentence can be represented by using a plurality of one-dimensional vectors, so that a matrix is obtained to represent one sentence; the word vector can be pre-trained word vector of any one of word2vec, GloVe and fasttext, or fixed dimension vector which is randomly generated and obeys normal distribution can be used as the word vector.
In order to further extract depth features among enterprise data information and relevance among the information, in the process of extracting the depth convolution features, a feature depth fusion layer fuses each obtained dimension word vector feature, first, a first dimension feature passes through two branches, different layers and convolution kernel size processing are respectively designed, fusion features under the dimension are obtained, a first channel of the first dimension feature is 1 x 1, 5 x 5 and 1 x 1 after passing through convolution kernels, and a convolution kernel of a second branch is 1 x 1. Further, in order to extract feature relevance among enterprise information of each dimension, word vector features of a second dimension are subjected to fusion add calculation through convolution operation and features obtained through the second branch of the first dimension, and finally two-dimension fusion features and the first-dimension features are subjected to concatemate calculation, so that not only can fusion of data of the second dimension through different channels obtain fusion features under different receptive fields, but also fusion of the features of the different dimensions is performed through a plurality of channels of the different dimensions, and depth semantic features with high word vector feature relevance can be obtained. Similar second dimension and third dimension features are subjected to fusion feature processing in the same process, so that the accuracy of model analysis can be greatly improved.
After the input word vectors are fused through the feature fusion channel, as features with higher latitude are generated, subsequent pooling layer operation and full-connection layer calculation are carried out, and finally a demand result is output through a classification layer, wherein the activation function adopts a Relu activation function, and the loss function adopts a cross entropy (cross entropy) loss function:
wherein K is the number of label values of the required classes in model training, and y is a real label for a sample point i, and then y isi,k=1, otherwise equal to 0, i being the number of samples, the probability of a sample i belonging to a label k being pi,kN is the total number of samples, LlossTo train the loss value.
And carrying out actual enterprise demand diagnosis by using the constructed recognition model to obtain demand service matched with the current stage demand of the enterprise.
In the description of the present invention, it is to be noted that, unless otherwise explicitly specified or limited, the terms "connected" and "connected" are to be interpreted broadly, e.g., as being fixed or detachable or integrally connected; can be mechanically or electrically connected; may be directly connected or indirectly connected through an intermediate. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
In the description of the present invention, unless otherwise specified, the terms "upper", "lower", "left", "right", "inner", "outer", and the like indicate orientations or positional relationships based on those shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention.
Finally, it should be noted that the above-mentioned technical solution is only one embodiment of the present invention, and it will be apparent to those skilled in the art that various modifications and variations can be easily made based on the application method and principle of the present invention disclosed, and the method is not limited to the above-mentioned specific embodiment of the present invention, so that the above-mentioned embodiment is only preferred, and not restrictive.
Claims (10)
1. An enterprise demand intelligent diagnosis method comprises the following steps:
step 1, acquiring source data in a corresponding industry according to the industry type of an enterprise;
step 2, filtering and preprocessing the acquired source data under the industry category, extracting key information effective to demand analysis, and eliminating noise and invalid information;
step 3, extracting deep fusion characteristics of enterprise data;
performing word vector extraction on the enterprise text data preprocessed in the step 2 according to preset data dimensions, performing deep fusion on the acquired enterprise data word vector characteristics of multiple dimensions, performing convolution operation on a first dimension characteristic through two branches, wherein the first branch comprises three convolutional layers, the second branch comprises one convolutional layer, and obtaining word vector fusion characteristics under the dimensions through the convolutional layer processing of the two branches; obtaining second dimension information and first dimension information to carry out correlation fusion characteristics, carrying out fusion add calculation on the word vector characteristics of the second dimension through characteristics obtained by convolution of a third branch and characteristics obtained by convolution of a second branch of the first dimension characteristics, finally carrying out concat calculation on the word vector fusion characteristics under the first dimension and the obtained correlation fusion characteristics among the dimensions, carrying out the fusion operation on the characteristics among the dimension characteristics, and finally obtaining the fusion characteristics of the dimension and the depth correlation characteristics among the dimensions;
step 4, constructing a demand diagnosis identification model by using the obtained depth correlation characteristics;
and 5, carrying out actual enterprise demand diagnosis by using the constructed recognition model to obtain demand service matched with the current stage demand of the enterprise.
