Construction method and device of OTA upgrade package and electronic equipment

Abstract

The embodiment of the invention relates to the technical field of software upgrade processing, and discloses an OTA upgrade package construction method and device and electronic equipment, and the method comprises the steps: obtaining an upgrade information set which is an information set for classifying a plurality of upgrade contents after collecting the upgrade contents fed back by a plurality of users; performing priority ranking on the plurality of upgrading contents of the upgrading information set, and constructing a development task by adopting the upgrading contents of the upgrading information set according to a priority ranking result; after the development data is collected, the OTA upgrade package is constructed by adopting the development data, and the development data is code data which is developed by technicians according to development tasks after the development tasks are allocated to the technicians. By means of the technical scheme, upgrading requirements and application problems of multiple users can be collected at the same time, tasks are automatically arranged according to the priorities of the requirements of the multiple users, the corresponding upgrading packages are constructed for upgrading processing, the upgrading processing process can be simplified, and the upgrading processing efficiency can be effectively improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of software upgrade processing, and specifically to a method, device and electronic device for constructing an OTA upgrade package. Background Art

[0002] With the rapid development of science and technology, the computing power of hardware devices is becoming more and more powerful, and more and more functions can be realized. In the process of using different devices, different users have increasing personal needs. Users can select the functions they want on the device, and then send function activation or OTA upgrade packages to the device through the cloud to meet the personalized needs of users.

[0003] Since different devices have different developers, they have their corresponding management systems (for example, different car series or models have their corresponding management systems). Therefore, when it is necessary to build OTA upgrade packages for users according to different management systems, the current common method is to manually determine the version and format of the management system, and then edit each user's required functions into the corresponding OTA upgrade package according to the corresponding format and send it to the smart terminal for terminal upgrade to meet the needs of each user.

[0004] However, the commonly used methods currently have the following technical problems: manual development and editing are not only prone to errors, but the upgrade process for each user needs to execute the above process once, resulting in a long upgrade process and low efficiency. It often takes a lot of time to integrate and complete the build process, making it difficult to meet the upgrade needs of multiple different users. Summary of the invention

[0005] In view of the above problems, the embodiments of the present invention provide a method, device and electronic device for constructing an OTA upgrade package, which are used to solve one or more of the above technical problems existing in the prior art.

[0006] According to one aspect of an embodiment of the present invention, a method for constructing an OTA upgrade package is provided, the method comprising:

[0007] Acquire an upgrade information set, wherein the upgrade information set is an information set obtained by classifying and processing multiple upgrade contents after collecting upgrade contents fed back by multiple users;

[0008] Prioritizing the multiple upgrade contents of the upgrade information set, and constructing a development task using the upgrade contents of the upgrade information set according to the priority ranking result;

[0009] After collecting the development data, the OTA upgrade package is constructed using the development data, wherein the development data is code data that the technicians complete the development according to the development tasks after the development tasks are assigned to the technicians.

[0010] In an optional manner, the step of using the development data to construct an OTA upgrade package includes:

[0011] The preset automated coding tool is called to integrate the development data to obtain an OTA upgrade package.

[0012] According to another aspect of an embodiment of the present invention, a device for constructing an OTA upgrade package is provided, including:

[0013] An acquisition module, used to acquire an upgrade information set, wherein the upgrade information set is an information set that classifies and processes multiple upgrade contents after collecting upgrade contents fed back by multiple users;

[0014] A sorting and task building module, used to prioritize the multiple upgrade contents of the upgrade information set, and build a development task using the upgrade contents of the upgrade information set according to the priority ranking result;

[0015] The upgrade package construction module is used to construct an OTA upgrade package using the development data after collecting the development data, wherein the development data is the code data developed by the technicians after the development tasks are assigned to the technicians.

[0016] In an optional manner, the step of using the development data to construct an OTA upgrade package includes:

[0017] The preset automated coding tool is called to integrate the development data to obtain an OTA upgrade package.

[0018] According to another aspect of an embodiment of the present invention, there is provided an electronic device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;

[0019] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute the operation of the method for constructing the OTA upgrade package as described above.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the storage medium stores at least one executable instruction, and the executable instruction enables an electronic device / apparatus to perform the following operations:

[0021] Acquire an upgrade information set, wherein the upgrade information set is an information set obtained by classifying and processing multiple upgrade contents after collecting upgrade contents fed back by multiple users;

[0022] Prioritizing the multiple upgrade contents of the upgrade information set, and constructing a development task using the upgrade contents of the upgrade information set according to the priority ranking result;

[0023] After collecting the development data, the OTA upgrade package is constructed using the development data, wherein the development data is code data that the technicians complete the development according to the development tasks after the development tasks are assigned to the technicians.

[0024] The embodiment of the present invention collects upgrade contents fed back by multiple users, and classifies the multiple upgrade contents to form an upgrade information set; prioritizes the multiple upgrade contents of the upgrade information set, and constructs development tasks using the upgrade contents of the upgrade information set according to the result of the priority sorting; after collecting development data, the development data is used to construct an OTA upgrade package, which can simultaneously collect upgrade requirements and application problems of multiple users for upgrade processing, which can not only simplify the upgrade processing process, but also effectively improve the efficiency of the upgrade processing.

[0025] The above description is only an overview of the technical solution of the embodiment of the present invention. In order to more clearly understand the technical means of the embodiment of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiment of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings are only used to illustrate the embodiments and are not to be considered as limiting the present invention. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings. In the accompanying drawings:

[0027] Figure 1 A schematic diagram showing a flow chart of a first embodiment of a method for constructing an OTA upgrade package provided by the present invention;

[0028] Figure 2 A connection diagram of the OTA operation management platform provided by the present invention is shown;

[0029] Figure 3 A schematic diagram showing a flow chart of a second embodiment of a method for constructing an OTA upgrade package provided by the present invention;

[0030] Figure 4 A schematic diagram showing the structure of a first embodiment of a device for constructing an OTA upgrade package provided by the present invention is shown;

[0031] Figure 5 A schematic structural diagram of an embodiment of an electronic device provided by the present invention is shown. DETAILED DESCRIPTION

[0032] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0033] Figure 1 A flow chart of a first embodiment of a method for constructing an OTA upgrade package provided by the present invention is shown. The method can be executed by a background server or an electronic device, and the background server or the electronic device can be an OTA operation management platform.

