TWI653529B - Method for debugging os image and electronic device using the same - Google Patents

Abstract

A method for debugging an operating system image and an electronic device using the same. The method of debugging the operating system image includes the following steps. A basic information corresponding to a question and an execution record are obtained. According to a semantic definition rule and basic information, the type of problem corresponding to the problem is obtained. A target historical execution record is obtained based on the type of problem. A record of the execution of the problem and the target history execution record to achieve a first comparison similarity. In a case where the first alignment similarity is greater than or equal to a first similarity critical value, a historical execution record closing report corresponding to the target historical execution record is obtained.

Description

Method for debugging operating system image and electronic device using same

The present invention relates to a method for operating system image debugging and an electronic device using the same.

Image of the operating system (OS) Many different types of issues can occur during development and mass production. Analysis, debugging, and debugging of problems that occur are also time consuming, and are a significant burden for developers of operating systems. At the same time, the log file related to the problem may be stored in multiple storage paths and debugged for one problem. It may also need to view multiple running records to find the root cause of the problem. Therefore, It takes quite a while to find the right execution record.

The above situation has caused a heavy workload for developers, and it is also not conducive to the work education training for new developers. Therefore, there is a need for a method for debugging the image of the operating system, assisting in the analysis, debugging and debugging of the problems that occur, to provide front-end testers with test directions and suggestions, and to reduce the workload of developers.

The present invention relates to a method for operating system image debugging and an electronic device using the same. When a problem occurs in the production system image during production or testing, the front-end tester or developer can compare the problem with those past cases, and then analyze and reason the problem to generate a problem for the problem. Wrong suggestion.

According to a first aspect of the invention, a method for operating system image debugging is presented. The method for operating system image debugging includes the following steps. A basic information corresponding to a question and an execution record are obtained. According to a semantic definition rule and basic information, the type of problem corresponding to the problem is obtained. A target historical execution record is obtained based on the type of problem. A record of the execution of the problem and the target history execution record to achieve a first comparison similarity. In a case where the first alignment similarity is greater than or equal to a first similarity critical value, a historical execution record closing report corresponding to the target historical execution record is obtained.

According to a second aspect of the present invention, an electronic device for operating system image debugging is provided. The electronic device for debugging the operating system image includes a case database and a semantic recognition module. The case database is used to store a plurality of historical execution records and a plurality of historical execution record closing reports corresponding to historical execution records. The semantic recognition module is configured to obtain a basic information corresponding to a problem, an execution record, a semantic recognition rule and basic information, a problem type corresponding to the problem, and a target according to the problem type. The historical execution record, the execution record of the comparison problem, and the target historical execution record to obtain a first comparison similarity, and obtain a correspondence if the first comparison similarity is greater than or equal to a first similarity threshold A historical execution record closing report on the target history execution record.

In order to better understand the above and other aspects of the present invention, the following detailed description of the embodiments and the accompanying drawings

The various embodiments are described in detail below, however, the examples are intended to be illustrative only and not to limit the scope of the invention. Further, the drawings in the embodiments omits some of the elements to clearly show the technical features of the present invention. The same reference numerals will be used in the drawings to refer to the same or the like.

Please refer to FIG. 1A, which illustrates a schematic diagram of an electronic device 10 for operating system image debugging according to an embodiment of the invention. The electronic device 10 can be, for example, a server such as a local server and a cloud server, or a personal computer capable of performing a large number of operations. The electronic device 10 includes a semantic recognition module 102, a case database 104, and an inference module 106. The semantic recognition module 102 is coupled to the case database 104 and the inference module 106. The case database 104 can be, for example, a hard disk, a flash memory, a read-only memory (ROM), a non-volatile memory (Non-Volatile Memory), or a computer system. The electronic device such as the server executes the database system to store a plurality of historical execution records, common abnormal execution records, and a pass log file, and can store the historical execution records, common abnormal execution records, and the like. Through the execution of the record history record closing report, the common abnormal record closing report and the completion of the record through the execution of the record. The semantic recognition module 102 and the inference module 106 can be, for example, by using a wafer, a circuit block in the wafer, a firmware circuit, a circuit board containing a plurality of electronic components and wires, or a memory module. The storage medium is implemented, and can also be implemented by executing an electronic device or a program by an electronic device such as a computer system or a server.

In another embodiment of the present invention, the electronic device 10 further includes a user interface module 108. As shown in FIG. 1B, the user interface module 108 is coupled to the semantic recognition module 102 and the case database 104. And an inference module 106. The user interface module 108 is used to provide an issue in which the front-end tester or the developer reports the image of the operating system during development and mass production. The information received by the user interface module 108 includes, but is not limited to, basic information about the problem that occurred, an execution record, a probability of occurrence, and a hardware configuration comparison table. The user interface module 108 can also display case information obtained by the semantic recognition module 102, the case database 104, or the reasoning module 106, a suggestion scheme for the problem, and the like. The basic information of the above problems may further include a description of the problem and a copying process of the problem (or called a copying method, a copying step, or a reproducing program). The user interface module 108 can be, for example, a touch screen, a combination of a screen and an input device, or a chip, a circuit block in the chip, a firmware circuit, and a plurality of electronic components. The circuit board of the wire or a storage medium storing the complex array code may be implemented by executing a corresponding software or program by an electronic device such as a computer system or a server.

