TW202113618A - Method for checking output data with expert system that generally applies piece by piece comparison in an intelligent item-listing fashion to data output from an information system - Google Patents

Abstract

A method for checking output data with an expert system generally applies piece by piece comparison in an intelligent item-listing fashion to data output from an information system, namely using expert knowledge files built in an expert knowledge database of an expert system to automatically check if each piece of output data possesses correctness, integrity, effectiveness, undeniability, and security, so that those exhibiting correctness, integrity, effectiveness, undeniability, and security are allowed to output. Those do not possess one of correctness, integrity, effectiveness, undeniability, and security must be automatically corrected by the computer or manually corrected before output. The expert system also includes an automatic learning function so as to perform analysis on all the output data and information to automatically generate data of a corresponding expert knowledge file to be included in the expert knowledge database to thereby greatly reduce the time and cost for manual maintenance of the expert knowledge files.

Description

Method of checking output data with expert system

The present invention relates to a method for checking whether the output data is correct, especially a method for checking the output data with an expert system, and has an automatic learning function, which can automatically generate relevant expert knowledge files by analyzing all the output data and data And join the expert knowledge database, which greatly reduces the time and cost of manually maintaining expert knowledge.

Press, the output of computer data is a very important verification program in the development and maintenance of information systems. For the results of data output, the correctness, completeness, validity, non-repudiation and safety of the data should be verified. Performance, such as the total number of posts, the confirmation rate, the number of replies... and other output checks. Information system development should also consider the form of data output, and only authorized personnel should be allowed to access the data output by the application system.

Within a unit, there may be various heterogeneous information systems (take the teaching unit as an example, that is, it will also have educational, academic, general, accounting, research and development, personnel, library, environmental safety, or school safety... and other systems) coexist In order to check the information output by different information systems, program developers need to develop different data output check programs individually, which not only increases the difficulty of program development and maintenance, but also increases the risk of computer processing program errors and maintenance costs.

However, the output of data using manual verification is slow, and human errors often occur, verification and correction are slow, and it is not easy to be completely correct.

In addition, due to the advancement of information technology, when newer data output methods and verification and verification technologies appear, if the corresponding data output program cannot be modified automatically or the verification function can be added with the times, the output data cannot be increased. safety. For example, face recognition technology can be used to determine the legal and reasonable identity of the person requesting the output of the data, the IP address of the output device can also be used to determine whether the geographic information of the data output is in a legal and reasonable area, and the above-mentioned verification methods can be used as data verification. One of the security mechanisms.

In view of this, how to solve the above-mentioned problems is the primary problem to be solved by the present invention.

The main purpose of the present invention is to solve the above-mentioned problems and provide a method for checking output data with an expert system. The original computer hardware and information system software do not need to be changed, and the present invention is only connected to the output end of the information system. The provided expert system can use its artificial intelligence checking software and expert knowledge database with expert knowledge files in various fields to achieve an intelligent output data checking mechanism to safely avoid "computerized" or "automated" system output failure The phenomenon of data, and the result of checking the output data of the information system through its artificial intelligence checking software and expert knowledge database, can more ensure its correctness, completeness, validity, non-repudiation and safety.

To achieve the foregoing objective, the present invention provides a method for checking output data with an expert system, the steps of which include: a. Use an expert system to check the data prepared for output by an information system. The expert system includes a rule-based statement (Statements) method to establish an expert knowledge database with expert knowledge files in various fields, and multiple Artificial intelligence checking software that can check data of different content levels; b. The expert system selects the corresponding artificial intelligence check software based on the content level of the data to be output, and reads the expert knowledge file corresponding to the expert knowledge database with the selected artificial intelligence check software, and compares the rules in a row Ways to check and compare; c. The expert system judges whether the data to be output is correct, complete, valid, non-repudiation, and safe according to the method of tabular comparison rules, and has correctness, completeness, validity, non-repudiation and safety When the time, its output is allowed.

It is not difficult to obtain an in-depth understanding of the above-mentioned objectives and advantages of the present invention from the detailed description of the selected embodiments and the accompanying drawings.

Please refer to Figures 1 to 3, the figures shown in the figures are the preferred embodiments of the present invention, which are for illustrative purposes only, and are not limited by the embodiments in the patent application.

