Interface design method and device, electronic equipment and storage medium

Abstract

An interface design method and device, electronic equipment and storage medium, relate to the technical field of human-computer interaction, the method comprises: using a crawler tool, automatically collecting evaluation data on interface design on the network; based on text classification, the collected evaluation data is divided into positive evaluation data, negative evaluation data and neutral evaluation data; through a predefined syntax structure, identify the requirement evaluation and negative evaluation in the evaluation data, extract the key information in the evaluation data, and generate a visual report; based on the number and score of user comments, set the weighting coefficient of each function, and calculate the function priority; through A / B test, the feasibility of the design method is verified. The interface design method, device, electronic equipment and storage medium provided by the application not only improve the efficiency of user demand analysis, but also provide a scientific basis for interface design, and enhance the market competitiveness of the product.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of human-computer interaction, and particularly relates to an interface design method, an interface design device, an electronic device and a storage medium. BACKGROUND

[0002] In the interaction process between human and machine, there is a level, that is, interface. From the psychological meaning, the interface can be divided into two levels of feeling (vision, touch, hearing, etc.) and emotion. User interface design is an important part of screen products. Interface design is a complex engineering with the participation of different disciplines, and cognitive psychology, design, linguistics, etc. play an important role. The three principles of user interface design are: placing the interface under the control of the user; reducing the memory burden of the user; maintaining the consistency of the interface

[0003] However, interface design has not been valued in the long development of software. People who do interface design are also derogatorily called "art designers". In fact, software interface design is like industrial modeling design in industrial products, which is an important buying point of the product. A friendly and beautiful interface will bring people comfortable visual enjoyment, shorten the distance between people and the frame, and create a selling point for the business. Interface design is not just a simple art of painting, it needs to position the user, the use environment, the use method and design for the end user, and it is a purely scientific art design.

[0004] At present, the HMI design of the vehicle enterprise exists in the subjective experience and aesthetics of the designer as the guide; the actual needs of the user are ignored or difficult to capture; the designer needs to try multiple rounds of design and subjective evaluation in order to capture the needs of the user; and the standard for testing an interface is not the opinion of a project development group leader or the result of a project member vote, but the feelings of the end user. How to objectively and reasonably capture the subjective feelings of the user and reflect them in the interface design; provide a scientific design method for the designer; has become a key link in the development of interface design. SUMMARY

[0005] The purpose of the present application is to provide an interface design method, device, electronic device and storage medium, which not only improves the efficiency of user demand analysis, but also provides a scientific basis for interface design and enhances the market competitiveness of the product.

[0006] The present application provides the following solutions:

[0007] According to one aspect of the present application, an interface design method is provided, which comprises:

[0008] Using a crawler tool, automatically collecting evaluation data on interface design on the network;

[0009] Based on text classification, the collected evaluation data is divided into positive evaluation data, negative evaluation data and neutral evaluation data;

[0010] By predefining the syntax structure, the demand evaluation and negative evaluation in the evaluation data are identified, the key information in the evaluation data is extracted, and the visual report is generated;

[0011] Based on the number and score of user comments, the weighting coefficient of each function is set, and the function priority is calculated;

[0012] Through A / B test, the feasibility of the design process is verified.

[0013] Optionally, a crawler tool is used to automatically collect evaluation data on interface design on the network, including:

[0014] The crawler tool is used to automatically collect the evaluation data;

[0015] The collected batch of comments are preliminarily identified and corrected;

[0016] The crawler tool includes: BeautifulSoup or Selenium.

[0017] Optionally, based on text classification, the collected evaluation data is divided into positive evaluation data, negative evaluation data and neutral evaluation data, including:

[0018] The text to be classified is obtained;

[0019] The text to be classified is converted into a vector by using Word2Vec;

[0020] The SVM classifier is trained by using the vector;

[0021] The text to be classified is divided into positive evaluation data, negative evaluation data and neutral evaluation data by using the trained SVM classifier.

