CN110377526A - A kind of test method of AI intelligence software - Google Patents
A kind of test method of AI intelligence software Download PDFInfo
- Publication number
- CN110377526A CN110377526A CN201910699230.4A CN201910699230A CN110377526A CN 110377526 A CN110377526 A CN 110377526A CN 201910699230 A CN201910699230 A CN 201910699230A CN 110377526 A CN110377526 A CN 110377526A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- test
- ability
- data
- intelligence
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3684—Test management for test design, e.g. generating new test cases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3688—Test management for test execution, e.g. scheduling of test suites
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The invention discloses a kind of test methods of AI intelligence software, comprising the following steps: S1: establishes model；S2: framework is established；S3: supplement；S4: table is built；S5: library is built；S6: verifying；S7: test；S8: it calculates；S9 calculates verifying；S10: data result is drawn.The present invention is by establishing intelligent system model, form unification, the test mode for the AI intelligence software that can be compared to each other, it can be by data by detection, information and knowledge are input to the acquisition capability of detection system knowledge in intelligence system, by the grasp ability for checking the capacity check systematic knowledge of intelligence system knowledge base, by checking intelligence system can be by how many data, information and Knowledge conversion are new knowledge base content to detect the innovation ability of the systematic knowledge, test operation is convenient and efficient, by carrying out data verification before testing, it can prevent the exception occurred due to data, it is verified by calculating, guarantee the accuracy of test data, it is suitble to be widely popularized.
Description
Technical field
The present invention relates to AI intelligence software technical field, specially a kind of test method of AI intelligence software.
Background technique
Artificial intelligence (Artificial Intelligence), english abbreviation AI.It is research, develop for simulating,
Extend and the theory of the intelligence of extension people, method, a new technological sciences of technology and application system.
Artificial intelligence is a branch of computer science, it attempts to understand essence of intelligence, and is produced a kind of new
The intelligence machine that can be made a response in such a way that human intelligence is similar, the research in the field include robot, language identification, image
Identification, natural language processing and expert system etc..Artificial intelligence is since the birth, and theory and technology is increasingly mature, application field
Also constantly expand, it is contemplated that the following artificial intelligence bring sci-tech product, it will be the wisdom of humanity " container ".Artificial intelligence
Can consciousness to people, thinking information process simulation.Artificial intelligence is not the intelligence of people, but can think deeply as people,
It may also be more than the intelligence of people.
In AI intelligence software test process, now lack a kind of unified test model, not only test is chaotic, but also tests
It is inconvenient, it is easy occur abnormal conditions because of data in test process, the accuracy of test is also relatively low.
Summary of the invention
The purpose of the present invention is to provide a kind of test method of AI intelligence software, formed it is unified, can mutually be compared
Compared with AI intelligence software test mode, test operation is convenient and efficient, by carrying out data verification before testing, can prevent by
In the exception that data occur, is verified by calculating, guarantee the accuracy of test data, to solve to propose in above-mentioned background technique
The problem of.
To achieve the above object, the invention provides the following technical scheme: a kind of test method of AI intelligence software, including with
Lower step:
S1: establishing model, by historical data training and establishes intelligent system model, and issues prediction service；
S2: establishing framework, AI intelligence software test structure is established with reference to von Karman framework, by people, machine and people
Work intelligence system is showed with framework；
S3: supplement according to existing knowledge, finds new knowledge element and new rule, feeds them into memory, supplies
Computer and controller use, and by input/output and external progress Knowledge interaction, the outside for carrying out knowledge sharing is known
Know library or cloud storage；
S4: building table, establishes internet works software IQ appraisement system according to the feature main points of intelligent system model, is formed artificial
Intelligence software IQ test scale；
S5: building library, according to artificial intelligence software's IQ test scale is formed, establishes internet works software IQ test library；
S6: verifying carries out data verification before testing, and verifying system receives the number from various sources and various formats
According to whether the taken in data of verifying are converted according to the desired format of goal systems；
S7: test, for AI intelligence software and human subject, when answering exam pool, according to the answer result of feedback
As judge object；
S8: calculating, and carries out software according to the absolute IQ calculation formula and AI deviation IQ formula of AI intelligence software test
Measuring and calculation；
S9: verifying is calculated, basic data seven or three is opened, for training and testing, is understood using training dataset and is built
Modular system behavior, and the accuracy of system is verified using test data, for sorting algorithm, it can pass through and calculate nicety of grading
Or index is obscured to obtain the accuracy of model；
S10: data result is drawn, and the result of test is depicted as data form record.