2. The intelligent diagnosis method for enterprise demand according to claim 1, characterized in that: in the enterprise data deep fusion feature extraction, the convolution kernels of the first branch network are 1 × 1, 5 × 5 and 1 × 1 in sequence, and the convolution kernel of the second branch network is 1 × 1.
3. The intelligent diagnosis method for enterprise demand according to claim 1, characterized in that: the enterprise sample data acquired by the identification model at least comprises six-dimensional large data of enterprise basic data, enterprise financing data, enterprise operation data, enterprise known production data, enterprise tag data and enterprise competitive product data; the service type data comprises data service, operation consultation, entrepreneurship service and professional service basis.
4. The intelligent enterprise demand diagnosis method of claim 1, wherein the demand diagnosis identification model further comprises pooling and full-join computing after deep fusion feature extraction, and the demand diagnosis identification model classification function is a SoftMax function.
5. An enterprise demand intelligent diagnosis system is characterized by comprising the following modules: the system comprises an intelligent acquisition and source data acquisition module, a data filtering and data preprocessing module, an enterprise data deep fusion feature extraction module, a demand diagnosis identification model construction module and a diagnosis result output module, wherein:
the intelligent acquisition and source data acquisition module acquires source data in the corresponding industry according to the industry type of the enterprise;
the data filtering and data preprocessing module is used for filtering and preprocessing the acquired source data under the industry category, extracting key information effective to demand analysis and eliminating noise and invalid information;
the enterprise data deep fusion feature extraction module is used for extracting the enterprise data deep fusion feature;
performing word vector extraction on the preprocessed enterprise text data according to preset data dimensions by an enterprise, performing deep fusion on the acquired enterprise data word vector characteristics of multiple dimensions, performing convolution operation on a first dimension characteristic through two branches, wherein the first branch comprises three convolutional layers, the second branch comprises one convolutional layer, and obtaining word vector fusion characteristics under the dimensions through the convolutional layer processing of the two branches; obtaining second dimension information and first dimension information to carry out correlation fusion characteristics, carrying out fusion add calculation on the word vector characteristics of the second dimension through characteristics obtained by convolution of a third branch and characteristics obtained by convolution of a second branch of the first dimension characteristics, finally carrying out concat calculation on the word vector fusion characteristics under the first dimension and the obtained correlation fusion characteristics among the dimensions, carrying out the fusion operation on the characteristics among the dimension characteristics, and finally obtaining the fusion characteristics of the dimension and the depth correlation characteristics among the dimensions;
the demand diagnosis identification model construction module is used for constructing a demand diagnosis identification model by using the obtained depth correlation characteristics;
and the diagnosis result output module is used for carrying out actual enterprise demand diagnosis by utilizing the constructed recognition model to obtain demand service matched with the current stage demand of the enterprise.
6. The system of claim 5, wherein the convolution kernels of the first branch network and the second branch network in the deep fusion feature extraction of the enterprise data are 1 x 1, 5 x 5 and 1 x 1 in sequence, and the convolution kernel of the second branch network is 1 x 1.
7. The system of claim 5, wherein: the enterprise sample data acquired by the identification model at least comprises six-dimensional large data of enterprise basic data, enterprise financing data, enterprise operation data, enterprise known production data, enterprise tag data and enterprise competitive product data; the service type data comprises data service, operation consultation, entrepreneurship service and professional service basis.
8. The system of claim 5, wherein the demand diagnosis recognition model further comprises pooling and full-join computing after deep fusion feature extraction, and the demand diagnosis recognition model classification function is a SoftMax function.
9. A computer device comprising a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method of any one of claims 1 to 4.
