CN110909133B - Intelligent question and answer testing method and device, electronic equipment and storage medium - Google Patents
Intelligent question and answer testing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN110909133B CN110909133B CN201811083158.4A CN201811083158A CN110909133B CN 110909133 B CN110909133 B CN 110909133B CN 201811083158 A CN201811083158 A CN 201811083158A CN 110909133 B CN110909133 B CN 110909133B
- Authority
- CN
- China
- Prior art keywords
- question
- test
- result
- feedback
- intelligent
- 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.)
- Active
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The invention discloses a test method and a test device for intelligent question answering, electronic equipment and a storage medium, wherein the method comprises the following steps: sending a preset test question to an intelligent question-answering system to detect the feedback correctness of the intelligent question-answering system to obtain a feedback result; matching and analyzing the feedback result and the expected result to obtain a test result of the test question; counting the accuracy of the test data based on the test result of each test question sentence, and adjusting the intelligent question-answering system according to the accuracy; according to the method, the device, the electronic equipment and the storage medium, the preset test question is automatically sent to the intelligent question-answering system, and then the feedback result of the intelligent question-answering system is subjected to matching analysis through the preset expected result, so that whether the feedback of the intelligent question-answering system for the test question is correct or not and the accuracy rate of the feedback of the intelligent question-answering system are determined; and determining whether to adjust the intelligent question-answering system according to the accuracy so as to improve the project quality of the intelligent question-answering system.
Description
Technical Field
The invention relates to the field of computers, in particular to a method and a device for testing intelligent question answering, electronic equipment and a storage medium.
Background
The chat robot can be used for practical purposes such as customer service or information acquisition. Some chat robots carry natural language processing systems, but most simple systems only capture input keywords and then search for the most appropriate answer sentence from the database. Currently, chat robots are part of virtual assistants (e.g., Google smart assistants) that can interface with applications, websites, and instant messaging platforms (Facebook messengers) of many organizations, such as: the customer service robot. Non-assistant applications include chat rooms for entertainment purposes, research and special product promotions such as: a social robot.
In addition, in order to ensure that the interaction between the chat robot and the user can be smoother, the accuracy of the output result fed back to the user by the chat robot needs to be tested. However, in the prior art, a test question is manually input into the chat robot, and whether the output result fed back by the chat robot is correct and whether the test question is correct are manually analyzed, which causes the test process to be very complicated and time-consuming.
Disclosure of Invention
The invention mainly aims to provide an intelligent question and answer testing method, an intelligent question and answer testing device, electronic equipment and a storage medium, and solves the problem that in the prior art, the testing process is very complicated and time-consuming due to manual testing.
According to one aspect of the invention, a method for testing intelligent question answering is provided, which comprises the following steps: presetting test data, wherein the test data comprises a test question sentence and an expected result; sending each test question to an intelligent question-answering system to detect the feedback correctness of the intelligent question-answering system to obtain a feedback result; matching and analyzing the feedback result and the expected result to obtain a test result of the test question; and counting the accuracy of the test data based on the test result of each test question sentence, and adjusting the intelligent question-answering system according to the test result when the accuracy is less than the preset accuracy.
Optionally, the intelligent question-answering system includes a front-end knowledge base, the front-end knowledge base is a local database, and the feedback correctness detection includes: inquiring whether a preset result exists in a front-end knowledge base or not based on the test question; when the preset result is inquired in the front-end knowledge base, determining that the feedback result comprises the preset result and a front-end type identifier for representing the preset result from the front-end knowledge base; the desired results include: pre-storing a front-end identifier corresponding to the test question; the matching analysis comprises the following steps: judging whether the pre-stored front-end identification is the same as the front-end type identification, and if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question answering system responding to the test question sentence.
Optionally, the intelligent question-answering system further comprises a question-answering knowledge base; if the preset result is not inquired in the front-end knowledge base, the feedback correctness detection further comprises the following steps: inquiring whether a question-answer knowledge base has a preset result or not based on the test question; when a preset result is queried in the question-answer knowledge base, determining that the feedback result comprises the preset result, and expecting the result to comprise: and pre-storing results corresponding to the test question.
Optionally, the preset result includes one or more of the following: a standard question-answer sentence, at least two standard question-answer sentences, or an operation instruction, wherein the standard question-answer sentence comprises: standard question sentences and standard results which correspond to each other; when the preset result is a standard question-answer sentence, the matching analysis comprises the following steps: judging whether the standard question-answer sentence is the same as the standard question-answer sentence in the pre-stored result, if so, representing the test result as that the feedback of the intelligent question-answer system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question; when the preset result is at least two standard question sentences, the matching analysis comprises the following steps: judging whether the at least two standard question sentences comprise pre-stored standard question sentences represented by the test question sentences in the pre-stored results, if so, representing the test results that the feedback of the intelligent question answering system responding to the test question sentences is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question; when the preset result is an operation instruction, the matching analysis comprises the following steps: judging whether the preset result is a response result of the operation instruction in the prestored result, if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the intelligent question-answering system further comprises a chat knowledge base, wherein the chat knowledge base is a network database; if the preset result is not inquired in the question-answer knowledge base, the feedback correctness detection further comprises the following steps: inquiring whether a preset result exists in a chat knowledge base or not based on the test question; when a preset result is inquired in the chat knowledge base, determining that the feedback result comprises the preset result and a chat type identifier for representing that the preset result comes from the chat knowledge base; moreover, the desired results include: pre-storing chat type marks corresponding to the test question sentences; the matching analysis comprises the following steps: judging whether the pre-stored chat identifier is the same as the chat type identifier or not, if so, representing the test result as that the intelligent question-answering system responds to the feedback of the test question correctly; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the counting the accuracy of the test data based on the test result of each test question includes: counting the number of test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the sent test question sentences; and determining the accuracy of the test data based on the number of the test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the test question sentences which are sent.
Optionally, adjusting the intelligent question-answering system according to the test result includes: and storing the test question into a standard question feedback database of the intelligent question-answering system, wherein the standard question feedback database is used for storing standard questions.
According to a second aspect of the present invention, there is provided a test apparatus for intelligent question answering, the apparatus comprising: the device comprises a presetting module, a data processing module and a data processing module, wherein the presetting module is used for presetting test data, and the test data comprises a test question sentence and an expected result; the sending module is used for sending each test question to the intelligent question-answering system so as to detect the feedback correctness of the intelligent question-answering system and obtain a feedback result; the analysis module is used for matching and analyzing the feedback result and the expected result to obtain a test result of the test question sentence; and the adjusting module is used for counting the accuracy of the test data based on the test result of each test question sentence, and adjusting the intelligent question-answering system according to the test result when the accuracy is less than the preset accuracy.