[0022] Optionally, by predefining the syntax structure, the demand evaluation and negative evaluation in the evaluation data are identified, the key information in the evaluation data is extracted, and the visual report is generated, including:

[0023] The syntax structure of demand and negative comments is defined, and the key information is extracted;

[0024] The extracted key information is visualized by using the Matplotlib tool.

[0025] Optionally, the syntax structure of demand and negative comments is defined, and the key information is extracted, including:

[0026] The sentiment analysis algorithm in the NLP technology is used to calculate the sentiment score of the comment.

[0027] A keyword list of demand and negative comments is defined, and keyword matching is used to assist in distinguishing.

[0028] The NLP tool is used for syntax analysis to identify the syntax structure in the comment.

[0029] The demand comment information and the negative comment information are processed respectively.

[0030] The convolutional neural network is used to extract features in the evaluation data.

[0031] Optionally, the sentiment analysis algorithm in the NLP technology is used to calculate the sentiment score of the comment, including:

[0032] The VADER, TextBlob or deep learning model is used to calculate the sentiment score of the comment.

[0033] Optionally, the NLP tool is used for syntax analysis to identify the syntax structure in the comment, including:

[0034] The spaCy tool is used to identify the syntax structure in the comment.

[0035] According to the two aspects of the present application, an interface design device is provided, which comprises:

[0036] A collection module is configured to automatically collect evaluation data of the interface design on the network by using a crawler tool;

[0037] A distinguishing module is configured to distinguish the collected evaluation data into positive evaluation data, negative evaluation data and neutral evaluation data based on text classification;

[0038] An extraction module is configured to identify demand evaluation and negative evaluation in the evaluation data by using a pre-defined syntax structure, extract key information in the evaluation data, and generate a visual report;

[0039] A calculation module is configured to set a weighting coefficient of each function based on the number and score of user comments, and calculate the function priority;

[0040] A verification module is configured to verify the feasibility of the design process by A / B testing.

[0041] According to the three aspects of the present application, an electronic device is provided, which comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus.

[0042] The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the interface design method.

[0043] According to the four aspects of the present application, a computer readable storage medium is provided, comprising: which stores a computer program executable by an electronic device, when the computer program runs on the electronic device, causes the electronic device to perform the steps of the interface design method.

[0044] Through the above scheme, the following beneficial technical effects are obtained:

[0045] The present application combines automatic data collection, natural language processing, information extraction and data visualization to provide an efficient intelligent cockpit interface optimization method. The technical points to be protected include data collection and processing methods, text classification and semantic analysis technology, data visualization implementation and function priority calculation method. These technical points not only improve the efficiency of user demand analysis, but also provide a scientific basis for interface design, and enhance the market competitiveness of the product. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 is a flowchart of the interface design method provided by one or more embodiments of the present application.

[0047] Figure 2 is a flowchart of the collection operation in the interface design method provided by one or more embodiments of the present application.

[0048] Figure 3 is a flowchart of the web page information crawling in the collection operation provided by one or more embodiments of the present application.

[0049] Figure 4 is a flowchart of the distinguishing operation in the interface design method provided by one or more embodiments of the present application.

[0050] Figure 5 is a flowchart of the extraction operation in the interface design method provided by one or more embodiments of the present application.

[0051] Figure 6 is a flowchart of the syntax structure definition in the extraction operation provided by one or more embodiments of the present application.

[0052] Figure 7 is a structural diagram of the interface design device provided by one or more embodiments of the present application.

[0053] Figure 8 is a structural diagram of the electronic device provided by one or more embodiments of the present application. DETAILED DESCRIPTION

[0054] The technical solutions of the present application will be clearly and completely described below with reference to the drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0055] Human–Machine Interaction (HMI): refers to the information exchange process between human and computer using certain dialogue language and certain interactive mode to complete the determined task. Many famous companies and academic institutions are researching human-computer interaction. In the history of computer development, people rarely pay attention to the usability of computers. Now, many computer users complain that computer manufacturers have not invested enough effort in making their products "user-friendly". On the contrary, these computer system developers complain that designing and manufacturing computers is a very complex job, and they have already spent most of their efforts on studying how to apply computers in new fields, and they simply do not have the extra energy to study how to improve the usability of computers. An important problem of human-computer interaction (HMI) is that different computer users have different use styles - their educational backgrounds, understanding methods, learning methods and skills are different, for example, the use habits of a left-hander and an ordinary person are completely different. In addition, cultural and national factors also need to be considered. Secondly, when studying and designing human-computer interaction, the rapid changes in user interface technology need to be considered, and new interactive technologies provided may not be suitable for previous studies. In addition, when users gradually master the new interface, they may have new requirements.