10. A computer-readable storage medium having computer-executable instructions stored thereon which, when invoked and executed by a processor, cause the processor to implement the method of any of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210218099.7A CN114331226B (en) | 2022-03-08 | 2022-03-08 | Intelligent enterprise demand diagnosis method and system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210218099.7A CN114331226B (en) | 2022-03-08 | 2022-03-08 | Intelligent enterprise demand diagnosis method and system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114331226A true CN114331226A (en) | 2022-04-12 |
| CN114331226B CN114331226B (en) | 2022-06-24 |
Family
ID=81030511
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210218099.7A Active CN114331226B (en) | 2022-03-08 | 2022-03-08 | Intelligent enterprise demand diagnosis method and system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114331226B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116431711A (en) * | 2023-06-13 | 2023-07-14 | 北京长河数智科技有限责任公司 | Data intelligent acquisition method and system based on data characteristics |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108717439A (en) * | 2018-05-16 | 2018-10-30 | 哈尔滨理工大学 | A kind of Chinese Text Categorization merged based on attention mechanism and characteristic strengthening |
| CN110489749A (en) * | 2019-08-07 | 2019-11-22 | 北京航空航天大学 | Intelligent Office-Automation System Work Flow Optimizing |
| CN110674305A (en) * | 2019-10-10 | 2020-01-10 | 天津师范大学 | Deep feature fusion model-based commodity information classification method |
| CN111460830A (en) * | 2020-03-11 | 2020-07-28 | 北京交通大学 | Method and system for extracting economic events in judicial texts |
| CN112364638A (en) * | 2020-10-13 | 2021-02-12 | 北京工业大学 | Personality identification method based on social text |
| CN112686843A (en) * | 2020-12-21 | 2021-04-20 | 福建新大陆软件工程有限公司 | Wood board defect detection method and system based on neural network |
| CN112818892A (en) * | 2021-02-10 | 2021-05-18 | 杭州医典智能科技有限公司 | Multi-modal depression detection method and system based on time convolution neural network |
| CN113065698A (en) * | 2021-03-25 | 2021-07-02 | 珠海大横琴科技发展有限公司 | Service processing method and device |
| CN113177733A (en) * | 2021-05-20 | 2021-07-27 | 北京信息科技大学 | Medium and small micro-enterprise data modeling method and system based on convolutional neural network |
| CN113591525A (en) * | 2020-10-27 | 2021-11-02 | 蓝海（福建）信息科技有限公司 | Driver road rage recognition method with deep fusion of facial expressions and voice |
| CN113643133A (en) * | 2021-08-17 | 2021-11-12 | 南开大学 | LSTM stock price prediction optimization method based on 3D-CNN and Hash attention mechanism |
| CN113919606A (en) * | 2021-12-14 | 2022-01-11 | 山东建筑大学 | Intelligent site selection method and system for distributed photovoltaic power station |
-
2022
- 2022-03-08 CN CN202210218099.7A patent/CN114331226B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108717439A (en) * | 2018-05-16 | 2018-10-30 | 哈尔滨理工大学 | A kind of Chinese Text Categorization merged based on attention mechanism and characteristic strengthening |
| CN110489749A (en) * | 2019-08-07 | 2019-11-22 | 北京航空航天大学 | Intelligent Office-Automation System Work Flow Optimizing |
| CN110674305A (en) * | 2019-10-10 | 2020-01-10 | 天津师范大学 | Deep feature fusion model-based commodity information classification method |
| CN111460830A (en) * | 2020-03-11 | 2020-07-28 | 北京交通大学 | Method and system for extracting economic events in judicial texts |
| CN112364638A (en) * | 2020-10-13 | 2021-02-12 | 北京工业大学 | Personality identification method based on social text |
| CN113591525A (en) * | 2020-10-27 | 2021-11-02 | 蓝海（福建）信息科技有限公司 | Driver road rage recognition method with deep fusion of facial expressions and voice |
| CN112686843A (en) * | 2020-12-21 | 2021-04-20 | 福建新大陆软件工程有限公司 | Wood board defect detection method and system based on neural network |
| CN112818892A (en) * | 2021-02-10 | 2021-05-18 | 杭州医典智能科技有限公司 | Multi-modal depression detection method and system based on time convolution neural network |
| CN113065698A (en) * | 2021-03-25 | 2021-07-02 | 珠海大横琴科技发展有限公司 | Service processing method and device |
| CN113177733A (en) * | 2021-05-20 | 2021-07-27 | 北京信息科技大学 | Medium and small micro-enterprise data modeling method and system based on convolutional neural network |
| CN113643133A (en) * | 2021-08-17 | 2021-11-12 | 南开大学 | LSTM stock price prediction optimization method based on 3D-CNN and Hash attention mechanism |
| CN113919606A (en) * | 2021-12-14 | 2022-01-11 | 山东建筑大学 | Intelligent site selection method and system for distributed photovoltaic power station |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116431711A (en) * | 2023-06-13 | 2023-07-14 | 北京长河数智科技有限责任公司 | Data intelligent acquisition method and system based on data characteristics |
| CN116431711B (en) * | 2023-06-13 | 2024-03-15 | 北京长河数智科技有限责任公司 | Data intelligent acquisition method and system based on data characteristics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114331226B (en) | 2022-06-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Czimmermann et al. | Visual-based defect detection and classification approaches for industrial applications—a survey | |