Preferably, model is established in the step S1, and to include sound, image and text mode obtain data, information from the external world
With the ability of knowledge；Data, information and the knowledge transformation obtained from the external world is the ability that system grasps knowledge；According to the external world
Or demand caused by own system generation problem, pass through the ability innovated with grasped knowledge；Pass through sound, figure
System generating data, information and knowledge feedback to the external world or are transformed the external world by picture and text mode.
Preferably, the interface termination that prediction service is http in the step S1.
Preferably, in the step S2 AI intelligence software test structure include calculator, logic control device, memory,
Input system and output system.
Preferably, internet works software IQ appraisement system ability is established in the step S4 to include ability in knowledge acquisition, know
Know grasp ability, Innovation Ability of Knowledge and knowledge feedback ability.
Preferably, software IQ test library includes identification text ability, identification sound capabilities, identification figure energy
Power grasps common sense ability, grasps translation ability, grasp computing capability, grasp alignment capabilities, grasp ability of selecting, grasp association
Ability grasps creativity, grasps and find regular ability, with ability of language expression and with image expression ability.
Preferably, absolute IQ formula in the step S8 are as follows:AI deviation IQ formula are as follows:
Preferably, verification mode is calculated in the step S9 and is used is that Mean Square Error, root-mean-square error and the side R miss
Any one in difference.
Compared with prior art, the beneficial effects of the present invention are: the test side of the strict control of the present invention AI intelligence software
Method forms the test mode of AI intelligence software that is unified, being compared to each other, passes through by establishing intelligent system model
Can detection be input to data, information and knowledge the acquisition capability of detection system knowledge in intelligence system, by checking intelligence
The grasp ability of the capacity check systematic knowledge of system knowledge base, by checking that intelligence system by how many data, information and can be known
Know and is converted to new knowledge base content to detect the innovation ability of the systematic knowledge, by checking that can intelligence system will be grasped
Knowledge base content be decomposed into the mode that data, information and knowledge are outwardly transmitted according to demand and can both carry out AI intelligence software
Comprehensive test, test operation is convenient and efficient, by carrying out data verification before testing, can prevent from occurring due to data
Exception, by calculate verify, guarantee the accuracy of test data, be suitble to be widely popularized.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with specific embodiment, to this
Invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, not
For limiting the present invention.
A kind of test method of AI intelligence software, comprising the following steps:
S1: establishing model, by historical data training and establishes intelligent system model, and issues prediction service；
S2: establishing framework, AI intelligence software test structure is established with reference to von Karman framework, by people, machine and people
Work intelligence system is showed with framework；
S3: supplement according to existing knowledge, finds new knowledge element and new rule, feeds them into memory, supplies
Computer and controller use, and by input/output and external progress Knowledge interaction, the outside for carrying out knowledge sharing is known
Know library or cloud storage；
S4: building table, establishes internet works software IQ appraisement system according to the feature main points of intelligent system model, is formed artificial
Intelligence software IQ test scale；
S5: building library, according to artificial intelligence software's IQ test scale is formed, establishes internet works software IQ test library；
S6: verifying carries out data verification before testing, and verifying system receives the number from various sources and various formats
According to whether the taken in data of verifying are converted according to the desired format of goal systems；
S7: test, for AI intelligence software and human subject, when answering exam pool, according to the answer result of feedback
As judge object；
S8: calculating, and carries out software according to the absolute IQ calculation formula and AI deviation IQ formula of AI intelligence software test
Measuring and calculation；
S9: verifying is calculated, basic data seven or three is opened, for training and testing, is understood using training dataset and is built
Modular system behavior, and the accuracy of system is verified using test data, for sorting algorithm, it can pass through and calculate nicety of grading
Or index is obscured to obtain the accuracy of model；
S10: data result is drawn, and the result of test is depicted as data form record.