Optionally, the intelligent question-answering system includes a front-end knowledge base, the front-end knowledge base is a local database, and the sending module includes: the front-end query unit is used for querying whether a preset result exists in the front-end knowledge base or not based on the test question; the front-end determining unit is used for determining that the feedback result comprises a preset result and a front-end type identifier used for representing the preset result from the front-end knowledge base when the preset result is inquired in the front-end knowledge base; the desired results include: pre-storing a front-end identifier corresponding to the test question; the analysis module is specifically configured to: judging whether the pre-stored front-end identification is the same as the front-end type identification, and if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the intelligent question-answering system further comprises a question-answering knowledge base; if the preset result is not inquired in the front-end knowledge base, the sending module further comprises: the question-answer inquiring unit is used for inquiring whether a preset result exists in the question-answer knowledge base or not based on the test question; the question-answer determining unit is used for determining that the feedback result comprises a preset result when the preset result is inquired in the question-answer knowledge base, and the expected result comprises: and pre-storing results corresponding to the test question.
Optionally, the preset result includes one or more of the following: a standard question-answer sentence, at least two standard question-answer sentences, or an operation instruction, wherein the standard question-answer sentence comprises: standard question sentences and standard results which correspond to each other; when the preset result is a standard question-answer sentence, the analysis module is specifically configured to: judging whether the standard question-answer sentence is the same as the standard question-answer sentence in the pre-stored result, if so, representing the test result as that the feedback of the intelligent question-answer system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question; when the preset result is at least two standard question sentences, the analysis module is specifically configured to: judging whether the at least two standard question sentences comprise pre-stored standard question sentences represented by test question sentences in pre-stored results, if so, representing the test results that the feedback of the intelligent question answering system responding to the test question sentences is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question; when the preset result is the operation instruction, the analysis module is specifically configured to: judging whether the preset result is a response result of the operation instruction in the prestored result, if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the intelligent question-answering system further comprises a chat knowledge base, wherein the chat knowledge base is a network database; if the preset result is not inquired in the question-answer knowledge base, the sending module further comprises: the chat inquiry unit is used for inquiring whether a preset result exists in the chat knowledge base or not based on the test question; the chat determining unit is used for determining that the feedback result comprises the preset result and a chat type identifier for representing the preset result from the chat knowledge base when the preset result is inquired in the chat knowledge base; moreover, the desired results include: pre-storing chat type marks corresponding to the test question sentences; the analysis module is specifically configured to: judging whether the pre-stored chat identifier is the same as the chat type identifier or not, if so, representing the test result as that the intelligent question-answering system responds to the feedback of the test question correctly; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the adjusting module includes: the statistical unit is used for counting the number of the test question sentences which are correctly fed back by the intelligent question answering system and the total number of the sent test question sentences; and the accuracy calculation unit is used for determining the accuracy of the test data based on the number of the test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the sent test question sentences.
Optionally, the adjusting module further includes: and the storage unit is used for storing the test question to a standard question feedback database of the intelligent question-answering system, wherein the standard question feedback database is used for storing the standard question.
According to a third aspect of the present invention, there is provided an electronic device comprising a processor and a memory;
the memory is used for storing computer instructions, and the processor is used for operating the computer instructions stored by the memory so as to realize the intelligent question-answering testing method.
According to a fourth aspect of the present invention, there is provided a computer-readable storage medium storing one or more programs, the one or more programs being executable by one or more processors to implement a method for intelligent question and answer testing as described above.
The invention has the following beneficial effects: the method comprises the steps that a test question and a corresponding expected result are preset, then the preset test question is automatically sent to an intelligent question-answering system, then matching analysis is conducted on a feedback result of the intelligent question-answering system through the preset expected result, whether the feedback of the intelligent question-answering system for the test question is correct or not is determined, and the accuracy of the feedback of the intelligent question-answering system for the test question can be obtained; and determining whether to adjust the intelligent question-answering system according to the accuracy so as to improve the question-answering accuracy of the intelligent question-answering system.
Drawings
FIG. 1 is a flow chart of a method for testing intelligent question answering according to a first embodiment of the present invention;
FIG. 2 is a flowchart illustrating a method for testing intelligent question answering according to a second embodiment of the present invention;
FIG. 3 is a flow chart of a method for testing intelligent question answering according to a third embodiment of the present invention;
FIG. 4 is a flow chart of a method for testing intelligent question answering according to a fourth embodiment of the present invention;
FIG. 5 is a schematic flowchart illustrating a method for testing intelligent question answering according to a fifth embodiment of the present invention;
fig. 6 is a schematic structural diagram of a testing apparatus for intelligent question answering according to a fifth embodiment of the present invention.
The implementation, functional features and advantages of the present invention will be further described with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the invention.
Before discussing exemplary embodiments in greater detail, it should be noted that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although a flowchart may describe the operations as a sequential process, many of the operations can be performed in parallel, concurrently, or simultaneously. In addition, the order of the operations may be re-arranged. A process may be terminated when its operations are completed, but may have additional steps not included in the figure. A process may correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.
The term "electronic device", also referred to as "computer", refers to an intelligent electronic device that can execute predetermined processes such as numerical calculation and/or logic calculation by running predetermined programs or instructions, and may include a processor and a memory, wherein the processor executes a pre-stored instruction stored in the memory to execute the predetermined processes, or the processor executes the predetermined processes by hardware such as ASIC, FPGA, DSP, or a combination thereof. Electronic devices include, but are not limited to, servers, personal computers, laptops, tablets, smart phones, and the like.
The electronic equipment comprises user equipment and network equipment. Wherein the user equipment includes but is not limited to computers, smart phones, PDAs, etc.; the network device includes, but is not limited to, a single network server, a server group consisting of a plurality of network servers, or a cloud based computing (CloudComputing) consisting of a large number of computers or network servers, wherein cloud computing is one of distributed computing, a super virtual computer consisting of a collection of loosely coupled computers. The electronic device can be operated independently to realize the invention, and can also be accessed into a network to realize the invention through the interactive operation with other computer devices in the network. The network where the electronic device is located includes, but is not limited to, the internet, a wide area network, a metropolitan area network, a local area network, a VPN network, and the like.
It should be noted that the user equipment, the network device, the network, etc. are only examples, and other existing or future electronic devices or networks may be applicable to the present invention, and are included in the scope of the present invention and are included by reference.
The methods discussed below, some of which are illustrated by flow diagrams, may be implemented by hardware, software, firmware, middleware, microcode, hardware description languages, or any combination thereof. When implemented in software, firmware, middleware or microcode, the program code or code segments to perform the necessary tasks may be stored in a machine or computer readable medium such as a storage medium. The processor(s) may perform the necessary tasks.
Specific structural and functional details disclosed herein are merely representative and are provided for purposes of describing example embodiments of the present invention. The present invention may, however, be embodied in many alternate forms and should not be construed as limited to only the embodiments set forth herein.