[0056] User Interface (UI): refers to the overall design of the human-computer interaction, operation logic, and interface aesthetics of software. Good UI design not only makes the software have personality and taste, but also makes the operation of the software comfortable, simple, free, and fully reflects the positioning and characteristics of the software. User Interface (UI) is the medium for interaction and information exchange between system and user, which realizes the conversion between internal form and human-acceptable form. User Interface is the software that mediates the interaction and communication between users and hardware design, and the purpose is to enable users to conveniently and efficiently operate hardware to achieve bidirectional interaction and complete the work expected to be completed by hardware. User Interface is defined widely, including human-computer interaction and graphical user interface, and there is a user interface in the field of information exchange between human and machine.

[0057] Figure 1 A flowchart of the interface design method provided by one or more embodiments of the present application is shown. Referring to Figure 1The interface design method comprises the following operation steps:

[0058] S11, using a crawler tool to automatically collect evaluation data on the interface design on the network.

[0059] S12, based on text classification, the collected evaluation data is divided into positive evaluation data, negative evaluation data and neutral evaluation data.

[0060] S13, through a predefined syntax structure, the requirement evaluation and negative evaluation in the evaluation data are identified, the key information in the evaluation data is extracted, and a visual report is generated.

[0061] S14, based on the number and score of user comments, setting the weighting coefficient of each function, and calculating the function priority.

[0062] S15, through A / B test, the feasibility of the design process is verified.

[0063] To ensure the authenticity and strong guidance of the obtained user information, a large number of evaluations of car owners or car enthusiasts need to be collected.

[0064] For data sample collection, if the information collection is simply performed by manpower, it is not only low in efficiency and cumbersome, but also the cost of collection is increased. A crawler tool (such as BeautifulSoup and Selenium) also known as a network robot can be used to help automatically collect data information.

[0065] In order to obtain data in the automobile vertical field, a large number of professional automobile evaluation websites need to be read and collected. In order to ensure the data quality and availability, some useless information such as web advertisements needs to be filtered out. The extractable web features such as “vehicle HMI” and “user experience” are determined, and whether the determined features meet the requirements is used as the basis for page classification to collect available web pages. In order to ensure that the sample number is large enough, the determined features are identified and calculated to query the sample number.

[0066] In professional automobile websites, there are also some irrelevant or unclear statements. Data preprocessing can preliminarily identify, correct and include sentence breaking and spelling correction for the collected batch of comments.

[0067] For the user comments collected by the crawler, sentence processing is performed, and noise words and non-Chinese characters are deleted.

[0068] After obtaining the preprocessed comments, the user comments need to be preliminarily classified, and the positive / negative / neutral evaluation categories are distinguished.

[0069] In processing a large number of reviews, the model needs to understand our textual information; word vectors can map users' words into vector form, so processing user reviews can be transformed from classifying language semantics into computation between vectors.

[0070] Therefore, word vectors and support vector machines (SVM) can be used to classify user comments and filter out constructive comments that are related to UI improvements.

[0071] Based on the training described above, we have obtained the categorized user reviews. Now, how can we enhance the readability of these reviews? By using predefined syntactic structures, we can extract key information from the comments and generate visual reports.

[0072] Define the grammatical structure of demand and negative comments, and extract their features.

[0073] The grammatical structure of demand and negative comments:

[0074] The syntactic structure of demand-based comments:

[0075] Request comments typically include user requests or suggestions for certain features or services, and their syntax includes:

[0076] Verb phrases: Verbs that express hope, suggestion, or request (such as "hope", "suggest", "need").