[0034] Reference Figure 2 , shows a connection diagram of the OTA operation management platform provided by the present invention. The OTA operation management platform can be connected with different smart terminals and vehicles.

[0035] In one embodiment, with the rapid development of the automobile manufacturing industry, automobile companies are facing increasingly complex OTA upgrade management needs. Frequent upgrades of different car series, models and versions require an efficient management system. However, the existing manual management mode is inefficient and has a high error rate, making it difficult to meet the needs of parallel management across models and multiple versions. The current OTA upgrade package construction method requires manual development and editing. This operation method is not only prone to errors, but also requires the above process to be executed once for each user's upgrade process, resulting in a long upgrade process and low efficiency. It often takes a lot of time to integrate to complete the construction process, which makes it difficult to meet the upgrade needs of multiple different users.

[0036] Aiming at the construction scenario of OTA upgrade package for a vehicle, the present invention proposes a method, device, equipment and computer-readable storage medium for constructing an OTA upgrade package, aiming to solve the above-mentioned problems and realize efficient and automated parallel management across vehicle models and multiple versions.

[0037] In one implementation, the method is applicable to a vehicle backend management system or backend server, such as Figure 1 As shown, the method comprises the following steps:

[0038] Step 110: Acquire an upgrade information set, where the upgrade information set is an information set obtained by classifying the upgrade contents fed back by multiple users.

[0039] In one embodiment, information may be collected from different users, and the information may be feedback information about upgrading vehicles or feedback information about different problems existing in the system.

[0040] For example, user A wants to upgrade the interactive interface of the vehicle system, but believes that the existing interactive interface has few options.

[0041] For another example, user B finds that the wake-up word recognition of the vehicle's voice assistant is often wrong and hopes to modify the wake-up word.

[0042] For another example, user C finds that the vehicle's computer has freezes and a black screen and hopes for improvements.

[0043] After collecting feedback from multiple different users, an upgrade information set can be formed as a set of upgrade requirements for a user. Subsequently, an OTA upgrade package can be built based on the content of the information set to take into account the needs of multiple different users. There is no need to perform an upgrade process for each user's upgrade process, which can greatly shorten the processing time and improve the efficiency of the upgrade process.

[0044] Different users have different information, which may be information about upgrade requirements, feedback issues, tips or personal opinions. The way or channel of collecting information may also be different. It may be information uploaded directly by the user on the vehicle system or vehicle-mounted system, or information sent by the user through his terminal to the backend server, or information collected and entered by the technician online. The information collected through different channels has different formats. In order to unify various information, as an example, obtaining the upgrade information set may include the following sub-steps:

[0045] Step 1101: Acquire upgrade requirement information of multiple users, each of which includes application problem information, new function information and prompt information reported by the user.

[0046] Step 1102: pre-process each of the upgrade information to obtain processed information, wherein the pre-processing includes: text cleaning, text standardization and unified formatting.

[0047] Step 1103: Call a classification algorithm to classify the information content of each processed information to obtain upgrade information, and combine multiple upgrade information to obtain an upgrade information set. The classification processing includes classifying application problem information based on emotional tendency, classifying new function information based on product entity content, and classifying prompt information according to information type.

[0048] In one actual operation method, information can be collected by monitoring social media. For example, user comments and posts related to xx brand models / series products can be captured from social media platforms (such as Weibo, Douyin, WeChat, etc.) to collect user information.

[0049] It is also possible to receive user feedback information, for example, to receive information collected from direct feedback from user feedback systems (such as mini programs, H5, Apps, customer service chats, etc.).

[0050] Information can also be collected through questionnaires. For example, APP / Mini Program questionnaires can be sent regularly to collect users' specific needs and feedback on the product.

[0051] Information can also be collected through import, for example, accepting specific product requirements and issues imported in a fixed format (Excel).

[0052] After a plurality of upgrade requirement information are received, each upgrade information may be preprocessed to obtain processed information, wherein the preprocessing includes: text cleaning, text standardization and unified formatting.

[0053] The text cleaning process may include removing HTML tags, special characters, stop words, etc.

[0054] The text standardization process may be to convert the text into a unified format and encoding.

[0055] The unified formatting process can be to uniformly clean the collected data through big data (clean up invalid feedback, including positive / negative reviews without substantive content and feedback without practical meaning. These feedbacks often do not provide valuable information and may interfere with subsequent analysis), and escape them, so as to achieve unified formatting.

[0056] Since the collected information describes various contents, such as upgrade requirements and system problems, it is necessary to classify the information input by each user to determine what the information input by the user includes.

[0057] In one embodiment, sentiment analysis may be performed on each user's information to analyze the sentiment tendency (positive, negative, neutral) in the user's feedback information.

[0058] Entity recognition is then performed on each user’s information to identify entity issues such as product features, functions, and problems mentioned in user feedback.

[0059] Before analysis and identification, topic modeling and classification can be performed first: cluster the user feedback information into different topics or categories (such as business categories: car series, model, vehicle (frame number), parts, vehicle OS version, complaints, feedback, problems, bugs, feature suggestions, etc.); then analyze and identify it to determine the specific category.

[0060] Among them, the classification of physical problems can refer to bugs or technical problems in the product. The classification of non-physical problems may involve users' suggestions for improving existing functions, demands for new functions, or feedback on product interaction and experience).

[0061] In one operation mode, sentiment analysis can be performed using natural language processing (NLP) technology to automatically analyze the sentiment tendency in the text. This can be achieved by using a pre-trained deep learning model.

[0062] For the identification and classification of entity issues, for defined feedback categories (such as functional defects, performance issues, interface suggestions, etc.), automated tools or algorithms can be used to classify feedback. This involves text classification techniques such as Naive Bayes, Support Vector Machine (SVM) or deep learning models.

[0063] Optionally, keyword matching can also be used to determine the category of information provided by the user. For example, keyword matching can be used to automatically identify and classify feedback related to a specific problem or function. For example, if the feedback contains keywords such as "crash" or "error", it can be automatically classified as technical feedback.

[0064] By classifying the content of each user's information, the category of the information reported by the user can be determined, and then upgrade processing can be performed on the category to make the upgrade more targeted, thereby improving the accuracy of the upgrade and meeting the actual needs of the user.

[0065] Step 120: Prioritize the multiple upgrade contents of the upgrade information set, and construct a development task using the upgrade contents of the upgrade information set according to the priority ranking result.