The present invention provides a method for checking output data with an expert system, which mainly uses an expert system 1 to check data and data (hereinafter collectively referred to as data) prepared for output by an information system 2 (such as a computer). The expert system 1 includes a Based on the Rule-Based Statements method, an expert knowledge database 12 with expert knowledge files 121 in various fields is established, and multiple artificial intelligence checking software 11 that can check data of different content levels. The expert system 1 can automatically check the correctness, completeness, validity, non-repudiation and safety of the data prepared by the information system 2 to avoid illegal, unreasonable and unreasonable output data Undesirable phenomena, including:

As shown in Figure 1, the output data received by the expert system 1 can be used by various information systems 2 (take the school unit as an example, which can include educational affairs, academic affairs, general affairs, accounting, research and development, personnel, library , Environmental Safety or School Safety... etc.), the Expert System 1 (Expert System) is an artificial intelligence execution software built in a computer, which can store the required expert knowledge in a form that the computer can understand, and Adding a control strategy allows the computer to use this knowledge and rules of experience to solve problems like an expert. In other words, the expert system 1 is a knowledge-based program that can be used to solve specific domain problems and can provide "professional-level" answers like human experts. Moreover, the expert system 1 is based on the Rule-Based Statements method to establish its collational comparison rules. The core of the expert system 1 includes the construction and application of reasoning similar to or beyond humans (logical logic). ), prediction, knowledge and experience storage, automatic learning, environment sensing, speech, text and graphic action recognition, machine perception, intuition and imagination, automatic control... and other software and hardware.

Each artificial intelligence checking software 11 (AI APP) is built in the expert system 1 based on its rule-based model. Each artificial intelligence checking software 11 corresponds to a professional field and can be used to check the output data to ensure Data correctness, data integrity, data validity, data non-repudiation, data security, and can automatically correct wrong output data.

The expert knowledge database 12 can be one of a cloud database or a local area network database, and can store related expert knowledge files 121 automatically or manually. The expert knowledge files 121 stored in the expert knowledge database 12, According to various industries, fields, national conditions, etc., it is applicable to specific fields in the form of statements (Statements). The expert knowledge file 121 contains various knowledge and experience in various industries, daily life, military, politics, etc. Analyze and summarize logically according to lawfulness, reasonableness, sensibility, time, place, people and people; for example, when the data to be output is related to school affairs, the legal provisions, administrative orders, and rules of operation in the expert knowledge database 12 , It can be used by the corresponding artificial intelligence checking software 11, and the prepared output data will be checked and compared one by one to determine whether the output data is legal or not. , Reasonable and sensible.

Furthermore, the expert knowledge database 12 has a positive data file 122 (Clean File) storing good and usable data examples and a negative data file 123 (Dirty File) storing bad and unusable data examples. And through the verification result of the expert system 1, if the output is illegal, unreasonable, or unreasonable (not in accordance with the statute), it will be excluded from use and saved in the negative data file 123; if the output is legal, reasonable, and sensible The output command of the information system 2 is automatically executed and stored in the front data file 122 for future use. When the expert system 1 executes the artificial intelligence checking software 11, each artificial intelligence checking software 11 will read the corresponding expert knowledge file 121, the positive data file 122, and the negative data file 123 in the expert knowledge database 12, and list them together. Column comparison (ie comparison one by one) rule method, one by one to determine whether the output data violates the rules, if the rules are violated, the data is determined to be illegal, and the illegal data is stored in the negative data file 123 for record, and the process ends . If none of the checked data violates the relevant rules, it means that the data is judged to be legal through the automatic checking procedure of the expert system 1 and can be output to an output system 3 (such as the output of a screen, a printer, a wireless transmission device, etc.) Equipment), and by this, the expert system 1 can use various artificial intelligence checking software 11 to ensure the correctness, completeness, validity, non-repudiation and safety (legal, reasonable, and reasonable) of the output data to protect the computer The correct and reliable AI Upgrade method is output.

In addition, the expert system 1 converts the checked data into positive and negative rule statements, and automatically saves them in the positive data file 122 and the negative data file 123, so that the expert system 1 can use the analyzed data to automatically Amendment (that is, the function of automatic learning) and expand the new knowledge to its expert knowledge database 12, which greatly reduces the burden of time and cost of manually maintaining expert knowledge data. This means that the monitoring data can be fed back to the expert knowledge database 12 at the output end of the existing information system 2 and during execution, and new expert knowledge clauses can be automatically revised and expanded to the expert knowledge database 12.

And the data prepared to be output by the information system 2 may include: text, numbers, images (movies), graphics, photos, sounds, voices, etc., and when the output data is voice and voice, the correctness of the data has been checked, After the integrity, validity, non-repudiation and security (legitimate, reasonable, and reasonable) and the legality of the data recipient, it can be directly connected to the computer loudspeaker after natural voice recognition. Furthermore, in order to warn illegal persons or illegal or unreasonable data output, the artificial intelligence checking software 11 of the expert system 1 will also automatically issue an early warning notice when illegal persons or illegal or unreasonable persons are found. The AI APP with face recognition or fingerprint technology verifies the identity, and creates a separate notification file and sends it to the relevant authorized user of the information system 2, so that the relevant authorized user understands the occurrence of this problem and the improvement method.

