TWI767192B - Application method of intelligent analysis system - Google Patents

Abstract

The present invention relates to an application method of an intelligent analysis system, the steps of which include: retrieving an open data from a server, and obtaining a closed data from a client; a processing unit of the open data and the closed data pre-processing the data, and output a first data source and a second data source; store the first data source and the second data source in a database; analyze the first data source and the second data source, and output an analysis result; output a chart, a model or a table on an interface according to the analysis result, the user can realize intelligent operation and management by means of the chart, the model or the table.

Description

Application method of intelligent analysis system

The present invention relates to an application method of an analysis system, especially an application method of an intelligent analysis system by collecting, front-end analysis, in-depth analysis and outputting analysis results.

With the development of network technology, artificial intelligence has become the most popular industry nowadays. The combination of artificial intelligence with traditional industries can greatly help the data processing of traditional industries. Artificial intelligence (AI) is a field of computer science that addresses common cognitive problems associated with human intelligence, such as learning, problem solving, and pattern recognition.

Similarly, the related fields of "Business Intelligence & Analytics" (BI&A, Business Intelligence & Analytics) have become more and more important in business or academia, and there have been major breakthroughs and developments in many industrial studies. Therefore, in addition to obtaining information, how to store and analyze data is also a technology that needs to be broken through and developed. In short, the place where data is stored is the work to be achieved by a general database, and further targeted at a group of related data in the database. analysis. The data stored in the database can generally be divided into open data and closed data, that is, open data can be accessed by the public, while closed data is restricted to private companies or government agencies.

The applications of open data generally open to the outside world are mainly non-text data materials, such as maps, gene sequences, neural network bodies (or Connections), chemical molecular data, mathematical functions, scientific formulas, medical data and Applications, life sciences, biodiversity and other public information without privacy restrictions and without copyright restrictions, for private enterprises or government agencies, the open information provided by them without access restrictions is also open information, but also There are closed data for internal access only, for example: the results of the business activities of private enterprises are basically based on fixed reports and other models. The business report stage is the initial stage of data analysis. With the emergence of database technology, especially today, enterprises have begun to build information, and the informationization of business processes has precipitated a large amount of digital business data. In fact, everyone has always had the need for data analysis. Now that there is data precipitation, through these The demand for data reporting statistics and data analysis naturally arises.

However, as the database stores a large amount of digitized data, it is inevitable that there will be redundant data or too detailed data, which will take a long time to analyze the data. Data is the research direction of the present invention.

In view of the above problems, the present invention provides an application method of an intelligent analysis system, and combines the modern technologies such as artificial intelligence analysis and deep learning to quickly convert the above-mentioned data into practical technical data.

An object of the present invention is to provide an application method of an intelligent analysis system, by analyzing the internal and external structured data or semi-structured data of an enterprise, after data extraction, conversion and loading, using artificial intelligence analysis technology, the The knowledge of enterprise operation and management is externalized into a set of operation management system to help managers operate intelligently.

In order to achieve the above object, the present invention discloses an application method of an intelligent analysis system, which includes the following steps: a processing unit of a server retrieves an open data, and obtains a closed data from a client; the processing The unit performs pre-processing on the open data and the closed data to generate a first data source and a second data source; and stores the first data source and the second data source in a data source in the database; the processing unit executes an artificial intelligence program and analyzes the first data source and the second data source according to at least one analysis parameter to generate an analysis result; according to the analysis result, a graph and a model are displayed on a user interface Or a table, the user realizes intelligent operation and management through the diagram, the model or the table.

In an embodiment of the present invention, it is further disclosed that the pre-processing includes data cleaning, missing value processing, and normalization processing.

In an embodiment of the present invention, it further discloses performing Extract-Transform-Load (ETL) processing on the open data and the closed data.

In an embodiment of the present invention, it further discloses using a PDI (Pentaho Data Integration) tool to analyze the open data and the closed data.

In an embodiment of the present invention, it further discloses performing statistical induction, regression analysis, decision tree analysis and deep learning on the first data source and the second data source.

In an embodiment of the present invention, it is further disclosed that the deep learning is Long Short-Term Memory (LSTM).