[0066] In one embodiment, the upgrade information set includes information of multiple users, and the information of different users may be different. For example, the upgrade content of user A's feedback information is the upgrade wake-up word; the upgrade content of user B's feedback information is the upgrade interface; the upgrade content of user C's feedback information is the screen freeze problem; the upgrade content of user D's feedback information is the problem of multiple connection errors with the user's smart terminal.

[0067] If an OTA upgrade package is generated for multiple needs and issues at the same time, not only will the data capacity of the OTA upgrade package be large, the transmission and upgrade will take a long time, and the development will also take a long time, making it difficult to meet the needs of all users at once in a short period of time. To this end, the multiple upgrade contents of the upgrade information set can be prioritized according to different issues and needs to determine which upgrade content has a higher priority and needs to be upgraded immediately, and which upgrade content has a lower priority and can be upgraded later.

[0068] After determining the priority ranking of multiple upgrade contents, the multiple upgrade contents can be arranged one by one according to the ranking results, and then the development tasks are constructed in the order of arrangement, and the R&D personnel are prompted to perform development processing one by one to generate the corresponding OTA upgrade package.

[0069] For example, the four upgrade contents of the upgrade information set are A, B, C, and D. The priority order is C, D, A, and B. You can first build a development task according to the upgrade content of C to generate an OTA upgrade package that meets the upgrade content of C; then, build a development task according to the upgrade content of D to generate an OTA upgrade package that meets the upgrade content of D, and so on, until the OTA upgrade package of the upgrade content of B is finally generated.

[0070] In one embodiment, the content of the user feedback information may be a problem, an upgrade requirement, or both. In order to comprehensively calculate the two situations to determine the priority of the user feedback information, as an example, the priority sorting of the multiple upgrade contents of the upgrade information set may include the following sub-steps:

[0071] Step 1201: Call a preset scoring model to calculate the scoring value of each upgraded content in the upgrade information set to obtain a priority scoring value, wherein the preset scoring model is a model trained by word segmentation after collecting user feedback questions and requirements and performing word segmentation, and each priority scoring value corresponds to an upgraded content of a user.

[0072] Since the information content involved by the user includes multiple types and the amount of information is large, in order to improve the efficiency of calculating the priority score value, a scoring model can be pre-trained, and then the pre-trained scoring model can be called to calculate the score value of each upgrade content in the upgrade information set.

[0073] In actual operation, each upgrade content of the upgrade information set can be directly input into the pre-trained scoring model, and the pre-trained scoring model can directly output the priority scoring value corresponding to each upgrade content.

[0074] Because the upgrade content involves multiple contents such as problems and requirements, when training the scoring model, you can pre-collect the feedback problem content of multiple different users and various upgrade requirement contents, segment the various information contents, and use the segmentation to train the model, so that you can get a scoring model; when the scoring model calculates the priority score value of each upgrade content, it can also segment each upgrade content and calculate the priority score value based on its segmentation.

[0075] Step 1202: sorting the multiple upgrade contents of the upgrade information set according to the priority score values ​​to obtain a priority sorting.

[0076] After calculating the priority score value, the multiple upgrade contents in the upgrade information set may be sorted according to the numerical values ​​of the priority score value to obtain the priority sorting of each upgrade content.

[0077] For example, the upgrade information set has four upgrade contents, A, B, C, and D, and the calculated priority scores are 10, 20, 18, and 6, respectively. Then, if they are arranged from large to small according to the numerical values ​​of the priority scores, the order obtained is 20, 18, 10, 6, and the priority sorting result is B, C, A, D.

[0078] Sort the requirements and issues according to the calculated priority scores. Requirements and issues with higher priority scores should be processed first.

[0079] Subsequently, issues, defects, and requirements can be automatically created: according to the above strategy, important issues, defects, and requirements are filtered out, and JIRAopenAPI is called to automatically enter the JIRA requirement defects and problem management related issue sections (the project name of the related issue section is generally named according to the car series / model name). The following is an example of JIRA issue fields (each Project section field varies according to requirements):

[0080] Vehicle series, model, problem type, problem channel, person who raised the problem, summary, description, severity, priority, label, whether it occurs occasionally, vehicle version, vehicle VIN, controller, accessory list, occurrence time, expected closure time, etc.

[0081] Since the upgrade content involves problems and requirements, in order to score different contents respectively, in an optional embodiment, the preset scoring model includes: a problem scoring model, a requirement scoring model and a weight distribution model.

[0082] Among them, the problem scoring model is used to score the upgrade content according to the information content in the upgrade content about describing the problem; similarly, the requirement scoring model is used to score the upgrade content according to the information content in the upgrade content about describing the upgrade requirement.

[0083] As an example, the step of invoking a preset scoring model to calculate the scoring value of each upgrade content of the upgrade information set to obtain a priority scoring value may include the following sub-steps:

[0084] Step 12011: Call the problem scoring model to determine the severity scoring value of each upgrade content in the upgrade information set, call the requirement scoring model to determine the prevalence scoring value of each upgrade content in the upgrade information set, and call the weight allocation model to determine the severity weight value and prevalence weight value of each upgrade content in the upgrade information set.

[0085] Step 12012: Calculate the product of the severity score value and the severity weight value of each upgrade content to obtain the problem product value of each upgrade content, and calculate the product of the prevalence score value and the prevalence weight value of each upgrade content to obtain the demand product value of each upgrade content.

[0086] Step 12013: Calculate the sum of the problem product value and the requirement product value to obtain the priority score value of each upgrade content.

[0087] The priority score is calculated based on the severity of the problem and the prevalence of user needs.

[0088] In an actual operation mode, the problem scoring model can be called to calculate the severity scoring value of each upgrade content in the upgrade information set, the requirement scoring model can be called to calculate the prevalence scoring value of each upgrade content in the upgrade information set, and then the weight allocation model can be called to determine the severity weight value and prevalence weight value of each upgrade content in the upgrade information set.

[0089] The severity score value of the upgraded content indicates the severity of the upgraded content; the prevalence score value of the upgraded content indicates the prevalence of the upgraded content.

[0090] For example, the severity score value can be calculated by using the problem scoring model to perform word frequency statistics on the upgraded content, and count the frequency of each word in user feedback. High-frequency words usually have greater attention.