It will be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element may be termed a second element, and, similarly, a second element may be termed a first element, without departing from the scope of example embodiments. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may be present. In contrast, when an element is referred to as being "directly connected" or "directly coupled" to another element, there are no intervening elements present. Other words used to describe the relationship between elements (e.g., "between" versus "directly between.," adjacent to.. versus "directly adjacent to.. etc.) should be interpreted in a similar manner.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used herein, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It should also be noted that, in some alternative implementations, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed substantially concurrently, or the figures may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
In order to facilitate understanding of the embodiments of the present invention, the following detailed description of the embodiments of the present invention is provided.
Fig. 1 is a flowchart of a method for testing an intelligent question answering according to a first embodiment of the present invention. As shown in fig. 1, a method for testing an intelligent question answering according to a first embodiment of the present invention includes:
and S11, presetting test data, wherein the test data comprises a test question and an expected result.
And S12, sending each test question to the intelligent question-answering system to detect the feedback correctness of the intelligent question-answering system, and obtaining a feedback result.
And S13, matching and analyzing the feedback result and the expected result to obtain the test result of the test question.
And S14, counting the accuracy of the test data based on the test result of each test question sentence, and adjusting the intelligent question-answering system according to the test result when the accuracy is less than the preset accuracy.
In this regard, the test question is preset and the corresponding expected result is sent to the intelligent question-answering system automatically, then the feedback result of the intelligent question-answering system is subjected to matching analysis through the preset expected result, whether the feedback of the intelligent question-answering system for the test question is correct or not is determined, and the accuracy of the feedback of the intelligent question-answering system for the test question can be obtained; and determining whether to adjust the intelligent question-answering system according to the accuracy so as to improve the question-answering accuracy of the intelligent question-answering system.
Specifically, a method for testing an intelligent question answering according to a first embodiment of the present invention includes:
and S11, presetting test data, wherein the test data comprises a test question and an expected result.
Specifically, in this embodiment, the test question is any query information provided by the simulation user. The expected result is a corresponding correct result which is output by the intelligent question-answering system after receiving and responding the test question.
Wherein, the test question sentence includes but is not limited to: any question information, any instruction information. Furthermore, in the present embodiment, the expression form of the test question is not limited, such as: emoticons, question sentences, text symbols and operation instructions. Accordingly, the desired results include, but are not limited to: emoticons, answers, text symbols and response results of the operation instructions.
Moreover, the above-mentioned intelligent question-answering system belongs to a kind of virtual assistant, and is a computer chat system that carries on conversation via conversation, text, and operation triggering.
And S12, sending each test question to the intelligent question-answering system to detect the feedback correctness of the intelligent question-answering system, and obtaining a feedback result.
In this embodiment, the preset test question is sent to the intelligent question-answering system, so that the intelligent question-answering system queries a corresponding preset result in a database of the intelligent question-answering system, and thus a feedback result of the intelligent question-answering system is obtained.
And S13, matching and analyzing the feedback result and the expected result to obtain the test result of the test question.
And after the feedback result is obtained, performing matching analysis on the feedback result and a preset expected result to determine a corresponding test result, wherein the test result is used for representing whether the feedback result of the intelligent question answering system responding to the test question is correct or not.
And S14, counting the accuracy of the test data based on the test result of each test question sentence, and adjusting the intelligent question-answering system according to the test result when the accuracy is less than the preset accuracy.
After the test result corresponding to each test question sentence is determined, the corresponding accuracy can be determined.
For the calculation of the accuracy, an implementation manner provided by this embodiment includes:
counting the number of test question sentences which are correctly fed back by the intelligent question answering system and the total number of the sent test question sentences; and determining the accuracy of the test data based on the number of test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the test question sentences which are sent.
Specifically, the calculation method for obtaining the accuracy includes, but is not limited to:
the number of test questions that have been correctly fed back by the intelligent question-and-answer system/the total number of test questions sent is equal to the correct rate.
Moreover, in this embodiment, under the condition that the number of test questions that have been correctly fed back by the intelligent question-answering system and the total number of test questions that have been sent are known, the calculation method and possibly required parameters for obtaining the accuracy are not limited, and only the requirements of this embodiment need to be met.
After obtaining the accuracy, in this embodiment, the intelligent question-answering system needs to be adjusted according to the test result, so as to improve the accuracy of the intelligent question-answering system in responding to the test question.
Specifically, the test question is stored in a standard question feedback database of the intelligent question-answering system, wherein the standard question feedback database is used for storing standard questions.
In this regard, the test question is preset and the corresponding expected result is sent to the intelligent question-answering system automatically, then the feedback result of the intelligent question-answering system is subjected to matching analysis through the preset expected result, whether the feedback of the intelligent question-answering system for the test question is correct or not is determined, and the accuracy of the feedback of the intelligent question-answering system for the test question can be obtained; and determining whether to adjust the intelligent question-answering system according to the accuracy so as to improve the question-answering accuracy of the intelligent question-answering system.
Fig. 2 is a flowchart of a method for testing intelligent question answering according to a second embodiment of the present invention. According to fig. 2, on the basis of the first embodiment, for the feedback correctness detection in step S12 and the matching analysis in step S13, an implementation manner of the present embodiment includes the following steps:
and S22, sending each test question to the intelligent question-answering system, and inquiring whether a preset result exists in the front-end knowledge base or not based on the test question.
The intelligent question-answering system comprises a front-end knowledge base. The front-end knowledge base stores one or more of the following contents: and if the test question sentence is the emoticon or the character symbol, directly inquiring whether a preset result exists in the front-end knowledge base or not based on the test question sentence.
Optionally, the front-end knowledge base is cached in advance in a terminal loaded with the intelligent question-answering system.
And S23, when the preset result is inquired, obtaining a corresponding feedback result, wherein the feedback result comprises the preset result and a front-end type identifier for representing the preset result from the front-end knowledge base.
S24, judging whether the pre-stored front end identifier is the same as the front end type identifier or not; when the same, step S25 is performed, otherwise, step S26 is performed.
Wherein, the pre-stored front-end identification corresponds to the test question one by one. In addition, the front-end type identifier is used for representing that the feedback of the intelligent question answering system responding to the test question belongs to the category of front-end reply.
And S25, determining that the feedback of the intelligent question answering system responding to the test question is correct.
Specifically, a corresponding test result can be obtained, and the test result is characterized in that the feedback of the intelligent question-answering system responding to the test question is correct.
And S26, determining the feedback error of the intelligent question answering system responding to the test question.
Specifically, a corresponding test result can be obtained, and the test result is characterized in that the intelligent question-answering system responds to a feedback error of a test question;
such as: and if the test question is that the smiling emoticon is represented, locally inquiring whether a preset result corresponding to the smiling emoticon exists in the front-end knowledge base, and if the same smiling emoticon is inquired, feeding back the inquired smiling emoticon and the corresponding front-end type identifier. And if the front-end type identifier fed back is the same as the pre-stored front-end identifier corresponding to the test question, determining that the feedback of the intelligent question-answering system responding to the test question is correct, otherwise, determining that the feedback of the intelligent question-answering system responding to the test question is wrong.