Since the artificial intelligence checking software 11 built in the expert system 1 has many types, various feasible AI APPs are listed below for explanation:

As shown in Figure 1, the main function of the permission check AI APP is to ensure that only authorized users can request data output, whether the content of the data and the requesting data exporter are legal, reliable, and credible units or persons, especially whether Particular attention should be paid to blacklisted objectionable information content and information requested by users or prohibited from outputting to the computer.

The main function of the time check AI APP is to check the safety and reliability of the information at the time of information generation and output, and determine whether to allow the output of its data.

The main function of the environmental check AI APP is to check whether the information output environment will affect the accuracy of the information.

The main function of the AI APP is to check whether the content of the information to be output violates laws, rules and regulations, so as to avoid legal liabilities or illegal acts caused by improper output.

The experience check AI APP (historical experience success check AI APP) is used to target the information to be output, the implementation of the past processing results, and the experience that caused the failure, to reject or remind the operator to pay attention to the information.

The geographic check AI APP is used to check whether the information occurred and the place mentioned is harmless and correct.

The SOP check AI APP uses standard operations to check the information to be output, the required operation methods and procedures, and whether there is any violation of the standard procedures. Special instructions to the operator are required, and the information system is allowed to perform output only after special approval. instruction.

And the command source check AI APP is used to check whether the computer system and equipment required by the data content are legal, reliable, and credible, especially whether the bad information system and computer hardware equipment are blacklisted, and special attention should be paid to or prohibited from outputting to the computer.

The purpose/target check AI APP is used to check whether the purpose or target of the data output by the computer is legal, and will not be used for illegal or unethical use.

The current status check AI APP is used to check whether the data output by the computer is consistent with the person, event, material, and environmental data before it can be output.

It is hoped that the result check AI APP can be used to check the desired output result and whether the output data can be satisfied.

Next, as shown in Figure 2, the method for checking output data with an expert system provided by the present invention includes the following steps:

a. Use an expert system 1 to check data prepared to be output by an information system. The expert system 1 includes a Rule-Based Statement method to establish an expert knowledge database 12 with expert knowledge files 121 in various fields. , And multiple artificial intelligence checking software 11 that can check data of different content levels respectively;

b. The expert system 1 selects the corresponding artificial intelligence check software 11 according to the content level of the data to be output, and uses the selected artificial intelligence check software 11 to read the expert knowledge file 121 corresponding to the expert knowledge database 12, and then write Column comparison (ie comparison one by one) rule-based check and comparison;

c. The expert system 1 judges whether the data to be output is correct, complete, valid, non-repudiation, and safe according to the method of tabular comparison rules, and has correctness, completeness, validity, non-repudiation and safety When it is sexual, the output is allowed, and the output is prohibited if one of the correctness, completeness, validity, non-repudiation and safety is not available.

In addition, please refer to Fig. 3, which is an explanation of the steps of the present invention when actually checking the output data: First, the expert system 1 will first determine the content level of the output data prepared by the information system, and determine the content level according to the content level. It is different to decide which artificial intelligence checking software 11 to execute, so not every kind of output data needs to be checked by all artificial intelligence checking software 11, which can save checking time.

Then, after the expert system 1 selects the artificial intelligence checking software 11 to be executed, the expert system 1 will execute the artificial intelligence checking software 11 one by one. If one of the artificial intelligence checking software 11 checks that the output data is illegal, the illegal data will be stored in the negative data file 123, an early warning notice of the problem information will be issued and a separate notification file will be created, and artificial intelligence can be used The verification software 11 displays similar registration data in the expert knowledge database 12, and the warning notification is sent to the relevant authorized user of the information system in a wired or wireless manner. The transmission object can be a preset smart phone in the group or each output Equipment and computers, so that users with relevant permissions can understand the occurrence of this problem in the first time, and the data that has been checked as illegal can be corrected automatically or manually through the expert system 1 before it is allowed to correct, safe and good information. Output.