[0091] Next, the upgrade content can be contextually analyzed through the question scoring model, and the sentiment and problem expressed by the word can be judged by combining the context of each word in the specific feedback. For example, "crash" appears in negative feedback and is related to key functions (key functions can be pulled out based on the featureList of each vehicle version / product version), and the severity is high.

[0092] In addition, users can pre-define rules and set the severity scores of specific words based on the existing knowledge base and expert experience. For example, words such as "flash back", "crash", "unusable", and "blank page" are preset with high severity scores.

[0093] Moreover, the problem scoring model can be a classification model trained by machine learning, and the problem scoring model maps each word to a severity score. The input of the model can be the word itself and its context features, and the output is the severity score value.

[0094] Through the above steps, the severity of each word can be quantified for subsequent priority sorting.

[0095] For severity assessment, a severity score can be set for each problem based on factors such as the urgency of the problem, the scope of impact, and the cost of repair. For example, a score range of 1-10 can be set, where 1 represents very low and 10 represents very high. It should be noted that the prevalence score can also be calculated in the same way as above.

[0096] For the prevalence assessment, a prevalence score can be set for each requirement based on factors such as the number of people who proposed the requirement, the size of the user group, the market potential of the requirement, etc. A rating range of 1-10 can also be used.

[0097] The priority score value may be calculated as a weighted sum of the severity score and the prevalence score, ie, priority score = severity score * severity weight + prevalence score * prevalence weight.

[0098] Using the above example, assuming that the four upgrade contents of the upgrade information set are A, B, C, and D, the upgrade content A can be input into the problem scoring model, and the severity score and severity weight of the upgrade content A can be output. The upgrade content A is input into the demand scoring model, and the prevalence score and prevalence weight of the upgrade content A can be output.

[0099] Next, the product of the severity score value and the severity weight value of the upgraded content A can be calculated to obtain the problem product value of the upgraded content A, and then the product of the universality score value and the universality weight value of the upgraded content A can be calculated to obtain the demand product value of the upgraded content A. Finally, the sum of the problem product value and the demand product value is calculated to obtain the priority score value of the upgraded content A.

[0100] The processing methods of the remaining B, C, and D upgrade contents can be calculated by referring to the above method, so as to obtain the priority score values ​​of the A, B, C, and D upgrade contents.

[0101] As an example, the training operation of the question scoring model may include the following sub-steps:

[0102] S21. Obtain a user's problem data set, where the problem data set includes problem data fed back by multiple users, and each problem data is annotated with a severity score.

[0103] S22, performing word segmentation processing on each question data in the question data set to obtain question segmentation.

[0104] S23, extracting context features from the question segmentation to obtain segmentation features, and performing sentiment analysis on each question data to obtain sentiment results.

[0105] S24. After mapping the sentiment result and the word segmentation feature to question word segmentations respectively, model training is performed using the mapped question word segmentations to obtain a question scoring model.

[0106] Specifically, data collection can be performed first, in which a large amount of problem data reported by users can be collected, and the severity score of each piece of problem data reported can be marked.

[0107] Next, the question data may be subjected to text segmentation, and each piece of feedback question data may be subjected to word segmentation processing to obtain a word segmentation result.

[0108] Next, context features may be extracted from the question data, and context features of the word segments of the question data may be extracted, such as preceding and following words, sentence structure, etc.

[0109] The sentiment of each question data can then be analyzed to determine the sentiment tendency (positive, negative, neutral) of each feedback question data.

[0110] Then, the vocabulary of each word segment is scored and mapped, and each word segment can be mapped to a severity score by combining contextual features and sentiment analysis results. This can be achieved using predefined rules or trained machine learning models.

[0111] The model can be optimized later, using the collected annotated data to train the machine learning model, taking the word segmentation and its context features as input and the severity score as output. The model is continuously optimized to improve the scoring accuracy.

[0112] Similarly, the training operation of the demand scoring model can also be trained in the above manner.

[0113] As an example, the training operation of the demand scoring model may include the following sub-steps:

[0114] S31. Obtain a user demand data set, where the demand data set includes functional demand data fed back by multiple users, and each functional demand data is annotated with a score on universality.

[0115] S32: Perform context analysis on each real-time functional requirement data to obtain a general score for each word in the functional requirement data.

[0116] S33, using the word segmentation of the functional requirement data and the general score of the word segmentation to perform model training to obtain a requirement scoring model.

[0117] Specifically, data collection can be performed first to collect functional requirement data from user feedback, and the universality score of each piece of functional requirement data from feedback can be marked.

[0118] Then, the word frequency statistics can be performed on the functional requirement data of each feedback to count the frequency of occurrence of each word in the functional requirement data of the user feedback. High-frequency words usually have a higher prevalence.

[0119] When conducting user group analysis based on functional requirement data, we can analyze the characteristics of the user group that provides feedback and determine the size of the user group covered by the feedback.

[0120] By using the functional requirement data of each feedback to perform context analysis, we can combine the context of the word segmentation in the specific feedback to determine the universality of the demand expressed by the word.

[0121] Before training, users can also pre-define rules and set the prevalence score of specific words based on existing knowledge base and expert experience.

[0122] After the rules are defined, model training and optimization can be performed. The collected annotated data can be used to train the machine learning model, with the word segmentation and its context features as input and the prevalence score as output. The model is continuously optimized to improve the scoring accuracy.

[0123] Similarly, the training operation of the weight distribution model can also be trained in the above manner.

[0124] The training operation of the weight distribution model includes:

[0125] S41. Acquire an emergency data set of a user, wherein the emergency data set includes emergency problem data fed back by a plurality of users, and each emergency problem data is marked with an emergency level value.

[0126] S42: Perform context analysis on each urgent question data to obtain the urgency value of each word in the urgent question data.

[0127] S43, using the word segmentation of the urgent problem data and the urgency value of the word segmentation to perform model training to obtain a weight distribution model.

[0128] In one embodiment, the weights of severity and prevalence can be set according to the company's needs and strategic goals. For example, if the company attaches more importance to responding to urgent issues quickly, a higher weight can be set for severity; if the company attaches more importance to meeting the needs of a wide range of users, a higher weight can be set for prevalence.

[0129] Specifically, the training operation of the weight distribution model may be as follows:

[0130] First, data collection can be performed to collect urgent problem data reported by users, and the urgency value of each piece of urgent problem data reported can be marked.