| Raval et al. | Lean Six Sigma implementation: modelling the interaction among the enablers | |
| CN110597735A (en) | Software defect prediction method for open-source software defect feature deep learning | |
| CN106572001B (en) | A kind of dialogue method and system of intelligent customer service | |
| CN111737485A (en) | Human-sentry matching method and human-sentry matching system based on knowledge graph and deep learning | |
| CN114331226B (en) | Intelligent enterprise demand diagnosis method and system and storage medium | |
| CN112836809A (en) | Device characteristic extraction method and fault prediction method of convolutional neural network based on differential feature fusion | |
| CN114238655A (en) | Enterprise association relation identification method, device, equipment and medium | |
| Varghese et al. | INFOPLANT: Plant recognition using convolutional neural networks | |
| CN113111158B (en) | Intelligent data visualization oriented conversational question-answering implementation method | |
| CN113935880A (en) | Policy recommendation method, device, equipment and storage medium | |
| CN112347252B (en) | Interpretability analysis method based on CNN text classification model | |
| CN113378024A (en) | Deep learning-based public inspection field-oriented related event identification method | |
| CN117252459A (en) | Fruit quality evaluation system based on deep learning | |
| CN114331230B (en) | Full-chain full-flow intelligent scientific and creative service data processing method and system | |
| Arsirii et al. | Models and methods of intellectual analysis for medical-sociological monitoring’s data based on the neural network with a competitive layer | |
| CN112528038B (en) | Method, device and medium for ensuring communication body uniqueness identification based on multilayer graph structure | |
| CN114816321A (en) | Unsupervised method for determining importance of user perception index | |
| CN113191569A (en) | Enterprise management method and system based on big data | |
| Dai et al. | Intelligent audit question answering system based on knowledge graph and semantic similarity | |
| CN115687632B (en) | Criminal investigation plot decomposition analysis method and system | |
| CN114548325B (en) | Zero sample relation extraction method and system based on dual contrast learning | |
| Ma et al. | An Automatic Extraction Method of Rebar Processing Information Based on Digital Image | |
| CN116796723B (en) | Text set matching method and device, electronic equipment and storage medium | |
| CN112541085B (en) | Method for structuring questionnaire, apparatus for structuring questionnaire, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20220412 Assignee: Liantai Kechuang (Tianjin) Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2023120000051 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20231222 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20220412 Assignee: Tianjin Liankai Technology Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2023980054259 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20231227 |
|
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20220412 Assignee: Tianjin United Chuangxing Technology Development Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2023980054801 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20240102 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20220412 Assignee: Yunneng Zhongchuang Investment Management (Tianjin) Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2024980000234 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20240105 Application publication date: 20220412 Assignee: Tianjin 1919 Technology Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2024980000180 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20240104 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20220412 Assignee: Tianjin Xingdu Space Network Technology Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2024980000344 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20240109 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20220412 Assignee: Tianjin Huihe Investment Management Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2024980000705 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20240112 Application publication date: 20220412 Assignee: TIANJIN YONGRUI BUSINESS INFORMATION CONSULTATION CO.,LTD. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2024980000704 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20240112 Application publication date: 20220412 Assignee: Gongneng Enterprise Management Consulting (Tianjin) Co.,Ltd. Assignor: Tianjin Lianchuang Technology Development Co.,Ltd. Contract record no.: X2024980000518 Denomination of invention: A method, system, and storage medium for intelligent diagnosis of enterprise requirements Granted publication date: 20220624 License type: Common License Record date: 20240111 |
|
| EE01 | Entry into force of recordation of patent licensing contract |