[0077] Noun phrases: indicating specific functions or characteristics (such as "better navigation system", "greater security").

[0078] Modifiers: Describe the attributes of the required functionality or service (such as "more efficient" or "more flexible").

[0079] Example:

[0080] "I hope the navigation system will be more accurate."

[0081] "It is recommended to add an autonomous driving function."

[0082] The grammatical structure of negative comments:

[0083] Negative comments typically express a user's dissatisfaction or negative emotions, and their grammatical structure includes:

[0084] Subject: The subject of the user or a specific function (such as "I", "system").

[0085] Verb phrases: Verbs that express negative emotions or problems (such as "hate", "break down", "slow").

[0086] Noun phrases: Nouns that describe specific functions or characteristics (such as "interface" or "performance").

[0087] Emotion words: Adjectives or adverbs that express negative emotions (e.g., "terrible," "very slow").

[0088] Examples:

[0089] "I hate that the system always crashes."

[0090] "The interface is very slow and the user experience is terrible."

[0091] (2). Differences in grammatical structure between demanding and negative comments

[0092] Verb usage:

[0093] Verbs used in demanding comments are usually words that express requests, such as "hope," "suggest," and "need."

[0094] Verbs used in negative comments are usually words that express dissatisfaction or problems, such as "hate," "crash," and "slow."

[0095] Emotional expression:

[0096] Demanding comments usually have constructive suggestions and neutral or positive emotions.

[0097] Negative comments usually have strong negative emotions and clearly negative emotion words.

[0098] Target object:

[0099] The target object of demanding comments is usually a function or service that needs to be improved.

[0100] The target object of negative comments is usually a function or feature that the user believes has problems.

[0101] Provide designers with low-code or zero-code foundations with charts and interactive interfaces to display analysis results, helping designers quickly understand market trends and user needs.

[0102] Matplotlib, as the earliest data visualization library in Python, provides rich plotting functions and flexible configuration options to meet most basic plotting needs; the extracted key information can be visualized.

[0103] In terms of basic graph drawing, we need to master several core concepts, including axes, legends, titles, and labels. Axes are the foundation of charts, used to show the distribution of data; legends are used to distinguish different series of data; titles and labels are used to explain the theme and content of the chart.

[0104] Statistical IVI module user complaints rate; the statistics we choose pie chart; because the pie chart is suitable for showing the proportion of various categories, suitable for relative comparison of data.

[0105] (1) Import library:

[0106] importmatplotlib.pyplot as plt

[0107] (2) Prepare data:

[0108] labels=['A','B','C','D']

[0109] sizes=[15,30,45,10]

[0110] (3) Draw pie chart:

[0111] plt.pie(sizes,labels=labels)

[0112] plt.show()

[0113] The above can get the pie chart of the percentage of each module complaint rate.

[0114] In the design of automotive HMI; To ensure driving safety and optimize passenger entertainment experience; We need to reduce the interface level and enhance the interface usability. According to the user's comments on the star level and the frequency of function use, we can calculate the optimization priority of each function; Provide a theoretical basis for interface level design.

[0115] Based on the number and rating of user comments, set the weighting coefficient of each function and calculate the function priority.

[0116] The priority algorithm needs to consider different weight settings and scoring mechanisms when dealing with demand and negative comments. The following will discuss the differences in weight settings, the basis for weight coefficient design, and the calculation method of scoring.

[0117] (1). The difference between weight settings

[0118] The weight settings of demand and negative comments have the following differences:

[0119] Demand comment weight:

[0120] The weight is mainly based on the urgency of user demand and the feasibility of the suggestion.

[0121] For example, a high-frequency demand may have a higher weight than a low-frequency demand.

[0122] Negative comment weight:

[0123] Weights are primarily based on the severity of the issue and the degree of negative impact on user experience.

[0124] For example, a problem that causes system crashes would have a higher weight than one that only affects the interface aesthetics.