[0131] Then, we can perform context analysis on the urgent problem data of each feedback, and judge the urgency of the problem expressed by the word based on the context of the word segmentation in the specific feedback. For example, words related to safety issues (intelligent driving safety, cabin safety, driving safety, etc.) are usually more urgent.

[0132] Before training, users can also pre-define rules and set the urgency score of specific words based on existing knowledge base and expert experience.

[0133] After the rule definition is completed, model training and optimization can be carried out. The collected annotated data can be used to train the machine learning model, with the word segmentation and its context features as input and the urgency score as output. The model is continuously optimized to improve the scoring accuracy.

[0134] For weight setting, weights are set according to urgency and company strategic goals. For example, words with high urgency are given higher weights, and words with high prevalence are given higher weights. Weights can be determined by expert scoring or model calculation.

[0135] After arranging the upgrade contents, a corresponding development task can be generated for each upgrade content. According to the company's R&D domain division table (table format: including domain, R&D leader, R&D manager, front-end development, back-end development, front-end native, testing, architect, product and other dimensions, which are updated regularly), the corresponding project owner can be associated to automatically create stories, tasks, issues, etc. in JIRA. The corresponding owner can choose to accept or reject the relevant tasks, and the accepted tasks will be tracked and implemented in the iteration plan.

[0136] Version planning can plan the upgraded iterative versions of the product based on factors such as priority, development resources, and schedule. You can first determine the priority, which is the priority score value.

[0137] It can evaluate development resources and analyze the available resources of the development team, including the skills, experience and available working hours of the developers. (Jira integrated Tempo plug-in or other agile tools have this function).

[0138] You can also create a schedule for development tasks, including start and end times.

[0139] It can also automatically plan and use optimization algorithms (such as linear programming and genetic algorithms) to automatically calculate the best version planning solution. ), integrate team calendars, work time management tools, etc., and automatically calculate available development time and resources.

[0140] Different upgrade contents require different time and have different difficulties. In order to generate corresponding development tasks for the upgrade contents and allow the adapted technical personnel to quickly develop the required data, as an example, the upgrade contents of the upgrade information set are used to construct development tasks according to the priority sorting result, which may include the following sub-steps:

[0141] S51. Obtain the level value of the technician developing and upgrading the software.

[0142] S52: Calling a preset evaluation model to calculate the task working hours in turn by using the grade value and each upgraded content of the upgrade information set according to the result of priority sorting.

[0143] S53. Generate development tasks using the task hours corresponding to each upgrade content.

[0144] In one embodiment, in order to reasonably develop according to the actual situation of the technicians, the working time parameters may be first obtained, and the working time parameters include: the working time data required for the technicians to develop the upgraded software and the level value of the technicians to develop the upgraded software.

[0145] The work time data may include the total available hours (Total Available Hours, (TAH)) and the allocated hours (Allocated Hours, (AH)) of each developer.

[0146] The level value may include the skill level and experience level (SL) of the developer. Then, the available working hours of each developer may be calculated based on the working hours data and the level value. Specifically, the available working hours of each developer may be as shown in the following formula:

[0147] Available Hours(AH i )=Total Available Hours(TAH i )-Allocated Hours(AH i );

[0148] Then the time required for the task (Estimated TaskHours, (ETH)) can be estimated based on historical data and task complexity. This can be calculated using a regression model or empirical formula, as shown below:

[0149] Estimated Task Hours(ETH j )=α+β1×Task Typeke+β2×Skill Level(SL i )+…+β n ×Other Factors;

[0150] The preset evaluation model may be a historical data regression model, or a model obtained by combining historical data and machine learning training. The preset evaluation model may be shown as follows:

[0151] Predicted Task Hours(PTH j)=α+β1×Task Complexity+β2×Skill Level+…;

[0152] In the above formula, α = intercept, β = regression coefficient.

[0153] After calculating the task hours corresponding to each upgrade content, the corresponding development task can be generated according to its task hours. Specifically, the developer with the corresponding level can be found according to the level value, and then the available hours of the developer can be obtained. If the available hours of the developer are greater than the task hours, the development task of the developer can be generated.

[0154] For example, consider the following developer and task data:

[0155] Developer 1: Total available working hours 160 hours, assigned working hours 40 hours, skill level 3.

[0156] Developer 2: Total available working hours 160 hours, assigned working hours 60 hours, skill level 2.

[0157] Upgrade content A corresponds to development task A: Advanced skills (skill level 3) are required, and the estimated task time is 50 hours.

[0158] Upgrade content B corresponds to development task B: intermediate skills (skill level 2) are required, and the estimated task time is 30 hours.

[0159] Calculate the current available working hours of the two developers as follows:

[0160] Available Hours of A=160-40=120;

[0161] Available Hours of B=160-60=100;

[0162] When a development task is assigned, if the available working hours are greater than the estimated task working hours, the development task can be executed.

[0163] Optionally, you can also confirm the developer of the corresponding function (problem->function->owner) according to the R&D domain division table (function-owner).

[0164] Assume that development task A is assigned to developer 1 (skills match), and development task B is assigned to developer 2 (skills match):

[0165] Then the available working hours of developer A are:

[0166] Updated Available Hours of A=120-50=70;

[0167] Then the available working hours of developer B are:

[0168] Updated Available Hours of B=100-30=70;

[0169] In another optional embodiment, the available development resources of the developers can also be calculated, and the development resources include personnel and their skill levels. Assuming that there are multiple tasks to be assigned, the total available development time and resources are calculated as follows:

[0170] The total available development time is as follows:

[0171]

[0172] The total available resources are as follows:

[0173] For each open task j select matching developer i:

[0174] ResourceMatching Score(RMS ij )=γ1×Skill Level Match(SLM ij )+γ2×Experience Level Match(ELM ij )+…;

[0175] Allocating development tasks and calculating available development time can match development tasks according to resources and then adjust the available time. The adjustment of available time is shown in the following formula:

[0176] Updated Available Hours(UAH i )=TAH i -(AH i +ETH j );

[0177] Then, according to the task time, the working hours are allocated. Specifically, the developer u with the highest resource matching degree can be selected to allocate working hours ETH for task j. i , the allocation of working hours can be shown as follows:

[0178] Assigned Hours(AH ij )=ETH j ;

[0179] Finally, open tasks can be constructed based on the allocated working hours.