In an embodiment of the present invention, it further discloses that the open data is economic index open data or industry financial document data.

In an embodiment of the present invention, it further discloses that the closed data is structured data or semi-structured data.

In an embodiment of the present invention, it is further disclosed that the model includes a regression model.

In order to make your examiners have a further understanding and understanding of the features of the present invention and the effects achieved, I would like to assist with the detailed descriptions of examples and cooperation, and the descriptions are as follows:

The present invention is an application method of an intelligent analysis system, which captures open data on the server and obtains closed data on the client, and then analyzes the above-mentioned data, after data extraction, conversion and loading, the application Artificial intelligence analysis technology externalizes enterprise management knowledge into a set of management system to help managers operate intelligently.

Below, the application method of a kind of intelligent analysis system of the present invention will be further described:

Please refer to FIG. 1, which is a flowchart of an intelligent analysis system of the present invention. As shown in the figure, an application method of an intelligent analysis system of the present invention includes the following steps:

S1: Capture open data and obtain closed data;

S3: Pre-processing open data and closed data;

S9: Generate a first data source and a second data source;

S11: store the first data source and the second data source;

S13: analyze the first data source and the second data source;

S17: output the analysis result;

S19: Display a chart, model or table on the user interface according to the analysis result.

Next, in order to achieve an intelligent analysis system of the present invention, please refer to FIG. 1 and FIG. 2 together, which are schematic diagrams of the intelligent analysis system of the present invention. As shown in the figure, an intelligent analysis system of the present invention includes: a server 10 , a client 20 , a processing unit 30 , a database 40 and an operation interface 50 . The server 10 and the client 20 are connected to the processing unit 30 , the database 40 is connected to the processing unit 30 , and the operation interface 50 further includes a display unit for displaying an analysis result of the database 40 . Wherein, the server 10 is a web server, and an open data is stored on the server 10 , the client 20 is an internal client of an enterprise, and a closed data is stored on the customer list 20 . The specific steps of an intelligent analysis system of the present invention are as follows:

As shown in step S1, an open data is retrieved, and a closed data is obtained. In this step, firstly, an open data is retrieved from the server 10 , and a closed data is obtained from the client 20 . Among them, the open data is related data such as economic indicators open data or industry financial documents. The collection methods of the open data include cross-platform data connection and conversion, open data collection and retrieval, and web crawler components written in the Python programming language. the open data. The collection of closed data is the data provided by customers, which is structured data or semi-structured data. The closed data is required only after the data has undergone the data quality check procedure.

As shown in step S3, preprocessing is performed on the open data and the closed data. In this embodiment, after the step of retrieving an open data and obtaining a closed data, it further includes a step S5: extract-transform-load the open data and the closed data. -Load, ETL) processing, in other words, extract-transform-load processing is the process of extracting, transforming, and loading data from the source to the destination. That is to say, after the step of retrieving an open data and obtaining a closed data, first do some simple analysis work, the preprocessing includes data cleaning, missing value processing, normalization processing, etc., for example: 1. Extract: read data, the data source may be txt, excel or database; 2. Transform: Analyze the data, such as verifying whether the data format is correct; 3. Load: The processed data may be exported into txt, excel, written or updated to the specified database.