[0125] (2). Design basis of weight coefficients

[0126] The design of weight coefficients can be based on the following aspects:

[0127] Frequency of occurrence: The number of times a comment appears or the proportion of total comments.

[0128] User impact: The degree of impact of the problem or demand mentioned in the comment on user experience, which can be measured through user feedback or survey results.

[0129] Solution difficulty: The complexity and resource requirements of solving the demand or problem.

[0130] Urgency: The urgency of solving the demand or problem.

[0131] (3). Calculation method of score

[0132] The calculation of the score can use a weighted scoring model to combine various weight coefficients to get the total score. The formula is as follows:

[0133] Score = ∑ (Weight coefficient × Related index)

[0134] Specifically, the following steps can be used:

[0135] 1. Determine the weight coefficient:

[0136] Frequency weight: Weight frequency

[0137] User impact weight: Weight impact

[0138] Solution difficulty weight: Weight difficulty

[0139] Urgency weight: Weight urgency

[0140] 2. Calculate the relevant index:

[0141] Frequency: Frequency

[0142] User impact: Impact

[0143] Solution difficulty: Difficulty

[0144] Urgency: Urgency

[0145] 3. Calculate score:

[0146] Score = (Weight frequency × frequency) + (Weight impact × impact) + (Weight difficulty × difficulty)

[0147] + (Weight urgency × urgency)

[0148] Priority calculation explanation:

[0149] def calculate_priority(comments):

[0150] weight_factors = [1.5, 1.25, 1, 1.25, 1.5] # weight factors

[0151] priorities = {}

[0152] for feature, ratings in comments.items():

[0153] priority = sum(r * weight_factors[i] for i, r in enumerate(ratings))

[0154] priorities[feature] = priority

[0155] return priorities

[0156] The above can derive the weight coefficient of each function, which is used as a reference for the design of the human-computer interaction interface level.

[0157] Compare the design after the designer adopts the above scheme with the original design; confirm the feasibility of the scheme.

[0158] To ensure the number of user samples and not affect the current HMI design that has been mass-produced; we use A / B testing method for pre-online testing.

[0159] In the same time dimension, randomly select car owners / car enthusiasts to form the same (similar) visitor group (target group) to randomly access A / B versions, collect user experience data and business data of each group, and the current design scheme is tested before being put into formal use.

[0160] 1. Determine the test target: Determine the test target, such as improving interface operation accuracy, reducing operation level, or improving user satisfaction, etc.

[0161] 2. Select test variables: Select variables that need to be tested, such as multimedia modules of HMI interface, navigation color matching, etc.

[0162] 3. Develop a test plan: Develop a test plan, including test time, test sample size, test scheme, etc.

[0163] 4. Create test versions: Select two stable software versions before and after the application of the scheme.

[0164] 5. Randomly assign users: Randomly assign website users into two or more groups, each group is shown different versions of test variables.

[0165] 6. Collect data: Collect user behavior and feedback data, we can use eye tracking, electroencephalogram, capture user gaze rate, etc. objective data; also can use user interviews, etc. to understand user's subjective evaluation.

[0166] 7. Analyze data: Analyze the data, compare the performance of each group, and determine whether the application scheme version is optimized.

[0167] 8. Application results: According to the test results, apply the most effective version, optimize HMI design and content.

[0168] Continuous improvement: A / B testing is a continuous process; we can regularly monitor and analyze user behavior and feedback data, optimize and improve different modules of HMI, and improve the effectiveness of products or services.

[0169] In the design of automotive HMI; through the method of automatic data collection, preprocessing, classification analysis and visualization, the interface design guided by the designer's subjective aesthetic can be optimized to a scientific interface design based on data and user-oriented, which not only simplifies the work flow of interface designers, shortens the development cycle, but also meets the user's voice; improves user experience and market competitiveness.

[0170] Figure 2 The flowchart of the operation collection in the interface design method provided by one or more embodiments of the present application is shown. Referring to Figure 2 , the crawler tool is used to automatically collect evaluation data on interface design on the network, including the following operation steps:

[0171] S21, using a crawler tool, automatically collecting evaluation data.