[0180] In addition, you can also make a schedule to set the start and end time for each development task. The start time is the idle time of the assigned task owner who has not assigned tasks in the time management tool. If the assigned owner has a task preemption, he will be replaced by an alternative. If multiple alternatives still have tasks preemption, the task will be assigned to the R&D manager to coordinate resources and the company will send WeChat / email reminders. The end time is based on the following criteria: First, experience-based evaluation: If similar tasks have been completed in the past, historical data can be used to estimate the working hours. Second, the working hours are evaluated according to the job level of the responsible person and the skills required for the task (the HR system has relevant information), such as senior R&D engineer (level 3), intermediate R&D engineer (level 3), and ordinary R&D engineer (level 1). The R&D cycle can be set to a fixed value of 1 day to 2 weeks. The required R&D leader tracks the problem repair status and function iteration status every day to ensure on-time delivery.

[0181] Through the above method, the available development time and resources can be automatically calculated, and tasks can be reasonably planned and allocated to improve the efficiency and accuracy of the OTA management system.

[0182] Step 130: After collecting the development data, the OTA upgrade package is constructed using the development data, wherein the development data is code data that the technicians complete the development according to the development task after the development task is assigned to the technicians.

[0183] After the development task is built, the development task will be assigned to the corresponding developer or technician, who will develop according to the development task and upload the code data after the development is completed to obtain the code data.

[0184] Finally, the development data can be used to build an OTA upgrade package, which can be used to modify bugs in the vehicle or system or to improve functionality.

[0185] In an actual operation mode, regarding the production of OTA upgrade packages, the developed code can be integrated, packaged / deployed, and then put into testing (automated testing tools can be introduced). When the test is completed, the OTA operation management platform subscribes to the JIRA test completion task status notification webhook, automatically calls the OTA platform API, and produces an OTA software package for a single ECU function. According to the ECU list of each vehicle model, an OTA vehicle upgrade package is produced.

[0186] Specifically, the process of making a vehicle OTA upgrade package may include the following steps:

[0187] Step 1, Preparation: Collect all the changes that need to be included in the upgrade package from the requirements management system (such as JIRA). Confirm the priority and status of these changes, and ensure that they have passed the test and are ready for release. Code merging and version control, merge the relevant code from the version control system (such as Git) into a specific branch such as master. Generate version tags.

[0188] Step 2: Compile and build (compile code): Use the compilation toolchain (such as Makefile, CMake) to compile the code. Ensure that there are no errors during the compilation process and generate the target binary file. Build a single ECU software package and build an independent software package for each electronic control unit (ECU). Generate the executable file and related configuration files for a single ECU.

[0189] Step 3: Integration and packaging (integrate single ECU software packages and generate vehicle software packages): Integrate all single ECU software packages into one overall package. Ensure that the dependencies between different ECUs are handled correctly to avoid conflicts. Generate an OTA vehicle upgrade package and make a vehicle upgrade package that includes all ECU software packages. Including (such as vehicle model, version number, upgrade instructions, etc.), the software package version is named according to the rules, and the software package status becomes ready.

[0190] At this stage, the following tools can be introduced to improve efficiency:

[0191] For example, use code generation tools and automated unit testing frameworks (such as JUnit, TestNG, JMeter, etc.) to improve code development and testing efficiency.

[0192] You can also use continuous integration platforms, such as Jenkins, GitLab CI, etc., to automate code review, integration, and deployment to ensure code quality and version stability.

[0193] Then, establish the corresponding upgrade activities and tasks on the OTA operation management platform, select the corresponding vehicle version software package, circle the vehicle to be upgraded, and after BPM approval, synchronize the upgrade activities to the OTA platform, and notify the vehicle users to upgrade by issuing tasks on OTA.

[0194] Upgrade monitoring: Monitor the upgrade process to ensure the success of the upgrade and collect user feedback after the upgrade.

[0195] In addition, it is possible to upgrade based on user feedback loops. For example, you can collect user feedback on the new version, including problem reports, feature suggestions, etc.

[0196] Then analyze user feedback to identify new problems or new needs. Finally, perform iterative optimization, incorporate new user feedback into the needs analysis of the next iteration, and continuously optimize the product.

[0197] Figure 3 The flowchart of the second embodiment of the method for constructing an OTA upgrade package provided by the present invention is shown. The method is executed by an electronic device. Figure 3 As shown, the method comprises the following steps:

[0198] The first step is data collection.

[0199] The second step is data preprocessing.

[0200] The third step is NLP classification processing.

[0201] Step 4: Automatically enter requirements and defects.

[0202] Step 5: Automatically formulate upgrade iteration plans.

[0203] Step 6: OTA upgrade implementation.

[0204] Step 7. Feedback loop.

[0205] In this embodiment, the embodiment of the present invention provides a method for constructing an OTA upgrade package, and its beneficial effects are: the present invention can obtain an upgrade information set, which is an information set that classifies and processes multiple upgrade contents after collecting upgrade contents from multiple users; prioritizes multiple upgrade contents in the upgrade information set, and constructs development tasks using the upgrade contents in the upgrade information set according to the result of the priority sorting; after collecting development data, the development data is used to construct an OTA upgrade package; the present invention collects upgrade requirements and application problems from multiple users at the same time, automatically arranges tasks according to the priorities of multiple user requirements, and constructs corresponding upgrade packages for upgrade processing, which not only simplifies the upgrade processing process and effectively improves the efficiency of the upgrade processing; and the processing process can handle multiple problems at the same time, which can further shorten the time and improve efficiency.

[0206] Figure 4 FIG. 1 is a schematic diagram showing the structure of a first embodiment of a device for constructing an OTA upgrade package provided by the present invention. Figure 4 As shown, the device 300 includes: an acquisition module 310, a sorting and building task module 320 and a building upgrade package module 330.

[0207] An acquisition module 310 is used to acquire an upgrade information set, wherein the upgrade information set is an information set that classifies and processes multiple upgrade contents after collecting upgrade contents fed back by multiple users;

[0208] The sorting and task building module 320 is used to prioritize the multiple upgrade contents of the upgrade information set, and build a development task using the upgrade contents of the upgrade information set according to the priority ranking result;

[0209] The upgrade package building module 330 is used to build an OTA upgrade package using the development data after collecting the development data, wherein the development data is code data developed by the technicians after the development tasks are assigned to the technicians.