As shown in Figure 3, when the expert system 1 of the present invention executes the artificial intelligence checking software 11, the artificial intelligence checking software 11 will read the corresponding expert knowledge data file 121 located in the expert knowledge database 12. Data file 122, negative data file 123, and check the rules one by one to determine whether the output data violates the rules. If the rules are violated, the data is judged to be illegal, and the artificial intelligence check software 11 will store the illegal data into the negative data. File 123 to keep records. If the relevant artificial intelligence checking software 11 consistently determines that the data is legal, the legal data is stored in the positive data file 122 and the data can be output from the set output system 3 to end the process.

Therefore, whether it meets or does not meet the results of the expert system 1, it can be converted into positive and negative rule statements, and automatically stored in the positive data file 122 and the negative data file 123, so that the expert system 1 can learn and accumulate experience by itself . Of course, when an item with errors is checked through the expert system 1, it can also be checked by manual intervention to determine whether the check of the expert system 1 is incorrect. If the information content of the data is found to be unavailable after manual intervention and review When an error occurs, this situation can be stored in the positive data file 122 to expand the positive instantiation of the expert system 1 and improve the correctness of the expert system 1's subsequent checking procedures.

From the above description, it is not difficult to find that the advantages of the present invention are:

1. With the item-by-item check and output data method of the present invention, illegal, unreasonable and unreasonable data can be effectively filtered out, so as to avoid outputting erroneous data leading to undesirable results, so it can effectively improve the accuracy of system output data. Integrity, validity, non-repudiation and safety (legal, reasonable and reasonable).

2. The expert system provided by the present invention also has an automatic learning function, which can automatically modify and expand the new knowledge to the expert knowledge database by analyzing the output data, so as to reduce the burden of manual maintenance of expert knowledge.

3. The artificial intelligence check software (AI APP) of the present invention is mainly implemented in software, and can expand functions and add new check AI APPs at any time, in order to respond to the new information output check in the future, and increase the flexibility of the entire system And reliability.

However, the disclosure of the above embodiments is only used to illustrate the present invention, not to limit the present invention. Therefore, all changes in numerical values or replacement of equivalent elements should still belong to the scope of the present invention.

To sum up, when people who are familiar with this technique can understand that this invention can clearly achieve the aforementioned purpose, it has actually complied with the provisions of the Patent Law, so they file an application in accordance with the law.

1: Expert system 11: Artificial intelligence checking software 12: Expert knowledge database 121: Expert Knowledge File 122: positive data file 123: Negative data file 2: Information system 3: output system

Figure 1 is a system architecture diagram of the expert system provided by the present invention. Figure 2 is a block diagram of the steps of the method of the present invention. Figure 3 is a flow chart of the verification operation of the present invention.

1: Expert system

11: Artificial intelligence checking software

12: Expert knowledge database

121: Expert Knowledge File

122: positive data file

123: Negative data file

2: Information system

3: output system

Claims (5)

A method for checking output data with an expert system, the steps include: a. Use an expert system to check the data prepared for output by an information system. The expert system includes a rule-based statement (Statements) method to establish an expert knowledge database with expert knowledge files in various fields, and multiple Artificial intelligence checking software that can check data of different content levels; b. The expert system selects the corresponding artificial intelligence check software based on the content level of the data to be output, and reads the expert knowledge file corresponding to the expert knowledge database with the selected artificial intelligence check software, and compares the rules in a row Ways to check and compare; c. The expert system judges whether the data to be output is correct, complete, valid, non-repudiation, and safe according to the method of tabular comparison rules, and has correctness, completeness, validity, non-repudiation and safety When the time, its output is allowed. The method for checking output data with an expert system as described in claim 1, wherein the expert knowledge database further has a positive data file with examples of good and usable data and a negative data file with examples of bad and unusable data. , And each artificial intelligence checking software will read the corresponding expert knowledge files, negative data files, and positive data files in the expert knowledge database, and determine whether the output data violates the rules one by one by means of a list of comparison rules. If the rules are violated, the data is judged to be illegal and the illegal data is stored in the negative data file. If the checked data does not violate the relevant rules, the data is judged to be legal and allowed to be exported. The method for checking output data with an expert system as described in claim 2, in which, when one of the artificial intelligence checking software checks that the output data is illegal, an early warning notification of one of the problem information will be issued and a separate notification file will be created. It can also display similar registration data in the expert knowledge database through artificial intelligence checking software, and the warning notification is sent to the relevant authorized user of the information system in a wired or wireless manner. The method for checking output data with an expert system as described in claim 3, wherein the sending object of the early warning notification can be a preset smart phone or each output device and computer in the group. The method for checking output data with an expert system as described in claim 3, in which data that has been checked as illegal can be automatically corrected or manually corrected to correct, safe and good information through the expert system, and then the output can be allowed.

Images