Specifically, establishing model in the step S1, to include sound, image and text mode obtain data, information from the external world
With the ability of knowledge；Data, information and the knowledge transformation obtained from the external world is the ability that system grasps knowledge；According to the external world
Or demand caused by own system generation problem, pass through the ability innovated with grasped knowledge；Pass through sound, figure
System generating data, information and knowledge feedback to the external world or are transformed the external world by picture and text mode.
Specifically, the interface termination that prediction service is http in the step S1.
Specifically, in the step S2 AI intelligence software test structure include calculator, logic control device, memory,
Input system and output system.
Include ability in knowledge acquisition, know specifically, establishing internet works software IQ appraisement system ability in the step S4
Know grasp ability, Innovation Ability of Knowledge and knowledge feedback ability.
Specifically, software IQ test library includes identification text ability, identification sound capabilities, identification figure energy
Power grasps common sense ability, grasps translation ability, grasp computing capability, grasp alignment capabilities, grasp ability of selecting, grasp association
Ability grasps creativity, grasps and find regular ability, with ability of language expression and with image expression ability.
Specifically, absolute IQ formula in the step S8 are as follows:AI deviation IQ formula are as follows:
Specifically, calculating verification mode in the step S9 and using is that Mean Square Error, root-mean-square error and the side R miss
Any one in difference.
In summary: the test method of the strict control of the present invention AI intelligence software, by establishing intelligent system model, shape
It, can be defeated by data, information and knowledge by detection at the test mode of AI intelligence software that is unified, being compared to each other
The acquisition capability for entering into intelligence system detection system knowledge, by the capacity check systematic knowledge for checking intelligence system knowledge base
Grasp ability, by checking that how many data, information and Knowledge conversion can be new knowledge base content to examine by intelligence system
The innovation ability for surveying the systematic knowledge, by checking that can intelligence system be decomposed into number for the knowledge base content of grasp according to demand
The comprehensive test of AI intelligence software can be both carried out according to, the mode outwardly transmitted of information and knowledge, test operation facilitates fast
Victory can prevent the exception occurred due to data by carrying out data verification before testing, be verified by calculating, and guarantee to survey
The accuracy for trying data, is suitble to be widely popularized.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Anyone skilled in the art in the technical scope disclosed by the present invention, according to the technique and scheme of the present invention and its
Inventive concept is subject to equivalent substitution or change, should be covered by the protection scope of the present invention.
Claims (8)
1. a kind of test method of AI intelligence software, it is characterised in that: the following steps are included:
S1: establishing model, by historical data training and establishes intelligent system model, and issues prediction service；
S2: establishing framework, establishes AI intelligence software test structure with reference to von Karman framework, by people, machine and artificial intelligence
Energy system is showed with framework；
S3: supplement according to existing knowledge, finds new knowledge element and new rule, feeds them into memory, for calculating
Machine and controller use, and by input/output and external progress Knowledge interaction, carry out the external knowledge library of knowledge sharing
Or cloud storage；
S4: building table, establishes internet works software IQ appraisement system according to the feature main points of intelligent system model, forms artificial intelligence
Software IQ test scale；
S5: building library, according to artificial intelligence software's IQ test scale is formed, establishes internet works software IQ test library；
S6: verifying carries out data verification before testing, and verifying system receives the data from various sources and various formats, tests
Demonstrate,prove whether taken in data are converted according to the desired format of goal systems；
S7: test, for AI intelligence software and human subject, when answering exam pool, according to the answer result conduct of feedback
Judge object；
S8: calculating, and carries out software test according to the absolute IQ calculation formula and AI deviation IQ formula of AI intelligence software test
It calculates；
S9: verifying is calculated, basic data seven or three is opened, for training and testing, system is understood and modeled using training dataset
System behavior, and verify using test data the accuracy of system can be by calculating nicety of grading or mixed for sorting algorithm
Index confuse to obtain the accuracy of model；
S10: data result is drawn, and the result of test is depicted as data form record.