In another implementation manner of this embodiment, when it is determined that the feedback result includes the front-end type identifier, it may be directly determined that the intelligent question-and-answer system responds to the feedback error of the test question.
In this regard, it may be further determined that the intelligent question-answering system at least includes a front-end knowledge base cached in a terminal loading the intelligent question-answering system, wherein the front-end knowledge base stores one or more of the following contents: emoticons, text symbols, and the like. In addition, according to the front-end type identifier, it can be determined that the intelligent question-answering system does not pass through an intelligent engine when responding to the test question sentence, but directly replies by the front-end function; and whether the feedback result corresponding to the test question sentence is correct can be determined directly through the front-end type identification, so that the response rate of the intelligent question-answering system can be increased, and meanwhile, the test efficiency can be improved.
Fig. 3 is a flow chart of a method for testing an intelligent question answering according to a third embodiment of the present invention. As shown in fig. 3, based on the second embodiment, for the feedback correctness detection in step S12 and the matching analysis in step S13, the present embodiment includes the following steps:
the intelligent question-answering system further comprises a question-answering knowledge base.
Wherein, the question-answer knowledge base stores one or more of the following contents: a standard question-answer sentence, at least two standard question-answer sentences, or an operation instruction, wherein the standard question-answer sentence comprises: standard question sentence and standard result corresponding to each other.
Optionally, the question and answer knowledge base is a network database, and of course, the question and answer knowledge base may also be cached in advance in a terminal loading the intelligent question and answer system.
The priority of the front-end knowledge base is greater than that of the question-answering knowledge base.
In addition, in this embodiment, specific expression forms of the standard question and the test question are not limited, and they may be in other forms such as characters and pictures, and only need to be expressed with the same meaning, they can be regarded as the same question. Also, the predetermined result and the standard result are also the same.
Such as: what is the price of the cup lid? "and" how much money is sold on the cup lid? "may be considered the same.
Under the condition that the preset result is not inquired in the front-end knowledge base based on the test question sentence, one implementation manner of the embodiment comprises the following steps:
s32, inquiring whether a question-answer knowledge base has a preset result or not based on the test question;
s33: when a preset result is inquired, determining that the feedback result comprises the preset result;
s34: determining the type of the preset result; when it is determined that the preset result is the standard question-answer sentence, performing step S35;
s35, judging whether the standard question-answer sentence is the same as the standard question-answer sentence in the pre-stored result, if so, representing the test result as that the feedback of the intelligent question-answer system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question;
because the standard question-answer sentence includes: standard question sentence and standard result corresponding to each other. Therefore, specific meanings of this step S35 include:
judging whether the standard question and the standard result in the preset result are respectively the same as the standard question and the standard result in the prestored result, if so, representing the test result as that the intelligent question-answering system responds to the correct feedback of the test question; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question;
such as: if the question is "how much money is in the cup lid? "the standard question corresponding to the test question in the prestored result is" what is the price of the cup lid? And, the standard result corresponding to the test question in the pre-stored result is "25 yuan". In this case, if the standard question and the standard result of the standard question-answer sentence in the preset result are also "what is the price of the cup lid", respectively? And 25 Yuan, the test result can be determined to be characterized as that the feedback of the intelligent question-answering system responding to the test question is correct. However, if the standard question of the standard question-and-answer sentence in the preset result is "what is the price of cup? "or" 35 yuan "in the preset result, the test result may be determined to be characterized as a positive error in the response of the intelligent question answering system to the test question.
Of course, in this embodiment, the feedback result also includes a corresponding identifier, and the identifier is used to characterize that the feedback of the intelligent question-answering system responding to the test question belongs to the category of "standard answer".
Of course, in another embodiment, when the preset result is at least two standard question sentences, step S36 is executed;
s36, judging whether the at least two standard question sentences include the pre-stored standard question sentences represented by the test question sentences in the pre-stored results, if so, the test results are represented as that the feedback of the intelligent question-answering system responding to the test question sentences is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question;
such as: if the pre-stored standard question represented by the test question in the pre-stored result is' what is the price of the cup cover? "and the test question is" how much money is in the cup lid? "what are the prices of the cup covers" are the at least two standard questions included in the preset result? "and" how much money is sold on the cup lid? ". After the test question is input into the intelligent question-answering system, the preset result, namely the price of the cup cover, can be obtained under the condition that the accurate meaning of the test question cannot be identified through the intelligent question-answering system? "and" how much money is sold on the cup lid? At this time, the preset result includes a pre-stored standard question represented by the test question in the pre-stored result, so that it can be determined that the obtained test result is represented as that the feedback of the intelligent question-answering system responding to the test question is correct. Of course, if the preset result does not include the pre-stored standard question represented by the test question in the pre-stored result, it may be determined that the test result is characterized as a feedback error of the intelligent question-answering system responding to the test question.
Of course, in this embodiment, the feedback result also includes a corresponding identifier, and the identifier is used to characterize that the feedback of the intelligent question-answering system responding to the test question belongs to the category of "question suggestion".
Of course, in another embodiment, when the preset result is the operation instruction, step S37 is executed;
s37, judging whether the preset result is the response result of the operation instruction in the prestored result, if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Specifically, when the intelligent question-answering system receives the operation instruction, the front-end knowledge base is sequentially inquired, if the corresponding target instruction is not found in the front-end knowledge base, the corresponding target instruction is found in the question-answering knowledge base, and if the corresponding target instruction is found, the intelligent question-answering system sends the corresponding operation instruction to the corresponding instruction execution main body so as to instruct the execution main body to execute the preset operation.
Such as: the intelligent question-answering system receives a balance inquiry command, firstly inquires a front-end knowledge base, and if the corresponding balance inquiry command is not found, the corresponding balance inquiry command is found in the question-answering knowledge base; if the corresponding balance inquiry instruction is found in the question-answering knowledge base, the intelligent question-answering system sends a corresponding balance inquiry operation instruction to the balance system to inquire the balance, and feeds back and displays the balance, and at the moment, the test result can be determined to be represented as the correct feedback of the intelligent question-answering system responding to the test question; otherwise, failure prompt information is fed back, and at the moment, the test result can be determined to be characterized as a feedback error of the intelligent question answering system responding to the test question.
Of course, in this embodiment, the feedback result also includes a corresponding identifier, and the identifier is used to characterize that the feedback of the intelligent question answering system in response to the test question belongs to the category of "instruction".
In this regard, it may be further determined that the intelligent question-answering system at least includes a front-end knowledge base and a question-answering knowledge base, where the question-answering knowledge base stores one or more of the following contents: a standard question-answer sentence, at least two standard question sentences, or an operation instruction. Moreover, whether various feedback results in the intelligent question-answering system are correct or not can be further tested, the response speed of the intelligent question-answering system can be increased, and meanwhile, the testing efficiency can be improved.