In another embodiment of the present invention, the user interface module 108 is included in a remote device 20 connected to the electronic device 10. As shown in FIG. 1C, the user interface module 108 may not be included in the electronic device 10. in. The remote device 20 can be, for example, a personal computer, a mobile device, a smart phone, a number of assistants, and a server, but is not limited thereto. The remote device 20 can be connected to the electronic device 10 in a wired or wireless manner through various interfaces.

Referring to FIG. 2, a flow chart of a method for debugging an operating system image according to an embodiment of the present invention is shown. The flowchart of the method of debugging the operating system image shown in FIG. 2 can be applied to the electronic device 10 as shown in FIG. 1A, FIG. 1B or FIG. 1C. In order to clearly explain the operation of the above various elements and the data de-identification method of the embodiment of the present invention, the following will be described in detail with reference to the flowchart of FIG. However, those skilled in the art to which the present invention pertains can understand that the method of the embodiments of the present invention is not limited to the electronic device 10 of FIG. 1A, FIG. 1B or FIG. 1C, nor to FIG. The sequence of steps of the flow chart. The method for debugging the operating system image can be implemented, for example, by a software program. The software program can be stored on a disc, a hard disk or other non-transitory computer readable medium. The software program can include a plurality of computer processors or controllers. Instructions or software programs that can be loaded by an electronic device having a computer processor or controller to perform a data de-identification method. A detailed description of each step is as follows.

Please refer to FIG. 1A, FIG. 1B, FIG. 1C, and FIG. 2 at the same time. According to an embodiment of the present invention, first, in step S202, the semantic recognition module 102 obtains a basic information corresponding to the problem that occurred and a log file. This basic information may, for example, include current conditions at the time of the problem, such as system downtime, the end of an application, status alerts for system anomalies, unintended operational results, and reproducible procedures/steps to reproduce the problem. In step S204, the semantic recognition module 102 obtains a type of question corresponding to the problem that occurs according to a semantic recognition rule and the basic information. That is, the semantic recognition module 102 identifies which type of problem the problem occurred.

The following will further exemplify the basic information of the semantic recognition module 102 according to the semantic recognition rule and the problem that occurs, and obtain the type of the problem corresponding to the problem that occurs. In an embodiment of the invention, important keywords are listed first to create a bag of words as a basis for weight calculation. The preset important keywords may be, for example but not limited to, OOBE, D2D, Push Button Reset, PBR, User Alaunch, Audit Mode, NAPP, Recovery from a Drive, and the like. The semantic recognition module 102 normalizes the problem description in the basic information of the problem that occurs, that is, analyzes the words in the problem description, omits the punctuation marks and articles in the words, to obtain multiple words, each word has a word Hierarchical value. Next, look for these keywords to match the keywords listed in the keyword bag. The semantic recognition module 102 can further calculate the keyword weight of the keyword based on the obtained vocabulary and its corresponding vocabulary hierarchy value, and use the keyword weight as the degree of correlation between the problem and the keyword. For example, the keyword weight can be obtained by subtracting the lexical hierarchy value of the keyword from the maximum value of the vocabulary hierarchy values corresponding to the obtained vocabulary. The semantic recognition rule of the present invention refers to the above-mentioned formalization of paragraphs for problem description, taking out vocabulary and keywords, and calculating keyword weights. The semantic recognition module 102 uses the semantic recognition rules to find the key weights. Then, the semantic recognition module 102 obtains the type of the problem corresponding to the problem that occurs according to the keyword.

In an embodiment, for example, in the basic information of the problem that occurred, the problem description describes "During the PBR Resetting this PC 10%, the system call back to Desktop". Next, the semantic recognition module 102 normalizes the above-mentioned problem description into paragraphs, analyzes the words in the problem description, and omits the punctuation marks and articles in the words to obtain a plurality of words. The semantic recognition module 102 assigns a vocabulary hierarchy value to each vocabulary according to the order of the vocabulary in the problem description. In this embodiment, 10 words can be obtained, each of which has a vocabulary level value, as shown in Table 1 below. FIG. 4A is a schematic diagram of keyword weight calculation according to an embodiment of the invention. The semantic recognition module 102 can sort the acquired vocabulary in the order of the problem description to establish a binary tree as shown in FIG. 4A. Table 1  <TABLE border="1" borderColor="#000000" width="85%"><TBODY><tr><td> vocabulary</td><td> lexical hierarchy value</td></tr><tr ><td> During </td><td> 1 </td></tr><tr><td> PBR </td><td> 2 </td></tr><tr><td> Resetting </td><td> 3 </td></tr><tr><td> PC </td><td> 4 </td></tr><tr><td> 10% </ Td><td> 5 </td></tr><tr><td> system </td><td> 6 </td></tr><tr><td> call </td><td > 7 </td></tr><tr><td> back </td><td> 8 </td></tr><tr><td> to </td><td> 9 </ Td></tr><tr><td> Desktop </td><td> 10 </td></tr></TBODY></TABLE>