2. a kind of test method of AI intelligence software according to claim 1, it is characterised in that: established in the step S1
Model, which includes sound, image and text mode, obtains data, the ability of information and knowledge from the external world；From the external world obtain data,
Information and knowledge transformation are the ability that system grasps knowledge；According to need caused by the external world or own system generation problem
It asks, passes through the ability innovated with grasped knowledge；By sound, image and text mode by system generating data,
Information and knowledge feedback to the external world or are transformed the external world.
3. a kind of test method of AI intelligence software according to claim 1, it is characterised in that: predicted in the step S1
Service is the interface termination of http.
4. a kind of test method of AI intelligence software according to claim 1, it is characterised in that: AI intelligence in the step S2
Energy software test framework includes calculator, logic control device, memory, input system and output system.
5. a kind of test method of AI intelligence software according to claim 1, it is characterised in that: established in the step S4
Internet works software IQ appraisement system ability includes that ability in knowledge acquisition, acquisition of knowledge ability, Innovation Ability of Knowledge and knowledge are anti-
Energy regenerative power.
6. a kind of test method of AI intelligence software according to claim 1, it is characterised in that: the software IQ test
Exam pool includes identification text ability, identification sound capabilities, identification graph ability, grasps common sense ability, grasp translation ability, grasp
Computing capability, grasp alignment capabilities grasp ability of selecting, grasp associative ability, grasp creativity, grasp the regular energy of discovery
Power, with ability of language expression and with image expression ability.
7. a kind of test method of AI intelligence software according to claim 1, it is characterised in that: in the step S8 absolutely
IQ formula are as follows:AI deviation IQ formula are as follows:
8. a kind of test method of AI intelligence software according to claim 1, it is characterised in that: calculated in the step S9
It is any one in Mean Square Error, root-mean-square error and the side's R error that verification mode, which uses,.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910699230.4A CN110377526A (en) | 2019-07-31 | 2019-07-31 | A kind of test method of AI intelligence software |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910699230.4A CN110377526A (en) | 2019-07-31 | 2019-07-31 | A kind of test method of AI intelligence software |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110377526A true CN110377526A (en) | 2019-10-25 |
Family
ID=68257256
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910699230.4A Pending CN110377526A (en) | 2019-07-31 | 2019-07-31 | A kind of test method of AI intelligence software |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110377526A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110908919A (en) * | 2019-12-02 | 2020-03-24 | 上海市软件评测中心有限公司 | Response test system based on artificial intelligence and application thereof |
| CN112749072A (en) * | 2021-01-22 | 2021-05-04 | 北京聚云科技有限公司 | Method and device for testing cloud storage system storing training data |
| CN116450532A (en) * | 2023-06-13 | 2023-07-18 | 西安晟昕科技股份有限公司 | Multi-dimensional testing method for computer software performance |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107678935A (en) * | 2017-05-10 | 2018-02-09 | 平安科技（深圳）有限公司 | Data creation method, terminal and computer-readable recording medium |
| CN108038055A (en) * | 2017-12-07 | 2018-05-15 | 厦门理工学院 | A kind of software defect detection method of valid data processing |
| CN108764677A (en) * | 2018-05-17 | 2018-11-06 | 刘锋 | Artificial intelligence IQ evaluation method and system |
| CN109408827A (en) * | 2018-11-07 | 2019-03-01 | 南京理工大学 | A kind of software entity recognition methods based on machine learning |
| CN109684190A (en) * | 2017-10-19 | 2019-04-26 | 株式会社日立制作所 | Software testing device and method |
| CN109933309A (en) * | 2019-03-06 | 2019-06-25 | 上海工业控制安全创新科技有限公司 | Machine learning algorithm is applied to the flow and method of automobile software-development function safety |
-
2019
- 2019-07-31 CN CN201910699230.4A patent/CN110377526A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107678935A (en) * | 2017-05-10 | 2018-02-09 | 平安科技（深圳）有限公司 | Data creation method, terminal and computer-readable recording medium |