Fig. 4 is a flowchart of a method for testing an intelligent question answering according to a fourth embodiment of the present invention. As shown in fig. 4, based on the third embodiment, for the feedback correctness detection in step S12 and the matching analysis in step S13, the present embodiment includes the following steps:
the intelligent question answering system also comprises a chat knowledge base.
Wherein, the chat knowledge base stores the standby answer, and the standby answer comprises one or more of the following contents: and information such as emoticons and character symbols, and if the test question sentence is an emoticon or character symbol. And under the condition that the intelligent question-answering system does not inquire the preset result in the question-answering knowledge base, directly inquiring whether the chat knowledge base has the preset result or not based on the test question. Moreover, the chat knowledge base is a network database.
The priority of the front-end knowledge base is higher than that of the question and answer knowledge base, and the priority of the question and answer knowledge base is higher than that of the chat knowledge base.
Specifically, an implementation manner of this embodiment includes:
s42, inquiring whether a preset result exists in the chat knowledge base or not based on the test question;
s43, when the preset result is inquired in the chat knowledge base, a feedback result is obtained, wherein the feedback result comprises the preset result and a chat type identifier for representing the preset result from the chat knowledge base;
s44: judging whether the pre-stored chat identifier is the same as the chat type identifier, if so, executing step S45; otherwise, go to step S46;
s45: determining that the feedback of the intelligent question-answering system responding to the test question is correct;
specifically, the test result is characterized in that the intelligent question-answering system responds to the correct feedback of the test question;
s46: and determining the feedback error of the intelligent question answering system responding to the test question.
Specifically, the test result is characterized as a feedback error of the intelligent question-answering system responding to the test question.
Of course, in this embodiment, the chat type identifier is used to characterize that the feedback of the intelligent question-answering system in response to the test question belongs to the "chat" category.
Such as: and if the test question represents the smiling emoticon, under the condition that the corresponding type identifier is found in both the front-end knowledge base and the question-answer knowledge base, inquiring whether a preset result corresponding to the smiling emoticon exists in the network database-chat knowledge base, and under the condition, if the same smiling emoticon is inquired, feeding back the inquired smiling emoticon and the corresponding chat type identifier. And if the chat type identifier fed back is the same as the prestored chat identifier corresponding to the test question, determining that the intelligent question-answering system responds to the test question and feeds back correctly, otherwise, determining that the intelligent question-answering system responds to the test question and feeds back wrongly.
In another implementation manner of this embodiment, if it is determined that the feedback result includes the chat type identifier, it may be directly determined that the intelligent question-answering system responds to the feedback error of the test question.
In addition, in another embodiment, if the preset result is not found in the chat knowledge base, in this case, the feedback result includes: the reply-free mark is used for representing that the corresponding preset result cannot be inquired in the intelligent question-answering system based on the test question, and the feedback error of the intelligent question-answering system responding to the test question can be directly determined through the reply-free mark.
In this regard, it may be further determined that the intelligent question-answering system at least includes a front-end knowledge base, a question-answering knowledge base, and a chat knowledge base as a network database, wherein the alternative answers stored in the chat knowledge base include one or more of the following: emoticons, text symbols, and the like. In addition, according to the chat type identifier, it can be determined that the intelligent question-answering system does not inquire the corresponding preset results in the front-end knowledge base and the question-answering knowledge base when responding to the test question, and it can also be directly determined whether the feedback result corresponding to the test question is correct through the chat type identifier, so that the response rate of the intelligent question-answering system can be increased, and meanwhile, the test efficiency can be increased accordingly.
In order to better explain the implementation process of the method of the embodiment, the method of the embodiment is explained below with reference to a specific application example.
Fig. 5 is a schematic flow chart of a method for testing an intelligent question answering according to a fifth embodiment of the present invention. According to the illustration in fig. 5, a fifth embodiment of the present invention provides a method for testing intelligent question answering, which can be applied in the field of artificial intelligence software products.
In addition, the intelligent robot designed in the present embodiment includes: the system comprises a front-end knowledge base, a question-answer knowledge base and a chat knowledge base, wherein the priority of the front-end knowledge base is higher than that of the question-answer knowledge base, and the priority of the question-answer knowledge base is higher than that of the chat knowledge base.
In this embodiment, the technical solution of this embodiment is described by taking the test tool as the test terminal as an example:
specifically, test question data needs to be preset, such as: the test question data is imported into the corresponding database by preparing a test data import tool in advance. Wherein the test question data includes but is not limited to: the method comprises the following steps of testing a question sentence, expecting standard questions corresponding to the testing question sentence, expecting answers and corresponding platform information, wherein the platform for loading and operating the intelligent question-answering system can be determined by means of the imported platform information, such as: the platform may be: such as: qq, weixin and web pages, and ensures the correspondence of the test question sentence with the platform.
The question data is equivalent to the test data in the first to fourth embodiments, and the expected answer is equivalent to the expected result in the first to fourth embodiments. In addition, in this embodiment, the knowledge content online analysis module performs corresponding analysis, and certainly, an analysis module for configuring a test environment (interface, time, etc.) and a semantic matching result is also required.
And then, starting an accuracy detection mechanism according to the set time, and sending the test question sentence to the intelligent robot after the accuracy detection mechanism is started, wherein the intelligent robot is equivalent to the intelligent question-answering system in the first to fourth embodiments. After receiving the test question, the intelligent robot needs to inquire whether a preset result exists in the front-end knowledge base according to the test question. Wherein the front-end database is a local database. And when the preset result is inquired, storing data representing the preset result, and feeding back the preset result and a front-end type identifier used for representing the preset result from a front-end knowledge base by the intelligent robot. The front-end type identifier is used for representing that the feedback of the intelligent question answering system responding to the test question belongs to the category of front-end reply.
And determining the feedback error of the intelligent robot under the condition that the received feedback includes the front-end type identifier.
Of course, if the preset result is not queried in the front-end knowledge base, whether the preset result exists in the question-answer knowledge base or not is queried through the intelligent engine based on the test question sentence, and the preset result is fed back under the condition that the preset result is queried. Then judging whether the preset result is an expected answer which is the same as the pre-imported expected answer or not, if so, representing the test result as that the intelligent robot responds to the correct feedback of the test question; otherwise, the test result is characterized as the feedback error of the intelligent robot responding to the test question.
Of course, in this embodiment, the feedback result also includes a corresponding identifier, and the identifier is used to characterize that the feedback of the intelligent robot responding to the test question belongs to the category of "standard answer".
In addition, after receiving a preset result, if the preset result is detected to be at least two standard question sentences, at the moment, if the at least two standard question sentences are determined to comprise the pre-stored standard question sentences represented by the test question sentences, the test result is characterized in that the feedback of the intelligent robot responding to the test question sentences is correct; otherwise, the test result is characterized as a feedback error of the intelligent robot responding to the test question.
Of course, in this embodiment, the feedback result also includes a corresponding identifier, and the identifier is used to characterize that the feedback of the intelligent question-answering system responding to the test question belongs to the category of "question suggestion".