[0172] S22, preliminary identification and error correction of the collected batch of comments.

[0173] The process of automatically collecting comment information on a webpage through a crawler tool roughly includes the following operation steps:

[0174] (1) Read / access the webpage

[0175] Import BeautifulSoup and a parser; import the requests library to read webpage content from the network:

[0176] from bs4 import BeautifulSoup

[0177] import requests

[0178] from selenium import webdriver

[0179] url = 'https: / / example.com'

[0180] response = requests.get(url)

[0181] (2) Collect website

[0182] Use BeautifulSoup and the parser to collect website content:

[0183] soup = BeautifulSoup(response.text, 'html.parser')

[0184] (3) Webpage feature extraction

[0185] Use selectors to extract data for defined webpage features; for example, use the.find() or.find_all() methods to find the tag "car HMI" and use the.get_text() method to get the text inside the tag:

[0186] divs_with_class = soup.find_all('div', class_='some-class')

[0187] (4) Identify feature calculation parameters

[0188] Use the count function to count the feature data extracted by the selector:

[0189] def popular_words(text, words):

[0190] lower_count = text.lower().split().count

[0191] return{word:lower_count(word)forword inwords}

[0192] The preprocessing process described above is as follows:

[0193] (1) Word segmentation

[0194] Word segmentation, or sentence breaking, requires reasonable sentence segmentation to accurately capture user semantics. For example, "I like the car's voice assistant" can be segmented into "I / like / car / voice / assistant / "; "I like / car voice assistant / ", etc. During word segmentation, we should aim for large-granularity segmentation. That is, the more characters in a word, the more precise its meaning. With the same number of characters, a smaller total word count indicates fewer semantic units, resulting in a higher weight for each individual semantic unit and thus higher accuracy.

[0195] Use `jieba.Tokenizer(dictionary=DEFAULT_DICT)` to create a custom tokenizer, which can be used to work with different dictionaries simultaneously. `jieba.dt` is the default tokenizer, and all global tokenization-related functions are mappings to this tokenizer.

[0196] seg_list = jieba.cut("I like the car's voice system", cut_all = True)

[0197] print("Full Mode:"+" / ".join(seg_list))

[0198] (2) Cleaning

[0199] Data cleaning helps us remove useless data and reduce the size of training data, including removing punctuation, stop words, and rare words. Since these steps do not add any extra information to the text data, there is no need to retain them.

[0200] importre

[0201] defpreprocess_text(text):

[0202] text = re.sub(r'[^\u4e00-\u9fa5]',",text) # Only Chinese characters are retained

[0203] sentences = text.split('.')

[0204] return sentences

[0205] Remove stop words / rare words:

[0206] A list of stop words can be created as a personal stop word list or a predefined library can be used.

[0207] train['tweet']=train['tweet'].apply(lambda sen:"".join(x forx insen.split()ifx not in stop)

[0208] (3) Standardization

[0209] Standardization means spelling correction, the main task is to map the wrong word to the correct word. For example, "cheging" is corrected to "che machine". At the same time, compare the wrong word with the words in the dictionary, find the most matching word.

[0210] Delete characters from the wrong word one by one to generate all possible subsequences.

[0211] Compare these subsequences with the words in the dictionary to find the most matching word.

[0212] from textblob importTextBlob

[0213] train['tweet'][:5].apply(lambda x:str(TextBlob(x).correct()))

[0214] Figure 3 The flowchart of the web page information crawling in the collection operation provided by one or more embodiments of the application is shown. Referring to Figure 3 , a crawler tool is used to help automatically collect data information, including the following operation steps:

[0215] S31, reading / accessing a web page.

[0216] S32, collecting review data from the website.

[0217] S33, extracting features collected from the web page.

[0218] S34, storing the extracted web page collection features into a web page database.

[0219] S35, obtaining page features.

[0220] S36, classifying the page according to the extracted web page collection features and the page features.

[0221] S37, specifying the page features.

[0222] S38, identify feature calculation parameters.