[0210] In an optional manner, the prioritizing the multiple upgrade contents of the upgrade information set includes:

[0211] Calling a preset scoring model to calculate the scoring value of each upgrade content in the upgrade information set to obtain a priority scoring value, wherein the preset scoring model is a model trained by word segmentation after collecting user feedback questions and requirements and performing word segmentation, and each priority scoring value corresponds to an upgrade content of a user;

[0212] The multiple upgrade contents of the upgrade information set are sorted according to the priority score values ​​to obtain a priority sorting.

[0213] In an optional manner, the preset scoring model includes: a question scoring model, a demand scoring model and a weight distribution model;

[0214] The calling of a preset scoring model to calculate the scoring value of each upgrade content of the upgrade information set to obtain a priority scoring value includes:

[0215] Calling the problem scoring model to determine the severity score value of each upgrade content of the upgrade information set, calling the requirement scoring model to determine the prevalence score value of each upgrade content of the upgrade information set, and calling the weight allocation model to determine the severity weight value and prevalence weight value of each upgrade content of the upgrade information set;

[0216] Calculate the product of the severity score value and the severity weight value of each upgrade content to obtain the problem product value of each upgrade content, and calculate the product of the prevalence score value and the prevalence weight value of each upgrade content to obtain the demand product value of each upgrade content;

[0217] The priority score value of each upgrade content is obtained by calculating the sum of the problem product value and the demand product value.

[0218] In an optional manner, the training operation of the question scoring model includes:

[0219] Obtain a user's problem data set, wherein the problem data set includes problem data fed back by multiple users, and each problem data is annotated with a severity score;

[0220] Performing word segmentation processing on each question data of the question data set to obtain question word segmentations;

[0221] Extracting context features from the question segmentation to obtain segmentation features, and performing sentiment analysis on each question data to obtain sentiment results;

[0222] After the sentiment result and the word segmentation feature are respectively mapped to the question word segmentations, the mapped question word segmentations are used for model training to obtain a question scoring model.

[0223] In an optional manner, the training operation of the demand scoring model includes:

[0224] Obtaining a user's demand data set, wherein the demand data set includes functional demand data fed back by multiple users, and each functional demand data is annotated with a score on universality;

[0225] Performing context analysis on each real-time functional requirement data to obtain a general score of each word in the functional requirement data;

[0226] The word segmentation of the functional requirement data and the general score of the word segmentation are used for model training to obtain a requirement scoring model.

[0227] In an optional manner, the training operation of the weight distribution model includes:

[0228] Acquire an emergency data set of a user, wherein the emergency data set includes emergency problem data fed back by multiple users, and each emergency problem data is marked with an emergency level value;

[0229] Performing context analysis on each urgent question data to obtain the urgency value of each word in the urgent question data;

[0230] The word segmentation of the urgent problem data and the urgency value of the word segmentation are used to perform model training to obtain a weight distribution model.

[0231] In an optional manner, constructing a development task using the upgrade content of the upgrade information set according to the result of the priority sorting includes:

[0232] Obtain the level value of the technician developing and upgrading the software;

[0233] Calling a preset evaluation model to calculate the task working hours in turn using the grade value and each upgraded content of the upgrade information set according to the result of priority sorting;

[0234] The development tasks are generated using the task hours corresponding to each upgrade content.

[0235] In an optional manner, obtaining the upgrade information set includes:

[0236] Acquire upgrade requirement information of multiple users, each of which includes application problem information, new function information and prompt information reported by the user;

[0237] Preprocessing each of the upgrade information to obtain processed information, wherein the preprocessing includes: text cleaning, text standardization and unified formatting;

[0238] A classification algorithm is called to classify the information content of each processed information to obtain upgrade information, and multiple upgrade information are combined to obtain an upgrade information set. The classification processing includes classifying application problem information based on emotional tendency, classifying new function information based on product entity content, and classifying prompt information according to information type.

[0239] In an optional manner, the step of using the development data to construct an OTA upgrade package includes:

[0240] The preset automated coding tool is called to integrate the development data to obtain an OTA upgrade package.

[0241] Figure 5 The schematic diagram of the structure of an embodiment of the electronic device provided by the present invention is shown, and the specific embodiment of the present invention does not limit the specific implementation of the electronic device.

[0242] like Figure 5 As shown, the electronic device may include: a processor (processor) 402 , a communication interface (Communications Interface) 404 , a memory (memory) 406 , and a communication bus 408 .

[0243] Wherein: the processor 402, the communication interface 404, and the memory 406 communicate with each other through the communication bus 408. The communication interface 404 is used to communicate with other devices such as a client or other server network elements. The processor 402 is used to execute the program 410, which can specifically execute the relevant steps in the above-mentioned embodiment of the method for constructing an OTA upgrade package.

[0244] Specifically, the program 410 may include program code including computer executable instructions.

[0245] The processor 402 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the electronic device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0246] The memory 406 is used to store the program 410. The memory 406 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0247] Program 410 can be specifically called by processor 402 to enable the electronic device to execute the method for constructing the OTA upgrade package provided in the above embodiment.

[0248] Specifically, you can do the following:

[0249] Acquire an upgrade information set, wherein the upgrade information set is an information set obtained by classifying and processing multiple upgrade contents after collecting upgrade contents fed back by multiple users;

[0250] Prioritizing the multiple upgrade contents of the upgrade information set, and constructing a development task using the upgrade contents of the upgrade information set according to the priority ranking result;

[0251] After collecting the development data, the OTA upgrade package is constructed using the development data, wherein the development data is code data that the technicians complete the development according to the development tasks after the development tasks are assigned to the technicians.

[0252] In an optional manner, the program 410 is called by the processor 402 to enable the electronic device to execute the method for constructing the OTA upgrade package provided in the above embodiment.

[0253] Specifically, you can do the following:

[0254] Acquire an upgrade information set, wherein the upgrade information set is an information set obtained by classifying and processing multiple upgrade contents after collecting upgrade contents fed back by multiple users;

[0255] Prioritizing the multiple upgrade contents of the upgrade information set, and constructing a development task using the upgrade contents of the upgrade information set according to the priority ranking result;

[0256] After collecting the development data, the OTA upgrade package is constructed using the development data, wherein the development data is code data that the technicians complete the development according to the development tasks after the development tasks are assigned to the technicians.