| CN109684190A (en) * | 2017-10-19 | 2019-04-26 | 株式会社日立制作所 | Software testing device and method |
| CN108038055A (en) * | 2017-12-07 | 2018-05-15 | 厦门理工学院 | A kind of software defect detection method of valid data processing |
| CN108764677A (en) * | 2018-05-17 | 2018-11-06 | 刘锋 | Artificial intelligence IQ evaluation method and system |
| CN109408827A (en) * | 2018-11-07 | 2019-03-01 | 南京理工大学 | A kind of software entity recognition methods based on machine learning |
| CN109933309A (en) * | 2019-03-06 | 2019-06-25 | 上海工业控制安全创新科技有限公司 | Machine learning algorithm is applied to the flow and method of automobile software-development function safety |
Non-Patent Citations (1)
| Title |
|---|
| 颜媚等: "AI在软件测试领域中的应用与发展", 《信息通信技术与政策》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110908919A (en) * | 2019-12-02 | 2020-03-24 | 上海市软件评测中心有限公司 | Response test system based on artificial intelligence and application thereof |
| CN112749072A (en) * | 2021-01-22 | 2021-05-04 | 北京聚云科技有限公司 | Method and device for testing cloud storage system storing training data |
| CN112749072B (en) * | 2021-01-22 | 2024-03-01 | 北京聚云科技有限公司 | Testing method and device for cloud storage system for storing training data |
| CN116450532A (en) * | 2023-06-13 | 2023-07-18 | 西安晟昕科技股份有限公司 | Multi-dimensional testing method for computer software performance |
| CN116450532B (en) * | 2023-06-13 | 2023-09-05 | 西安晟昕科技股份有限公司 | Multi-dimensional testing method for computer software performance |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112836064B (en) | Knowledge graph completion method and device, storage medium and electronic equipment | |
| CN110377526A (en) | A kind of test method of AI intelligence software | |
| Zhang et al. | Delay-dependent guaranteed cost control for uncertain stochastic fuzzy systems with multiple time delays | |
| CN107330970A (en) | A kind of threedimensional model processing method, server and computer read/write memory medium | |
| CN113704410B (en) | Mood wave detection method and device, electronic equipment and storage medium | |
| CN109145362A (en) | A kind of power network modeling method and system | |
| EP4123516A1 (en) | Method and apparatus for acquiring pre-trained model, electronic device and storage medium | |
| JP2023092442A (en) | Integrated circuit chip verification method, device, electronic device and storage medium | |
| Sakharov et al. | Adaptive load balancing in the modified mind evolutionary computation algorithm | |
| CN108763707A (en) | The approximate Analysis of structural reliability method of second order saddle point is based under Hybrid parameter matrix | |
| CN110197037A (en) | The dynamic parameters identification method and system of robot based on singular value decomposition | |
| CN116305683A (en) | Power system transient stability evaluation method and system based on sample equalization | |
| CN113450612B (en) | Development method of complete teaching device applied to relay protection training | |
| CN108920810A (en) | Power Simulation Method Based on Big Data Architecture | |
| CN107102543A (en) | The forming method and device of a kind of energy router anti-interference controller | |
| Zhang | Application of 3D Digital Technology in Virtual Laboratory Training | |
| CN110363107A (en) | Face forehead point Quick Extended method, apparatus, storage medium and processor | |
| Chen et al. | Building Method for Metaverse Model of Power Information and Communication Network | |
| Huang et al. | Real-Time Interaction VR Design of Role-Based Autonomous Experiment on ERP | |
| Ning et al. | Research on automatic mapping of electrical system diagram based on artificial intelligence algorithm | |
| WO2024066159A1 (en) | Private network configuration method, apparatus and device, computer readable storage medium, and computer program product | |
| Jeyasurya | An educational interactive program for direct analysis of power system transient stability | |
| Wang et al. | A Model and Simulation of Energy Net Information Risk Transmission Based on Deep Learning | |
| Hao et al. | Virtual Simulation Resource Control Platform for Electric Power Training | |
| Al Jahdali et al. | PREPARING THE PATH FOR THE EFFICIENT SIMULATION OF TURBULENT COMPRESSIBLE INDUSTRIAL FLOWS WITH ROBUST COLLOCATED RK-DG SOLVERS |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191025 |
|
| RJ01 | Rejection of invention patent application after publication |