[0223] S39, count the number of queries.

[0224] In various embodiments of the present application, the automatic collection process of the comment data on the webpage by the crawler tool mainly consists of the following operation steps.

[0225] First, the target webpage containing interface comment data is read and accessed.

[0226] After completing the reading / accessing of the target webpage, the collection of comment data is performed on the target website according to the read information.

[0227] After the collection of comment data is completed, the webpage collection features are extracted.

[0228] It should be understood that different webpages have different features for the collection operation of the comment data. For example, different websites have their corresponding types. Some websites are car sales websites, from which real-time transaction data of vehicles can be easily captured. Some other websites are car review websites, on which a large number of car enthusiasts have published a large amount of use experience of specific vehicle models. On the car review website, the public's preference data for different vehicle models can be easily obtained.

[0229] Not only the type of the website, but also some other features closely related to the data collection operation can be classified into the category of webpage collection features. For example, for a webpage implemented by using JSP technology, the response speed is generally faster than that of a webpage implemented by using PHP and ASP technologies. Therefore, when collecting comment data on a webpage implemented by using JSP, the waiting time for collection can be set relatively short. For data collection of a webpage implemented by using PHP and ASP technologies, the setting value of the waiting time should be relatively large.

[0230] After the webpage collection features are extracted, the page is classified according to the read webpage data. The above classification aims to divide the comment data in the webpage into positive comment data, negative comment data, and neutral comment data according to the emotional tendency of the webpage text.

[0231] The above classification operation of the page utilizes the page features of the classified page.

[0232] It should be understood that some page features of the target webpage can be obtained in the process of collecting the comment data. These page features can include the specific location of the target comment area on the webpage, the location of the dynamic comment area, etc.

[0233] These page features can be obtained through experience summary of multiple data collection actions, that is, statistics of historical collection data. For example, for a target webpage, data collection is performed in different time periods, and it is found that the text contained in some specific areas of the webpage is dynamically changed. According to the dynamic change of such historical data, it can be determined that the area containing the dynamically changed text data is the dynamic comment area in the webpage, that is, the main target area of the comment data collection action.

[0234] Similar to the acquisition of the dynamic comment area described above, the extraction of the page features can be considered as the extraction action of the page features.

[0235] After the classification of the positive comment data, the negative comment data, and the neutral comment data is completed, the page features can be specified, and the feature data extracted by the selector can be counted according to the specified page features, that is, the feature calculation parameters are identified.

[0236] Further, after the feature calculation parameters are identified, the number of webpage queries can be further counted according to the identified feature calculation parameters.

[0237] Figure 4 A flowchart of the interface design method provided by one or more embodiments of the present application is shown. Referring to Figure 4 , based on text classification, the collected evaluation data is divided into positive evaluation data, negative evaluation data and neutral evaluation data, including the following operation steps:

[0238] S41, obtaining the text to be classified.

[0239] S42, converting the text to be classified into a vector using Word2Vec.

[0240] S43, training an SVM classifier using the vector.

[0241] S44, using the trained SVM classifier to divide the text to be classified into positive evaluation data, negative evaluation data and neutral evaluation data.

[0242] In the embodiment shown by Figure 4 , the division of the positive evaluation data, the negative evaluation data, and the neutral evaluation data is completed by the pre-constructed SVM classifier.

[0243] Support Vector Machine (SVM) is a generalized linear classifier for binary classification of data in a supervised learning manner, and its decision boundary is a maximum-margin hyperplane solved for learning samples.

[0244] SVM uses a hinge loss function to calculate the empirical risk and adds a regularization term to the solution system to optimize the structural risk, and is a classifier with sparsity and robustness. SVM can perform nonlinear classification through a kernel method, and is one of the common kernel learning methods.

[0245] The classification target of SVM is the comment data collected from the website. That is, the object of SVM classification is the text obtained from the network.

[0246] Before classifying the actual classification target, that is, the text, the Word2Vec tool needs to be used to convert the text into a vector.