[0257] An embodiment of the present invention provides a computer-readable storage medium, which stores at least one executable instruction. When the executable instruction is executed on an electronic device / apparatus, the electronic device / apparatus executes the method for constructing an OTA upgrade package in any of the above method embodiments.

[0258] The executable instructions can be specifically used to enable the electronic device / apparatus to execute the method for constructing the OTA upgrade package provided in the above embodiment.

[0259] Specifically, you can do the following:

[0260] Acquire an upgrade information set, wherein the upgrade information set is an information set obtained by classifying and processing multiple upgrade contents after collecting upgrade contents fed back by multiple users;

[0261] Prioritizing the multiple upgrade contents of the upgrade information set, and constructing a development task using the upgrade contents of the upgrade information set according to the priority ranking result;

[0262] After collecting the development data, the OTA upgrade package is constructed using the development data, wherein the development data is code data that the technicians complete the development according to the development tasks after the development tasks are assigned to the technicians.

[0263] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system or other device. In addition, the embodiments of the present invention are not directed to any particular programming language.

[0264] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. Similarly, in order to simplify the present invention and help understand one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, the various features of the embodiments of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. Wherein, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, wherein each claim itself is a separate embodiment of the present invention.

[0265] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive.

[0266] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order. These words may be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be understood as limitations on the order of execution.

Claims

1. A method for constructing an OTA upgrade package, characterized in that: The method comprises: Acquire an upgrade information set, wherein the upgrade information set is an information set obtained by classifying and processing multiple upgrade contents after collecting upgrade contents fed back by multiple users; Prioritizing the multiple upgrade contents of the upgrade information set, and constructing a development task using the upgrade contents of the upgrade information set according to the priority ranking result; After collecting the development data, the OTA upgrade package is constructed using the development data, wherein the development data is code data that the technicians complete the development according to the development tasks after the development tasks are assigned to the technicians.

2. The method according to claim 1, characterized in that The prioritizing of the multiple upgrade contents of the upgrade information set includes: Calling a preset scoring model to calculate the scoring value of each upgrade content in the upgrade information set to obtain a priority scoring value, wherein the preset scoring model is a model trained by word segmentation after collecting user feedback questions and requirements and performing word segmentation, and each priority scoring value corresponds to an upgrade content of a user; The multiple upgrade contents of the upgrade information set are sorted according to the priority score values ​​to obtain a priority sorting.

3. The method according to claim 2, characterized in that The preset scoring model includes: a question scoring model, a demand scoring model and a weight distribution model; The calling of a preset scoring model to calculate the scoring value of each upgrade content of the upgrade information set to obtain a priority scoring value includes: Calling the problem scoring model to determine the severity score value of each upgrade content of the upgrade information set, calling the requirement scoring model to determine the prevalence score value of each upgrade content of the upgrade information set, and calling the weight allocation model to determine the severity weight value and prevalence weight value of each upgrade content of the upgrade information set; Calculate the product of the severity score value and the severity weight value of each upgrade content to obtain the problem product value of each upgrade content, and calculate the product of the prevalence score value and the prevalence weight value of each upgrade content to obtain the demand product value of each upgrade content; The priority score value of each upgrade content is obtained by calculating the sum of the problem product value and the demand product value.

4. The method according to claim 3, characterized in that The training operation of the question scoring model includes: Obtain a user's problem data set, wherein the problem data set includes problem data fed back by multiple users, and each problem data is annotated with a severity score; Performing word segmentation processing on each question data of the question data set to obtain question word segmentations; Extracting context features from the question segmentation to obtain segmentation features, and performing sentiment analysis on each question data to obtain sentiment results; After the sentiment result and the word segmentation feature are respectively mapped to the question word segmentations, the mapped question word segmentations are used for model training to obtain a question scoring model.

5. The method according to claim 3, characterized in that: The training operation of the demand scoring model includes: Obtaining a user's demand data set, wherein the demand data set includes functional demand data fed back by multiple users, and each functional demand data is annotated with a score on universality; Performing context analysis on each real-time functional requirement data to obtain a general score of each word in the functional requirement data; The word segmentation of the functional requirement data and the general score of the word segmentation are used for model training to obtain a requirement scoring model.

6. The method according to claim 3, characterized in that The training operation of the weight distribution model includes: Acquire an emergency data set of a user, wherein the emergency data set includes emergency problem data fed back by multiple users, and each emergency problem data is marked with an emergency level value; Performing context analysis on each urgent question data to obtain the urgency value of each word in the urgent question data; The word segmentation of the urgent problem data and the urgency value of the word segmentation are used to perform model training to obtain a weight distribution model.

7. The method according to claim 1, characterized in that The step of constructing a development task using the upgrade content of the upgrade information set according to the result of the priority sorting includes: Obtain the level value of the technician developing and upgrading the software; Calling a preset evaluation model to calculate the task working hours in turn using the grade value and each upgraded content of the upgrade information set according to the result of priority sorting; The development tasks are generated using the task hours corresponding to each upgrade content.

8. The method according to any one of claims 1 to 7, characterized in that: The step of obtaining the upgrade information set includes: Acquire upgrade requirement information of multiple users, each of which includes application problem information, new function information and prompt information reported by the user; Preprocessing each of the upgrade information to obtain processed information, wherein the preprocessing includes: text cleaning, text standardization and unified formatting; A classification algorithm is called to classify the information content of each processed information to obtain upgrade information, and multiple upgrade information are combined to obtain an upgrade information set. The classification processing includes classifying application problem information based on emotional tendency, classifying new function information based on product entity content, and classifying prompt information according to information type.

9. A device for constructing an OTA upgrade package, characterized in that: The device comprises: An acquisition module, used to acquire an upgrade information set, wherein the upgrade information set is an information set that classifies and processes multiple upgrade contents after collecting upgrade contents fed back by multiple users; A sorting and task building module, used to prioritize the multiple upgrade contents of the upgrade information set, and build a development task using the upgrade contents of the upgrade information set according to the priority ranking result; The upgrade package construction module is used to construct an OTA upgrade package using the development data after collecting the development data, wherein the development data is the code data developed by the technicians after the development tasks are assigned to the technicians.

10. An electronic device, characterized in that: include: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform the operation of the OTA upgrade package construction method as described in any one of claims 1-8.

Images