Examples of open data are shown in Table 1 below: Function Primary external data source (Input) Processing method (Process) Output (Output) External statistical indicators: l comprehensive economic indicators, l labor wages and productivity, l industrial operation indicators l National Development Council Prosperity Index Inquiry System l Yahoo Stock Market Observation Station l Juheng Network l ROC Statistical Information Network 1. Data grouping and classification 2. Statistical operation 3. Chart presentation According to the characteristics of the indicators, it is presented to the user in the form of a suitable chart, which aims to make the user understand the performance indicators of other companies in the industry and serve as a reference for the company's operation. Industry News: l Summary Analysis, l Keyword Analysis l Play stock fun l MoneyDJ l Juheng.com Industry News Network l Taiwan Economic Research Institute 1. Text preprocessing analysis 2. Keyword frequency analysis 3. Abstract text mining The artificial intelligence algorithm is used to calculate the important keywords and abstract content of the collected external industry news, so that users can quickly grasp the external information related to the company's business. Trend analysis: l exchange rate trends, l commodity prices l Yingwei Finance l (investing.com) l Srock-AI l Finance M Square l (MacoMicro) 1. Pre-analysis of numerical data 2. Multivariate dimension analysis 3. Regression/DL model For the raw materials or exchange rates related to the high cost of the enterprise, the basic forecast of the future trend is provided for the enterprise to use as a reference for financial planning, sales or purchasing decisions. Current Situation Analysis l Internal information of the company l External information of the first three items above 1. Knowledge base (KB) model 2. Time series data analysis 3. Statistical analysis 4. Pattern analysis The analysis results obtained from the company's internal data and external data are comprehensively analyzed to give users a basic analysis of the current situation, especially reminders of some abnormal phenomena. future forecast l Internal data of the company l Results of the analysis of the first three external data above 1. Knowledge base (KB) model 2. Time series data analysis 3. Statistical analysis 4. Pattern analysis Conduct exploration and analysis on the future forecast results of the company's internal data and external data, and make future forecasts for the company's operation and management indicators. If abnormal phenomena are found, the operator can be warned in advance. Table I

Examples of open data of private companies are shown in Table 2 below: Comprehensive Prosperity Indicator leading indicator Leading Indicators Composite Index, Export Order Trend Index, Currency Aggregate M1B, Stock Price Index, Industrial and Service Employee Net Entry Rate, Floor Area under Construction, Semiconductor Equipment Import Value Simultaneous indicators At the same time indicators composite index, industrial production index, industrial production index, electricity (enterprise) total electricity consumption, manufacturing sales volume index, employment in the non-agricultural sector, customs export value, import value of machinery and electrical equipment lagging indicator Composite index of lagging indicators, unemployment rate, labor cost index per unit output of manufacturing industry, overnight lending rate in the financial industry, lending and investment of all financial institutions, manufacturing inventory value wage labor productivity Salary level Total monthly salary per person, monthly recurring salary per person, non-recurring monthly salary per person Human growth Entry rate, exit rate, turnover rate productivity cost Labor productivity index, unit output labor cost index Industrial management indicators Profitability Turnover per share, after-tax earnings per share, after-tax gross profit margin, pre-tax net profit ratio, after-tax net profit ratio, return on assets business performance Revenue growth rate, operating profit growth rate, pre-tax net profit growth rate, after-tax net profit growth rate Management capacity Inventory Turnover, Accounts Payable Turnover, Receivables Turnover, Total Assets Turnover financial structure Interest coverage ratio, current ratio, quick ratio solvency leverage ratio, debt ratio Table II

Examples of closed data are shown in Table 3 below: Production Output, production yield, per capita productivity, inventory amount, inventory age, inventory turnover days, overdue production ratio analysis, outsourcing delivery delay ratio analysis Sales Total revenue, customer sales ranking, product sales ranking, shipment delay analysis HR Staff turnover rate, salary cost rate, human output value research and development R&D expense rate, plan completion rate finance Gross Margin, Accounts Payable Turnover, Accounts Receivable Turnover, Debt Ratio, Operating Expense Ratio Table 3

In addition, step S7 is further included: use Pentaho Data Integration (PDI) to process the open data and the closed data. The so-called Pentaho Data Integration (PDI) is a Spoon-based data integration development environment. Pentaho data integration support is deployed on a single host in a cloud server architecture or a distributed cluster server architecture.

As shown in step S9, a first data source and a second data source are output. Wherein, after preprocessing the open data and the closed data, the first data source and the second data source are output.

As shown in step S11, the first data source and the second data source are stored. Wherein, after the first data source and the second data source are output, the first data source and the second data source are stored in the database 40, in other words, the open data and the closed data are processed before After processing, only the valuable information is retained, and the above-mentioned information is stored for further in-depth analysis to obtain more valuable information.

As shown in step S13, the first data source and the second data source are analyzed. In this step, the required data is first extracted from the database 40 or the next stage analysis is performed directly in the database. After step S13, step 15 is further included: performing statistical induction, regression analysis, decision tree analysis and deep learning on the first data source and the second data source. Wherein, in this embodiment, the deep learning is an analysis method such as Long Short-Term Memory (Long Short-Term Memory, LSTM). However, this method of analysis is not limited.