In this embodiment, according to the pre-established keyword word bag, the keyword "PBR" matching the keyword bag is found from the above vocabulary, and the vocabulary hierarchy value is 2. Among the vocabulary obtained by analyzing the problem description and its corresponding lexical hierarchy value, the vocabulary with the largest lexical hierarchy value is the last "Desktop" in the sentence of the problem description, and the vocabulary hierarchy value is 10, The keyword "PBR" found in the keyword bag has a lexical hierarchy value of 2. Therefore, the keyword weight of the keyword "PBR" can be the vocabulary hierarchy value 10 of the vocabulary "Desktop" minus the vocabulary hierarchy value 2 of the keyword "PBR", that is, 10-2=8, the key of the keyword "PBR" The word weight is 8.

In this example, the vocabulary obtained after analyzing the problem description has the keyword “PBR”, and the keyword weight is 8, and the semantic recognition module 102 according to the found keyword “PBR” will correspond to the problem that occurs. The problem type is defined as "PBR". Finding the type of problem corresponding to the problem that occurs is beneficial to directly searching for the case related to the problem that occurred in the case database 104, reducing the time of comparison of the case that is not related to the problem, thereby saving system computing time and performance.

In another embodiment of the present invention, the semantic recognition module 102 normalizes the problem description in the basic information of the question, and the obtained plurality of words have two or more keywords listed in the keyword bag. . For example, among the plurality of words obtained after the paragraph is normalized, the two words are consistent with the keywords in the keyword bag, and are the first keyword and the second keyword, respectively. Deriving the lexical hierarchy values of the first keyword and the second keyword respectively by the maximum value in the vocabulary hierarchy values corresponding to the obtained vocabulary to obtain the keyword weight of the first keyword and the keyword weight of the second keyword of. At this time, the semantic recognition module 102 respectively obtains the keyword weight of the first keyword (which may be referred to as the first keyword weight) and the keyword weight of the second keyword (which may be referred to as the second keyword weight). Determining whether a weight difference between the first keyword weight and the second keyword weight is greater than a weight threshold.

When the weight difference between the first keyword weight and the second keyword weight is greater than the weight threshold, that is, the weight difference between the first keyword weight and the second keyword weight is greater than the weight threshold In the case, the semantic recognition module 102 obtains the type of the problem corresponding to the problem that occurs according to one of the first keyword and the second keyword. For example, if the first keyword weight is greater than the second keyword weight, and the weight difference between the first keyword weight and the second keyword weight is greater than the weight threshold, the problem that occurs is compared with the first keyword. Correspondingly, the semantic recognition module 102 uses the first keyword as the type of the problem corresponding to the problem that occurs.

When the weight difference between the first keyword weight and the second keyword weight is less than or equal to the weight threshold, that is, the weight difference between the first keyword weight and the second keyword weight is less than or equal to the weight In the case of the threshold, the semantic recognition module 102 further analyzes the copying program in the basic narrative of the problem that occurs to obtain a first weight adjustment value corresponding to the first keyword and a second corresponding to the second keyword. Weight adjustment value. In an embodiment, the semantic recognition module 102 is configured to include a keyword (eg, a first keyword and/or a second key) in a plurality of steps of the first step of the copying process toward the first step of searching for the copying program. Word) steps. The weight adjustment value of the keyword is obtained according to the number of steps of the copying program and the step hierarchy in which the step of the keyword is located. For example, the semantic recognition module 102 subtracts the step hierarchy value of the step having the keyword by the number of steps of the copy program, and obtains the weight adjustment value corresponding to the keyword. The above-mentioned copying procedure refers to the execution of the steps described in the copying procedure, and the problem that occurs can be reproduced.

Then, the semantic recognition module 102 adjusts the keyword weight of the first keyword (the first keyword weight) and the second weight adjustment value according to the first weight adjustment value to adjust the keyword weight of the second keyword (the second keyword weight) ). Then, the semantic recognition module 102 obtains the type of the problem corresponding to the problem that occurs according to the adjusted first keyword weight and the adjusted second keyword weight. For example, if the weight difference between the adjusted first keyword weight and the adjusted second keyword weight is greater than the weight threshold, the semantic recognition module 102 is based on the keyword with the larger keyword weight ( It may be the first keyword or the second keyword) as the type of question corresponding to the problem that occurred. If the weight difference between the adjusted first keyword weight and the adjusted second keyword weight is still less than or equal to the weight threshold, the semantic recognition module 102 simultaneously uses the first keyword and the second keyword as The type of problem corresponding to the problem that occurs, that is, the problem that occurs in this case will correspond to both problem types, that is, the problem that occurs at the same time is classified into the problem type corresponding to the first keyword and the second key. The type of question corresponding to the word.

For example, in the basic information of the problem that occurred, the problem description describes "Between user alaunch and OOBE page, the system pop up error message", and the copy program (copy method/step) in the basic information is described as follows: Step1 : Install Win10 Image Step2: Image Deployment Step3: The system pop up error message during user alaunch where each step has its step hierarchy value, the step of step 1 of the first step (Step1) is 0, and the step of the second step (Step2) The level value is 1, and the step value of the third step (Step3) is 2. The weight threshold is set to 3.