In addition, after the preset result is received, if the preset result is detected to be the response result of the operation instruction, and the response result is determined to be correct, the test result is characterized to be that the intelligent robot responds to the correct feedback of the test question; otherwise, the test result is characterized as a feedback error of the intelligent robot responding to the test question.
Of course, in this embodiment, the feedback result also includes a corresponding identifier, and the identifier is used to characterize that the feedback of the intelligent question answering system in response to the test question belongs to the category of "instruction".
If the preset result is not inquired, inquiring whether the preset result exists in the chat knowledge base or not based on the test question; and when a preset result is inquired in the chat knowledge base, storing data representing the preset result, and feeding back the preset result and a chat type identifier for representing that the preset result comes from the chat knowledge base by the intelligent robot. The front-end type identifier is used for representing that the feedback of the intelligent robot responding to the test question belongs to the chat category.
In addition, in the case that the received feedback is detected to include the chat type identifier, the feedback error of the intelligent robot can be determined.
If the preset result is not found in the chat knowledge base, in this case, the feedback result includes: the reply-free mark is used for representing that the corresponding preset result cannot be inquired in the intelligent question-answering system based on the test question, and the feedback error of the intelligent question-answering system responding to the test question can be directly determined through the reply-free mark.
After determining whether the feedback corresponding to each test question is determined, counting the number of the test questions which are correctly fed back by the intelligent robot and the total number of the sent test questions; and determining the accuracy of the test data based on the number of the test question sentences which are correctly fed back by the intelligent robot and the total number of the sent test question sentences.
After the accuracy is determined, the database of the intelligent robot can be optimized and adjusted according to the accuracy, such as: and storing the test question into a robot content database of the intelligent robot, wherein the robot content database is equivalent to a standard question feedback database in the first to third embodiments, and the standard question feedback database is used for storing standard questions.
In addition, in this embodiment, the following optimization adjustments may be performed, so that the optimization is completed and the front end is effective:
1. summarizing and summarizing the test question sentences, and adding templates in the corresponding expected standard questions of the background of the robot;
2. by creating new knowledge points such as: adding a test question, an expected standard question corresponding to the test question, an expected answer, a corresponding platform (such as qq, weixin and webpage), dimensionality and classification;
3. ignoring the modification, marked as ignore, indicates that no addition or corresponding modification is needed.
In addition, after the optimization adjustment, if the effect is still not ideal, the engineer may check the change of the accuracy by modifying the accuracy threshold until the accuracy of the test sample reaches an acceptable range.
In this regard, a test question and a corresponding expected result are preset, the preset test question is automatically sent to the intelligent robot, and then the feedback result of the intelligent robot is subjected to matching analysis according to the preset expected result, so that whether the feedback of the intelligent robot for the test question is correct or not is determined, and the accuracy of the feedback of the intelligent robot for the test question can be obtained according to the accuracy; and determining whether to adjust the intelligent robot according to the accuracy so as to improve the question-answering accuracy of the intelligent robot.
Fig. 6 is a block diagram illustrating a flow chart of an intelligent question answering testing apparatus according to a fifth embodiment of the present invention. According to a fifth embodiment of the present invention, as shown in fig. 5, there is provided an intelligent question answering test apparatus, including:
a presetting module 110, configured to preset test data, where the test data includes a test question and an expected result; the sending module 120 is configured to send each test question to the intelligent question-answering system to perform feedback correctness detection of the intelligent question-answering system, so as to obtain a feedback result; the analysis module 130 is configured to perform matching analysis on the feedback result and the expected result to obtain a test result of the test question; the adjusting module 140 is configured to count a correctness of the test data based on the test result of each test question, and adjust the intelligent question-answering system according to the test result when the correctness is smaller than a preset correctness.
Optionally, the intelligent question-answering system includes a front-end knowledge base, the front-end knowledge base is a local database, and the sending module 120 includes: the front-end query unit is used for querying whether a preset result exists in the front-end knowledge base or not based on the test question; the front-end determining unit is used for determining that the feedback result comprises a preset result and a front-end type identifier used for representing the preset result from the front-end knowledge base when the preset result is inquired in the front-end knowledge base; the desired results include: pre-storing a front-end identifier corresponding to the test question; the analysis module 130 is specifically configured to: judging whether the pre-stored front-end identification is the same as the front-end type identification, and if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the intelligent question-answering system further comprises a question-answering knowledge base; if the preset result is not queried in the front-end knowledge base, the sending module 120 further includes: the question-answer inquiring unit is used for inquiring whether a preset result exists in the question-answer knowledge base or not based on the test question; the question-answer determining unit is used for determining that the feedback result comprises a preset result when the preset result is inquired in the question-answer knowledge base, and the expected result comprises: and pre-storing results corresponding to the test question.
Optionally, the preset result includes one or more of the following: a standard question-answer sentence, at least two standard question-answer sentences, or an operation instruction, wherein the standard question-answer sentence comprises: a standard question sentence and a standard result which correspond to each other; when the preset result is a standard question-answer sentence, the analysis module 130 is specifically configured to: judging whether the standard question-answer sentence is the same as the standard question-answer sentence in the prestored result or not, if so, representing the test result as that the feedback of the intelligent question-answer system responding to the test question sentence is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question; when the preset result is at least two standard question sentences, the analysis module 130 is specifically configured to: judging whether the at least two standard question sentences comprise pre-stored standard question sentences represented by test question sentences in pre-stored results, if so, representing the test results that the feedback of the intelligent question answering system responding to the test question sentences is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question; when the preset result is an operation instruction, the analysis module 130 is specifically configured to: judging whether the preset result is a response result of the operation instruction in the prestored result, if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the intelligent question-answering system further comprises a chat knowledge base, wherein the chat knowledge base is a network database; if the preset result is not queried in the question-answer knowledge base, the sending module 120 further includes: the chat inquiry unit is used for inquiring whether a preset result exists in the chat knowledge base or not based on the test question sentence; the chat determining unit is used for determining that the feedback result comprises the preset result and a chat type identifier for representing the preset result from the chat knowledge base when the preset result is inquired in the chat knowledge base; moreover, the desired results include: pre-storing chat type marks corresponding to the test question sentences; the analysis module 130 is specifically configured to: judging whether the pre-stored chat identifier is the same as the chat type identifier or not, if so, representing the test result as that the intelligent question-answering system responds to the feedback of the test question correctly; otherwise, the test result is characterized as the feedback error of the intelligent question-answering system responding to the test question.
Optionally, the adjusting module 140 includes: the statistical unit is used for counting the number of the test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the sent test question sentences; and the accuracy calculation unit is used for determining the accuracy of the test data based on the number of the test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the sent test question sentences.
Optionally, the adjusting module 140 further includes: and the storage unit is used for storing the test question to a standard question feedback database of the intelligent question answering system, wherein the standard question feedback database is used for storing the standard question.