As shown in step S17, an analysis result is output. Wherein, in this step, after further analyzing the first data source and the second data source, the analysis result is output.

As shown in step S19, a graph, a model or a table is output on an operation interface according to the analysis result. Wherein, in this embodiment, the graph is a statistical graph, and the model is a regression model, etc., but not limited thereto.

In one embodiment, the analysis of the first data source and the second data source is based on an indicator characteristic, and the analysis result is output in a suitable chart form, so that the user can understand the performance indicators of other companies in the industry, and use them as the company's performance indicators. business reference.

In another embodiment, an artificial intelligence algorithm is used to calculate the important keywords and abstract content of the first data source of the open data, and provide the user with relevant external information.

In another embodiment, the analysis of the first data source and the second data source is to provide a basic forecast of future trends for related raw materials or exchange rates with higher costs for the enterprise, which is used by the enterprise for financial planning, sales or procurement. reference for decision-making.

In another embodiment, prospecting and analyzing the future forecast results of the closed data and the open data of the company are performed to make future forecasts for the company's operation and management indicators, and an early warning message is sent to the operator if abnormal phenomena are found.

To sum up, an intelligent analysis system of the present invention captures the open data on the server and obtains the closed data from the client, and then analyzes the above-mentioned data, through data extraction, conversion, After loading, it is stored in the database, and secondly, artificial intelligence analysis technology is used to externalize the enterprise management knowledge into a set of management system to help managers operate intelligently.

Therefore, the present invention is indeed novel, progressive and available for industrial use, and it should meet the requirements of patent application in my country's patent law.

However, the above descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of implementation of the present invention. All changes and modifications made in accordance with the shape, structure, features and spirit described in the scope of the patent application of the present invention are equivalent. , shall be included in the scope of the patent application of the present invention.

10: Server 20: Client 30: Processing unit 40:Database 50: Operation interface S1~S19: Steps

Figure 1: It is a flow chart 1 of the application method of the intelligent analysis system of the present invention; and Figure 2: It is a schematic diagram of the application method of the intelligent analysis system of the present invention.

10: Server

20: Client

30: First App

40: Second application

50: Second Authentication Mechanism

Claims (8)

An application method of an intelligent analysis system, the steps of which include: a processing unit of a server fetches an open data, and obtains a closed data from a client, wherein the server is a web server; the processing unit Perform a pre-processing on the open data and the closed data to generate a first data source and a second data source; the processing unit stores the first data source and the second data source In a database; the processing unit executes an artificial intelligence program and analyzes the first data source and the second data source according to at least one analysis parameter, and generates an analysis result, which further includes the artificial intelligence program on the first data perform statistical induction, regression analysis, decision tree analysis and deep learning on the source and the second data source; and display a graph, a model or a graph on a user interface according to the analysis result sheet. The application method of the intelligent analysis system as described in item 1 of the patent scope, wherein the pre-processing includes data cleaning, missing value processing, and normalization processing. The application method of the intelligent analysis system as described in item 1 of the patent scope, wherein in the step of preprocessing the open data and the closed data in the processing unit, it further comprises the step of: the open data and the closed data Extract-Transform-Load (ETL) processing of sexual data. The application method of the intelligent analysis system as described in item 3 of the patent scope, wherein in the steps of extract-transform-load (ETL) processing of the open data and the closed data, More steps are included: Use PDI (Pentaho Data Integration) tool to analyze the open data and the closed data. The application method of the intelligent analysis system as described in item 1 of the patent scope, wherein the deep learning is long short-term memory (Long Short-Term Memory, LSTM). According to the application method of the intelligent analysis system described in item 1 of the patent scope, the open data is open data of economic indicators or data of industrial financial documents. The application method of the intelligent analysis system as described in item 1 of the patent scope, wherein the closed data is structured data or semi-structured data. The application method of the intelligent analysis system as described in item 1 of the patent scope, wherein the model includes a regression model.

Images