Next, the semantic recognition module 102 normalizes the above-mentioned problem description to obtain a vocabulary. In this example, the problem with the problem is described by omitting the punctuation and the article in the sentence, and 10 words can be obtained. At the same time, according to the order of each vocabulary in the problem description to establish a binary tree, and give each vocabulary a lexical hierarchy value, as shown in Figure 4B and Table 2 in the following table. FIG. 4B is a schematic diagram of key weight calculation according to another embodiment of the present invention. Table 2  <TABLE border="1" borderColor="#000000" width="85%"><TBODY><tr><td> vocabulary</td><td> lexical hierarchy value</td></tr><tr ><td> Between </td><td> 1 </td></tr><tr><td> user alaunch </td><td> 2 </td></tr><tr><td > and </td><td> 3 </td></tr><tr><td> OOBE </td><td> 4 </td></tr><tr><td> page </ Td><td> 5 </td></tr><tr><td> system </td><td> 6 </td></tr><tr><td> pop </td><td > 7 </td></tr><tr><td> up </td><td> 8 </td></tr><tr><td> error </td><td> 9 </ Td></tr><tr><td> message </td><td> 10 </td></tr></TBODY></TABLE>

In this embodiment, two keywords matching the keyword bag can be found from the above vocabulary according to the keyword word bag, which are the first keyword "user alaunch" and the second keyword "OOBE", respectively, and the vocabulary hierarchy. The values are 2 and 4 respectively. In this example, among the vocabulary obtained by analyzing the problem description and its corresponding lexical hierarchy value, the vocabulary with the largest lexical hierarchy value is the last message "message" in the sentence of the problem description, and the vocabulary hierarchy value is 10, and the first keyword "user alaunch" found by the keyword word bag has a vocabulary hierarchy value of 2 and the second keyword "OOBE" has a vocabulary hierarchy value of 4. Therefore, the keyword weight of the first keyword "user alaunch" may be the vocabulary hierarchy value 10 of the vocabulary "message" minus the lexical hierarchy value 2 of the first keyword "user alaunch", ie 10-2=8, A keyword "user alaunch" has a keyword weight of 8. The keyword weight of the second keyword "OOBE" may be the vocabulary hierarchy value 10 of the vocabulary "message" minus the vocabulary hierarchy value 4 of the keyword "OOBE", ie 10-4=6, the second keyword "OOBE" The keyword weight is 6.

Next, the semantic recognition module 102 compares the keyword weight of the first keyword "user alaunch" with the keyword weight of the second keyword "OOBE", and obtains the keyword weight of the first keyword "user alaunch" and the second key. A weight difference between the keyword weights of the word "OOBE". For example, the weight difference is the keyword weight 8 of the first keyword "user alaunch" minus the keyword weight 2 of the second keyword "OOBE", ie 8-6=2, the first keyword "user alaunch" And the weight difference between the keyword weights of the second keyword "OOBE" is 2.

Therefore, when the weight difference between the keyword weight of the first keyword "user alaunch" and the keyword weight of the second keyword "OOBE" (for example, 2) is less than the weight threshold (for example, 3), the semantic meaning The identification module 102 further analyzes the copying procedure in the basic narrative of the problem that occurred.

In this example, there are three steps in the copying process (the number of steps to copy the program is 3), which are: Step1: Install Win10 Image Step2: Image Deployment Step3: the system pop up error message during user alaunch The module 102 can search for and obtain the first keyword "user alaunch" in the third step (Step3) in the copying process, and the semantic recognition module 102 subtracts the first keyword "user alaunch" by the number of steps of the copying program. The step value of the third step is to obtain the weight adjustment value of the first keyword "user alaunch", that is, 3 (number of steps) - 2 (step value of the third step) = 1 (first keyword "user" The weight adjustment value of alaunch", the weight adjustment value of the first keyword "user alaunch" is 1. The semantic recognition module 102 adds the keyword weight 8 of the first keyword "user alaunch" to the weight adjustment value 1 of the first keyword "user alaunch", and obtains the adjusted keyword weight of the first keyword (first Keyword weight) 9, which is 8+1=9.

The semantic recognition module 102 fails to find and obtain the second keyword "OOBE" in the three steps in the copying process. Therefore, the weight adjustment value of the second keyword "OOBE" is set to zero. In an embodiment, since the weight adjustment value of the second keyword "OOBE" is set to 0, and the second keyword "OOBE" is not found in the copying program, the semantic recognition module 102 further sets the second keyword "OOBE". The keyword weight is set to 0 to reduce the correlation between the second keyword and the problem that occurred.