A seventh embodiment of the present invention provides an electronic device, including a processor and a memory; the memory is used for storing computer instructions, and the processor is used for operating the computer instructions stored by the memory so as to realize the intelligent question-answering testing method.
The terms and implementation principles related to the electronic device in the seventh embodiment of the present invention may specifically refer to the method for testing the intelligent question answering in the first to fifth embodiments of the present invention, and are not described herein again.
An eighth embodiment of the present invention provides a computer-readable storage medium, which stores one or more modules, where the one or more modules are executable by one or more processors to implement the method for testing intelligent question and answer.
The terms and implementation principles related to a computer-readable storage medium in the eighth embodiment of the present invention may specifically refer to the method for testing intelligent question answering in the first to fifth embodiments of the present invention, which is not described herein again.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.
While the present invention has been described with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, which are illustrative and not restrictive, and it will be apparent to those skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (9)
1. A method for testing intelligent question answering is characterized by comprising the following steps:
presetting test data, wherein the test data comprises a test question sentence and an expected result;
sending each test question to an intelligent question-answering system to detect the feedback correctness of the intelligent question-answering system to obtain a feedback result;
matching and analyzing the feedback result and the expected result to obtain a test result of the test question sentence;
counting the accuracy of the test data based on the test result of each test question sentence, and adjusting the intelligent question-answering system according to the test result when the accuracy is smaller than the preset accuracy;
the intelligent question-answering system comprises a front-end knowledge base, the front-end knowledge base is a local database, and the feedback correctness detection comprises the following steps:
inquiring whether a preset result exists in the front-end knowledge base or not based on the test question sentence;
when the preset result is inquired in the front-end knowledge base, determining that the feedback result comprises the preset result and a front-end type identifier for representing that the preset result comes from the front-end knowledge base;
the desired results include: pre-storing a front end identifier corresponding to the test question sentence; the matching analysis comprises: judging whether the pre-stored front-end identification is the same as the front-end type identification, and if so, representing the test result as that the feedback of the intelligent question answering system responding to the test question is correct; otherwise, the test result is characterized in that the intelligent question-answering system responds to the feedback error of the test question.
2. The method of claim 1, wherein the intelligent question-answering system further comprises a question-answering knowledge base;
if the preset result is not inquired in the front-end knowledge base, the feedback correctness detection further comprises:
inquiring whether a preset result exists in the question-answer knowledge base or not based on the test question;
when the preset result is queried in the question-answer knowledge base, determining that the feedback result comprises the preset result, and the expected result comprises: and pre-storing results corresponding to the test question sentence.
3. The method of claim 2,
the preset result comprises one or more of the following: the system comprises a standard question-answer sentence, at least two standard question-answer sentences and an operation instruction, wherein the standard question-answer sentence comprises: standard question sentences and standard results which correspond to each other;
when the preset result is the standard question-answer sentence, the matching analysis comprises: judging whether the standard question-answer sentence is the same as the standard question-answer sentence in the pre-stored result, if so, representing the test result as that the feedback of the intelligent question-answer system responding to the test question sentence is correct; otherwise, the test result is characterized in that the intelligent question-answering system responds to the feedback error of the test question;
when the preset result is at least two standard question sentences, the matching analysis comprises the following steps: judging whether the at least two standard question sentences comprise prestored standard question sentences represented by test question sentences in the prestored results, if so, representing the test results that the feedback of the intelligent question answering system responding to the test question sentences is correct; otherwise, the test result is characterized in that the intelligent question-answering system responds to the feedback error of the test question;
when the preset result is an operation instruction, the matching analysis comprises: judging whether the preset result is a response result of the operation instruction in the prestored result, if so, representing the test result as that the feedback of the intelligent question-answering system responding to the test question is correct; otherwise, the test result is characterized in that the intelligent question-answering system responds to the feedback error of the test question.
4. The method according to claim 2 or 3, wherein the intelligent question answering system further comprises a chat knowledge base, and the chat database is a network database;
if a preset result is not queried in the question-answer knowledge base, the feedback correctness detection further comprises:
inquiring whether a preset result exists in the chat knowledge base or not based on the test question;
when the preset result is inquired in the chat knowledge base, determining that the feedback result comprises the preset result and a chat type identifier for representing that the preset result comes from the chat knowledge base;
moreover, the desired results include: pre-storing a chat type identifier corresponding to the test question sentence;
the matching analysis comprises: judging whether the pre-stored chat type identification is the same as the chat type identification, if so, representing the test result as that the feedback of the intelligent question answering system responding to the test question sentence is correct; otherwise, the test result is characterized in that the intelligent question-answering system responds to the feedback error of the test question.
5. The method according to claim 1, wherein the counting the correctness of the test data based on the test result of each test question sentence comprises:
counting the number of the test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the sent test question sentences;
and determining the accuracy of the test data based on the number of the test question sentences which are correctly fed back by the intelligent question-answering system and the total number of the sent test question sentences.
6. The method of claim 1, wherein said adjusting said intelligent question-answering system according to the test results comprises:
and storing the test question to a standard question feedback database of the intelligent question-answering system, wherein the standard question feedback database is used for storing standard questions.
7. An intelligent question-answering testing device, characterized in that the device comprises:
the device comprises a presetting module, a judging module and a judging module, wherein the presetting module is used for presetting test data, and the test data comprises a test question sentence and an expected result;
the sending module is used for sending each test question to the intelligent question-answering system so as to detect the feedback correctness of the intelligent question-answering system and obtain a feedback result;
the analysis module is used for matching and analyzing the feedback result and the expected result to obtain a test result of the test question sentence;
the adjusting module is used for counting the accuracy of the test data based on the test result of each test question sentence, and adjusting the intelligent question-answering system according to the test result when the accuracy is smaller than the preset accuracy;
the intelligent question-answering system comprises a front-end knowledge base, the front-end knowledge base is a local database, and the feedback correctness detection comprises the following steps:
inquiring whether a preset result exists in the front-end knowledge base or not based on the test question sentence;
when the preset result is inquired in the front-end knowledge base, determining that the feedback result comprises the preset result and a front-end type identifier for representing that the preset result comes from the front-end knowledge base;
the desired results include: pre-storing a front end identifier corresponding to the test question sentence; the matching analysis comprises: judging whether the pre-stored front-end identification is the same as the front-end type identification, and if so, representing the test result as that the feedback of the intelligent question answering system responding to the test question is correct; otherwise, the test result is characterized in that the intelligent question-answering system responds to the feedback error of the test question.
8. An electronic device comprising a processor and a memory;
the memory is used for storing computer instructions, and the processor is used for executing the computer instructions stored by the memory to realize the intelligent question-answering test method of any one of claims 1 to 6.