The adjusted first keyword has a keyword weight (first keyword weight) of 9, and the adjusted second keyword has a keyword weight (second keyword weight) of 0, and the weight difference between the two is It is 9 and greater than the weight threshold (for example, 3). Therefore, the semantic recognition module 102 uses the keyword having the larger keyword weight as the type of the problem corresponding to the problem that occurs. That is, in the embodiment, after the keyword weights of the first keyword and the second keyword are adjusted, the keyword weight of the first keyword "user alaunch" and the keyword of the second keyword "OOBE" The weight difference between the weights (for example, 9) is greater than the weight threshold (for example, 3), and the first keyword "user alaunch" has a larger keyword weight, and therefore, the semantic recognition module 102 uses the first keyword. "user alaunch" is the type of problem that corresponds to the problem that occurred.

Please refer to FIG. 1A, FIG. 1B, FIG. 1C, and FIG. 2 again. After the semantic recognition module 102 identifies in step S204 what type of problem the problem is, in step S206, the semantic recognition module 102 obtains the plurality of historical execution records stored by the case database 104 according to the problem type. A target history execution record corresponding to this type of question. In step S208, the semantic recognition module 102 compares the similarity between the execution record of the problem occurring and the target history execution record to obtain a first comparison similarity. In step S210, the semantic recognition module 102 determines whether the first alignment similarity is less than a first similarity threshold (for example, 80%). For example, if the similarity is 90%, 90% of the contents recorded in the execution record and the target historical execution record of the problem are the same.

If the first comparison similarity is greater than or equal to the first similarity critical value (the determination result of step S210 is NO), in step S212, the semantic recognition module 102 obtains one corresponding to the target historical execution record from the case database 104. Historical execution record closing report. The historical execution record closing report is used as the recommended treatment plan for the problem and the analysis result, so that the front-end testers and developers can obtain the follow-up feasible treatment. In an embodiment, the semantic recognition module 102 can transmit a historical execution record closing report of the suggested processing solution as a problem to the user interface module 108, and the user interface module 108 displays the historical execution record closing report and The recommended processing plan and analysis results corresponding to the historical execution record closing report are for the front-end testers and developers to view.

If the first comparison similarity is less than the first similarity threshold (YES in step S210), in step S214, the semantic recognition module 102 obtains at least one common abnormal execution record stored by the case database 104. And compare the similarity between the execution record of the problem occurring and the common abnormal execution record to obtain a second comparison similarity. Subsequently, in step S216, the semantic recognition module 102 determines whether the second alignment similarity is less than the first similarity threshold (for example, 80%).

If the second comparison similarity is greater than or equal to the first similarity critical value (the determination result of step S216 is NO), in step S218, the semantic recognition module 102 obtains one of the common abnormal execution records from the case database 104. Common abnormal execution record closing report. The common abnormal execution record closing report is used as the recommended solution and the analysis result of the problem, so that the front-end testers and developers can understand the subsequent feasible treatment. In an embodiment, the semantic recognition module 102 can transmit the common abnormal execution record closing report to the user interface module 108, and the user interface module 108 displays the common abnormal execution record closing report, corresponding to the common abnormal execution record. The proposed solution and analysis results of the closing report are for the front-end testers and developers to view.

If the second comparison similarity is less than the first similarity threshold (YES in step S216), in step S220, the semantic recognition module 102 notifies the inference module 106 to analyze the execution record of the problem that occurred to generate and Get a proposal. In an embodiment, after the notification inference module 106 obtains a suggestion scheme, the notification scheme is transmitted to the user interface module 108 to provide a follow-up feasible manner for the developer or the front-end tester. In an embodiment, after the developer or the front-end tester receives the suggestion plan and performs the follow-up action according to the recommended plan, the developer or the front-end tester can provide feedback according to the situation encountered in the process of performing the follow-up action of the recommendation. (eg, if the proposed solution is feasible) to the inference module 106.

Next, in step S222, the inference module 106 integrates the basic information of the problem that occurred, the execution record, and the proposal corresponding to the problem that occurred, as a brand new case, and stores the new case in the case database 104 in step S224. In one embodiment, the inference module 106 integrates the basic information of the problem, the execution record, the suggested solution corresponding to the problem that occurred, and the feedback provided by the developer or front-end tester for the proposed solution as a new case.

Please refer to FIG. 1A, FIG. 1B, FIG. 1C and FIG. FIG. 3 is a flow chart showing an execution record of an issue occurring to obtain a suggested solution according to an embodiment of the invention. In step S220 of FIG. 2, the inference module 106 analyzes the execution record of the problem that occurred to obtain a suggested solution. Steps S302 to S310 of FIG. 3 further explain the flow of the analysis execution record of step S220 of FIG. 2 to obtain the recommended plan. In an embodiment, the inference module 106 can include a semantic knowledge network (not shown) that combines past knowledge, experience, analysis of operational records, and generates recommendations through analysis of operational records. A suggestion is provided to the user interface module 108.

In the semantic knowledge network of the inference module 106 or the inference module 106, an inference analysis word bag is pre-established, which includes a proper noun, a verb or a verb clause that performs an operation on the proper noun or generates an action. And adjectives or adjective clauses that describe abnormal results. The proper nouns in the reasoning analysis bag can include the name of the driver, the application name (for example, Microsoft Office, Internet Explorer), device identification information, event identification information, error codes, and the like. The verb or verb clause in the inference analysis bag may include install, restore, deploy, delete, open, and the like. The adjective or adjective clause in the reasoning analysis bag may include fail, abnormal, crash, critical error, and the like. In addition, a proposal package is also pre-established, which includes proper nouns, suggested verbs or suggested verb clauses that suggest operations on proper nouns, and suggested nouns or suggested noun clauses. The proper nouns in the suggested word bag may include the name of the driver, the application name, device identification information, event identification information, error code, and the like. Suggested verbs or suggested verb clauses in the suggested program bag may include remove, uninstall, update, upgrade, roll back, and the like. Suggested nouns or suggested noun clauses in the suggested word bag may include verify again, reproduce more than 5 runs, and the like.

In an embodiment, in step S302, the inference module 106 obtains a pass log according to the type of the problem corresponding to the execution record of the problem that occurred, that is, the problem of performing the record and the occurrence The system belongs to the same question type. Execution records refer to the execution records of an operating system image that did not cause any problems during production or testing. In step S304, the inference module 106 compares the execution record of the problem that occurred and performs a record to obtain a different content between the two.

In another embodiment, in step S302, the inference module 106 finds another target history execution record from the history execution record stored in the case database, and the execution of the other target history execution record and the occurrence problem There is a third alignment similarity between the records, the third alignment similarity being greater than or equal to a second similarity threshold (eg, 50%) and less than the first similarity threshold (eg, 80%). In step S304, the inference module 106 compares the execution record of the problem that occurred and the other target history execution record to obtain a different content between the two.

Next, in step S306, the inference module 106 analyzes the dissimilar content obtained in step S304, and obtains a proper noun in the dissimilar content according to the inference analysis term bag established in advance, and also finds corresponding content in the dissimilar content. An action of a proper noun or a verb or verb clause that produces an action, and an adjective or adjective clause that describes an abnormal or abnormal result corresponding to a proper noun. That is to say, the reasoning module 106 obtains a proper noun according to the dissimilar content and a verb or verb clause corresponding to the proper noun and/or an adjective or adjective clause.

In step S308, the inference module 106 is based on the proper noun obtained in step S306, the verb/verb clause corresponding to the proper noun, and/or the adjective/adjective clause corresponding to the proper noun. A suggested verb or verb clause is obtained in the bag. In another embodiment, the inference module 106 is based on the proper noun obtained in step S306, the verb/verb clause corresponding to the proper noun, and the adjective/adjective clause corresponding to the proper noun. A suggestive verb or verb clause and a suggested noun or noun clause are obtained in the word bag.

Then, in step S310, the inference module 106 generates and obtains a suggestion scheme according to the suggested verb/verb clause and proper noun. In another embodiment, the inference module 106 generates and obtains a suggestion based on the suggested verb/verb clause, proper noun, and suggested noun/noun clause.

For example, the type of problem that occurs is "PBR", and the description of the problem with the basic information of the problem is described as "During PBR resetting this PC 30%, the system call back to desktop". The difference between the execution record of the problem and the execution of the record is as follows: 1. Failed to open [C:\$SysReset\OldOS\Program Files\WindowsApps\Microsoft.MicrosoftOfficeHub_17.7909.7600. 0_x64 FilesCommonX64] for delete[gle=0x00000005] 2. Failed to delete reparse point [C:\$SysReset\OldOS\Program Files\WindowsApps\Microsoft.MicrosoftOfficeHub_17.7909.7600.0_x64][gle=0x00000005]

The semantic network of the inference module 106 and/or the inference module 106 analyzes the above-mentioned different content, and obtains the proper nouns "Microsoft OfficeHub_17.7909.7600.0", the verbs "open" and "delete", and the adjective "Failed". Through the architecture of the semantic knowledge network of the inference module 106 and/or the inference module 106, it can be inferred that there may be problems in opening and deleting the Microsoft OfficeHub 17.7909.7600.0 version during the PBR process.

The semantic network of the inference module 106 and/or the inference module 106 is based on the inferentially analyzed proper nouns, verb/verb clauses, and adjective/adjective clauses, and is obtained through the pre-established suggestion word bag. The program's proper term "MicrosoftOfficeHub_17.7909.7600.0", the suggested verb "remove", "delete" or "upgrade", and the suggested noun clause "verify again". The semantic knowledge network of the inference module 106 and/or the inference module 106 establishes and obtains a suggestion scheme based on proper nouns, suggested verbs/verb clauses, and/or noun/noun clauses. In an embodiment, the suggestion may be, for example, a suggested verb/verb clause plus a proper noun. In another embodiment, the suggested scheme may be, for example, a suggested verb/verb clause plus a proper noun plus a noun/noun clause. For example, the semantic network of the inference module 106 and/or the inference module 106 obtains the following two suggestions according to the above-mentioned proper nouns, suggested verb/verb clauses, and suggested noun/noun clauses: Remove Microsoft Office Hub 17.7909.7600 and verify again 2. Upgrade Microsoft Office Hub 17.7909.7600 and verify again

The semantic knowledge network of the reasoning module 106 and/or the reasoning module 106 described above is centered on the inference analysis term bag and/or the proper noun in the suggestion word bag, if the word bag and/or suggestion cannot be analyzed in the inference. The finding of a proper noun in the program word bag indicates that the problem occurring at this time is a problem built into the operating system. Therefore, the suggestion solution of the semantic module of the inference module 106 and/or the inference module 106 can be, for example, "Suggest filing bugs to Microsoft."