9. A computer-readable storage medium storing one or more programs, the one or more programs being executable by one or more processors to implement a method for testing intelligent question and answer according to any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811083158.4A CN110909133B (en) | 2018-09-17 | 2018-09-17 | Intelligent question and answer testing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811083158.4A CN110909133B (en) | 2018-09-17 | 2018-09-17 | Intelligent question and answer testing method and device, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110909133A CN110909133A (en) | 2020-03-24 |
CN110909133B true CN110909133B (en) | 2022-06-24 |
Family
ID=69813684
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811083158.4A Active CN110909133B (en) | 2018-09-17 | 2018-09-17 | Intelligent question and answer testing method and device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110909133B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112084767B (en) * | 2020-07-22 | 2024-07-23 | 联想(北京)有限公司 | Information response processing method, intelligent voice equipment and storage medium |
CN111881251B (en) * | 2020-07-23 | 2023-10-13 | 平安科技(深圳)有限公司 | AI telephone sales testing method and device, electronic equipment and storage medium |
CN111897936B (en) * | 2020-08-05 | 2023-08-08 | 腾讯科技(深圳)有限公司 | Method, device and equipment for evaluating recall accuracy of question-answering system |
CN111914078A (en) * | 2020-08-13 | 2020-11-10 | 北京捷通华声科技股份有限公司 | Data processing method and device |
CN114692647B (en) * | 2020-12-31 | 2024-06-07 | 北京猎户星空科技有限公司 | Data processing method, device, equipment and medium |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104820635A (en) * | 2015-04-07 | 2015-08-05 | 南京国电南自电网自动化有限公司 | Method of testing robustness of closed loop protocol based on semantic analysis of dynamic script |
CN106599312A (en) * | 2016-12-29 | 2017-04-26 | 上海智臻智能网络科技股份有限公司 | Knowledge base check method and device, and terminal |
CN106599317A (en) * | 2016-12-30 | 2017-04-26 | 上海智臻智能网络科技股份有限公司 | Test data processing method and device for question-answering system and terminal |
CN106782539A (en) * | 2017-01-16 | 2017-05-31 | 上海智臻智能网络科技股份有限公司 | A kind of intelligent sound exchange method, apparatus and system |
CN107193739A (en) * | 2017-05-23 | 2017-09-22 | 张泽祎 | A kind of black box regression testing method |
CN107305579A (en) * | 2016-04-25 | 2017-10-31 | 北京京东尚科信息技术有限公司 | The method of testing and device of intelligent Answer System |
CN107908545A (en) * | 2017-09-26 | 2018-04-13 | 深圳市牛鼎丰科技有限公司 | Interface test method, device, computer equipment and storage medium |
CN108228468A (en) * | 2018-02-12 | 2018-06-29 | 腾讯科技(深圳)有限公司 | A kind of test method, device, test equipment and storage medium |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8984650B2 (en) * | 2012-10-19 | 2015-03-17 | Pearson Education, Inc. | Privacy server for protecting personally identifiable information |
US10628749B2 (en) * | 2015-11-17 | 2020-04-21 | International Business Machines Corporation | Automatically assessing question answering system performance across possible confidence values |
-
2018
- 2018-09-17 CN CN201811083158.4A patent/CN110909133B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104820635A (en) * | 2015-04-07 | 2015-08-05 | 南京国电南自电网自动化有限公司 | Method of testing robustness of closed loop protocol based on semantic analysis of dynamic script |
CN107305579A (en) * | 2016-04-25 | 2017-10-31 | 北京京东尚科信息技术有限公司 | The method of testing and device of intelligent Answer System |
CN106599312A (en) * | 2016-12-29 | 2017-04-26 | 上海智臻智能网络科技股份有限公司 | Knowledge base check method and device, and terminal |
CN106599317A (en) * | 2016-12-30 | 2017-04-26 | 上海智臻智能网络科技股份有限公司 | Test data processing method and device for question-answering system and terminal |
CN106782539A (en) * | 2017-01-16 | 2017-05-31 | 上海智臻智能网络科技股份有限公司 | A kind of intelligent sound exchange method, apparatus and system |
CN107193739A (en) * | 2017-05-23 | 2017-09-22 | 张泽祎 | A kind of black box regression testing method |
CN107908545A (en) * | 2017-09-26 | 2018-04-13 | 深圳市牛鼎丰科技有限公司 | Interface test method, device, computer equipment and storage medium |
CN108228468A (en) * | 2018-02-12 | 2018-06-29 | 腾讯科技(深圳)有限公司 | A kind of test method, device, test equipment and storage medium |
Non-Patent Citations (2)
Title |
---|
Routing questions to appropriate answerers in community question answering services;Baichuan Li等;《CIKM "10: Proceedings of the 19th ACM international conference on Information and knowledge management》;20101026;第1585-1588页 * |
自动问答***中问句相似度计算方法研究;徐海洲;《中国优秀硕士学位论文全文数据库信息科技辑》;20150215(第2期);第I138-1491页 * |
Also Published As
Publication number | Publication date |
---|---|
CN110909133A (en) | 2020-03-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110909133B (en) | Intelligent question and answer testing method and device, electronic equipment and storage medium | |
US10423665B2 (en) | Method and system for generating a conversational agent by automatic paraphrase generation based on machine translation | |
JP2020537777A (en) | Methods and devices for identifying the user's intent of speech | |
US20210357587A1 (en) | An intelligent response method and device | |
CN111368043A (en) | Event question-answering method, device, equipment and storage medium based on artificial intelligence | |
EP2940551B1 (en) | Method and device for implementing voice input | |
CN109726108B (en) | Front-end code testing method, device, system and medium based on analog data | |
US20220358292A1 (en) | Method and apparatus for recognizing entity, electronic device and storage medium | |
US20170243116A1 (en) | Apparatus and method to determine keywords enabling reliable search for an answer to question information | |
CN104267953A (en) | Control and method for importing Word test questions based on browser | |
CN114757176A (en) | Method for obtaining target intention recognition model and intention recognition method | |
CN116595154B (en) | Task processing method and automatic question-answering method | |
CN110362826A (en) | Periodical submission method, equipment and readable storage medium storing program for executing based on artificial intelligence | |
US20180189298A1 (en) | Random Index Pattern Matching Based Email Relations Finder System | |
CN113901837A (en) | Intention understanding method, device, equipment and storage medium | |
CN111708867A (en) | Question and answer query method, device and equipment applied to electric power operation and inspection | |
CN118035401A (en) | Answer generation method, device, electronic equipment and storage medium | |
CN117520497A (en) | Large model interaction processing method, system, terminal, equipment and medium | |
CN112818689B (en) | Entity identification method, model training method and device | |
CN110929014B (en) | Information processing method, information processing device, electronic equipment and storage medium | |
CN117130589A (en) | Code processing method, apparatus, all-in-one machine and computer program product | |
CN111460811A (en) | Crowdsourcing task answer verification method and device, computer equipment and storage medium | |
CN116150376A (en) | Sample data distribution optimization method, device and storage medium | |
WO2023065638A1 (en) | Data retrieval method and apparatus, and electronic device and storage medium | |
CN115906797A (en) | Text entity alignment method, device, equipment and 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 |