At this time, the former tester or the developer obtains the analysis result and the suggestion solution provided by the semantic knowledge network of the inference module 106 and/or the inference module 106, and can know the recommended actions to be taken subsequently. If the problem occurs is a brand new case, the reasoning module 106 finally collects the basic information of the problem, the process of reasoning analysis, the execution record, the settlement report, etc. into a new case and transmits it to the case database. , for storage.

According to the method and the electronic device for debugging the image of the operating system according to the embodiment of the present invention, when a problem occurs during the production or testing of the operating system image, the front-end tester or the developer can set the background and basic problem of the problem. Information and execution records, etc. are provided to the electronic device (for example, a server) for debugging the operating system image. Interpret the type of problem based on the basic information of the problem and decide to compare it with those past cases (including cases of problems, common exceptions, and cases passed), and decide to review the execution of those past cases. Record to determine whether the problem occurred is a stand-alone problem. The method and the electronic device for operating system image debugging proposed by the embodiments of the present invention analyze and reason the problems that occur, thereby generating a debugging solution for the problem that occurs.

The method and the electronic device for debugging the image of the operating system according to the embodiment of the present invention provide a solution of artificial intelligence, which can reduce the workload of front-end testers and developers, and reduce labor expenditure. Furthermore, each module in the operating system image debugging electronic device proposed by the embodiment of the present invention has its own execution log, which is not easy to be faked, and reduces the probability and risk of fraud.

In conclusion, the present invention has been disclosed in the above embodiments, but it is not intended to limit the present invention. A person skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, the scope of the invention is defined by the scope of the appended claims.

10‧‧‧Electronic devices

102‧‧‧Speech Identification Module

104‧‧‧Case database

106‧‧‧Inference Module

108‧‧‧User interface module

20‧‧‧ Remote device

S202～S224‧‧‧ Process steps

S302～S310‧‧‧ Process steps

FIG. 1A is a schematic diagram of an electronic device for debugging a working system image according to an embodiment of the invention. FIG. 1B is a schematic diagram of an electronic device for debugging a working system image according to another embodiment of the present invention. FIG. 1C is a schematic diagram showing the connection of an electronic device for operating system image debugging and a remote device having a user interface module according to another embodiment of the invention. FIG. 2 is a flow chart showing a method for debugging a working system image according to an embodiment of the invention. FIG. 3 is a flow chart of analyzing a problem record to obtain a suggested solution according to an embodiment of the invention. FIG. 4A is a schematic diagram of keyword weight calculation according to an embodiment of the invention. FIG. 4B is a schematic diagram of key weight calculation according to another embodiment of the present invention.

Claims (18)

A method for debugging an image of a working system, comprising: obtaining a basic information corresponding to a problem and an execution record; obtaining a type of the problem corresponding to the problem according to the semantic recognition rule and the basic information; The type obtains a target historical execution record; compares the execution record of the problem with the target history execution record to obtain a first comparison similarity; and the first comparison similarity is greater than or equal to a first similarity In the case of the threshold value, a historical execution record closing report corresponding to the target history execution record is obtained. The method of claim 1, further comprising: in the case that the first comparison similarity is less than the first similarity threshold, comparing the execution record of the problem with a common abnormal execution record, And obtaining a second comparison similarity degree; and in the case that the second comparison similarity is greater than or equal to the first similarity critical value, obtaining a common execution abnormal record closing report corresponding to the common abnormal execution record. The method of claim 2, further comprising: analyzing the execution record of the problem in the case that the second alignment similarity is less than the first similarity threshold to obtain a proposal. For example, the method described in claim 3 includes: integrating the basic information of the problem, the execution record and the proposal as a brand new case; and storing the new case. The method of claim 3, wherein the step of analyzing the execution record of the problem to obtain the proposal includes: obtaining a target by executing the record according to the type of the problem; comparing the execution record of the problem with the The goal is to obtain a different content by executing the record; obtaining a proper noun according to the different content and a verb and an adjective corresponding to the proper noun; according to the proper noun and the corresponding to the proper noun The verb and the adjective obtain a suggested verb; and obtain the suggestion according to the suggested verb and the proper noun. The method of claim 3, wherein the step of analyzing the execution record of the problem to obtain the recommended solution comprises: obtaining another target historical execution record according to the type of the problem, the another target historical execution record and A third comparison similarity between the execution records is greater than or equal to a second similarity threshold and less than the first similarity threshold; comparing the execution record of the problem with the other target historical execution record to Obtaining a different content; obtaining a proper noun according to the different content and a verb and an adjective corresponding to the proper noun; obtaining the proper noun and the verb corresponding to the proper noun and the adjective according to the proper noun a suggested verb; and obtaining the proposed scheme based on the suggested verb and the proper noun. The method of claim 1, wherein the step of obtaining the type of the problem corresponding to the problem according to the semantic recognition rule and the basic information comprises: describing a keyword bag and a problem description of the basic information , get a keyword; and get the type of the question corresponding to the question based on the keyword. The method of claim 1, wherein the step of obtaining the type of the problem corresponding to the problem according to the semantic identification rule and the basic information comprises: obtaining a plurality of words according to a problem description in the basic information, Each of the vocabulary words has a vocabulary class value; according to a keyword word bag, a first keyword and a second keyword are found out from the vocabulary words; and the first value is determined according to a maximum value of the vocabulary class values And the first keyword weight of the first keyword and a second keyword weight of the second keyword; If the weight difference between the keyword weight and the second keyword weight is greater than a weight threshold, the problem corresponding to the problem is obtained according to the first keyword and the second keyword. Types of. The method of claim 7, wherein the step of obtaining the keyword weight of the basic information according to the semantic recognition rule further comprises: between the first keyword weight and the second keyword weight When the weight difference is less than or equal to a weight threshold, a first weight adjustment value corresponding to the first keyword and a second corresponding to the second keyword are obtained according to a copying procedure in the basic information. Weight adjustment value; adjusting the first key weight and the second key weight respectively according to the first weight adjustment value and the second weight adjustment value; and adjusting the adjusted first key weight and adjusted The second keyword weights the type of the problem corresponding to the problem. An electronic device for debugging an image of a working system, comprising: a case database for storing a plurality of historical execution records and a plurality of historical execution record closing reports corresponding to the historical execution records; and a semantic recognition module And obtaining a basic information corresponding to a problem, an execution record, a semantic recognition rule and the basic information, obtaining a type of the problem corresponding to the problem, and obtaining one of the historical records according to the type of the problem a target history execution record, an execution record of the problem, and the target history execution record to obtain a first comparison similarity, and wherein the first alignment similarity is greater than or equal to a first similarity threshold In the case, a historical execution record closing report corresponding to the target history execution record is obtained. The electronic device of claim 10, wherein the case database is further configured to store a common abnormal execution record and a common abnormal execution record closing report corresponding to the common abnormal execution record file, and the semantic recognition model. The group is further configured to perform a second comparison similarity by comparing the execution record of the problem with the common abnormality when the first alignment similarity is less than the first similarity threshold. When the second comparison similarity is greater than or equal to the first similarity critical value, a common abnormal execution record closing report corresponding to the common abnormal execution record is obtained. The electronic device of claim 11, further comprising: an inference module, configured to analyze the execution record of the problem if the second comparison similarity is less than the first similarity threshold To get a suggested solution. The electronic device of claim 12, wherein the reasoning module is further configured to integrate the basic information of the problem, the execution record and the proposal as a brand new case, and store the new case in the case data. Library. The electronic device of claim 12, wherein the case database is further configured to store a plurality of execution records, and the reasoning module is further used to compare the execution record of the problem and the execution records a target corresponding to the type of the problem by performing a record to obtain a dissimilar content, obtaining a proper noun according to the dissimilar content, and a verb and an adjective corresponding to the proper noun, according to the proper noun And the verb corresponding to the proper noun and the adjective, obtaining a suggested verb, and obtaining the suggestion according to the suggested verb and the proper noun. The electronic device of claim 12, wherein the reasoning module is further configured to compare the execution record of the problem with another target historical execution record of the historical execution records to obtain a different content, Obtaining a suggested verb according to the dissimilar content to obtain a proper noun and a verb and an adjective corresponding to the proper noun, according to the proper noun and the verb corresponding to the proper noun and the adjective, and Obtaining the suggestion according to the suggested verb and the proper noun; wherein a third comparison similarity between the other target historical execution record and the execution record is greater than or equal to a second similarity threshold and less than The first similarity threshold. The electronic device of claim 10, wherein the semantic recognition module obtains a keyword according to a problem description of the basic information and a keyword word bag, and obtains the corresponding question according to the keyword The type of problem. The electronic device of claim 10, wherein the semantic recognition module obtains a plurality of words according to a problem description in the basic information, each of the words having a vocabulary level value, according to a keyword word bag, Finding a first keyword and a second keyword from the vocabulary, determining a maximum value of the vocabulary hierarchy values, a vocabulary hierarchy value of the first keyword, and a vocabulary hierarchy of the second keyword a value, a first keyword weight of the first keyword, and a second keyword weight of the second keyword, and a weight difference between the first keyword weight and the second keyword weight When the value is greater than a weight threshold, the type of the problem corresponding to the problem is obtained according to one of the first keyword and the second keyword. The electronic device of claim 17, wherein the semantic recognition module, when the weight difference between the first keyword weight and the second keyword weight is less than or equal to a weight threshold Obtaining, according to a copying procedure in the basic information, a first weight adjustment value corresponding to the first keyword, and a second weight adjustment value corresponding to the second keyword, according to the first weight adjustment value, and the a second weight adjustment value, respectively adjusting the first keyword weight and the second keyword weight, and obtaining the corresponding problem according to the adjusted first keyword weight and the adjusted second keyword weight The type of problem.