[0247] After the above vector conversion is completed, the obtained vector is marked with the corresponding properties. That is, according to the original attributes of the text data, the converted vector is marked as positive evaluation data, negative evaluation data and neutral evaluation data.

[0248] After the labeling of the training data is completed, the above labeled data is input into the SVM classifier to complete the training of the SVM classification.

[0249] After the above classifier training and parameter tuning, the SVM classifier capable of actually classifying the original data is obtained. At this time, the trained SVM classifier is applied to the classification of the original comment data, and the classification result of whether the original comment data belongs to positive comment data, negative comment data or neutral comment data can be obtained.

[0250] Figure 5 A flowchart of an extraction operation in an interface design method provided by one or more embodiments of the present application is shown. Referring to Figure 5 , by predefining a syntax structure, the requirement evaluation and negative evaluation in the evaluation data are identified, the key information in the evaluation data is extracted, and a visual report is generated, including:

[0251] S51, define the syntax structure of the demand and negative comments, and extract the key information.

[0252] S52, use the Matplotlib tool to visualize the extracted key information.

[0253] In various embodiments of the present application, the extraction operation of the key information mainly includes two separate sub-operations. The two sub-operations are: feature extraction by syntax structure definition, and visualization operation of the extracted key information.

[0254] The definition of the syntax structure is respectively defined for the characteristics of the demand comments and the negative comments.

[0255] For example, demand comments often contain some verbs of hope, suggestion or request. For example, such verbs can be: hope, suggestion, need. Therefore, in the process of identifying demand comments, whether the above-mentioned verbs are contained in the comments can be identified first. If the verbs with the above-mentioned characteristics appear, the above-mentioned comments can be determined as demand comments.

[0256] For example, negative comments often contain some verbs expressing negative emotions or problems. These verbs can be: hate, collapse, slow. Therefore, in the process of identifying negative comments, these verbs expressing negative emotions or problems can be identified. The identified above-mentioned verbs can determine the above-mentioned comments as negative comments.

[0257] After identifying the demand comments and the negative comments, the key information can be extracted from the text data of the above-mentioned two kinds of comments respectively.

[0258] In various embodiments of the present application, the tool used for the above-mentioned key information extraction is a convolutional neural network. The convolutional neural network is a deep learning model for image and text data, which mainly consists of convolutional layer, pooling layer and full connection layer. The convolutional layer is used to extract the features in the text data, the pooling layer is used to reduce dimension and reduce calculation, and the full connection layer is used for classification prediction. After we design the key information definition, the user evaluation related feature extraction is mainly completed by the convolutional layer.

[0259] After the above-mentioned key information extraction is completed, the visualization operation of the above-mentioned key information can be performed, and various visualization reports can be generated, so that the extracted features are more intuitive to the user.

[0260] Figure 6 The flowchart of the syntax structure definition in the extraction operation provided by one or more embodiments of the present application is shown. Referring to Figure 6, define the grammatical structures of demand and negative reviews, and perform feature extraction, including the following steps:

[0261] S61, use sentiment analysis algorithms in NLP techniques to calculate sentiment scores for reviews.

[0262] S62, define a keyword list for demand and negative reviews, and use keyword matching to assist in differentiation.

[0263] S63, use NLP tools for syntactic analysis to identify the syntactic structure of the review.

[0264] S64, process demand and negative review information separately.

[0265] S65, use convolutional neural networks to extract features from evaluation data.

[0266] (1). Grammatical structure of demand and negative reviews

[0267] Grammatical structure of demand reviews:

[0268] Demand reviews usually include user requests or suggestions for certain functions or services, and their grammatical structure includes:

[0269] Verb phrase: verbs expressing hope, suggestion or request (such as "hope", "suggest", "need").

[0270] Noun phrase: specific functions or features (such as "better navigation system", "higher security").

[0271] Adjective: describes the attributes of the desired function or service (such as "more efficient", "more flexible").

[0272] Examples:

[0273] "I hope the navigation system is more accurate."

[0274] "Suggest adding automatic driving function."

[0275] Grammatical structure of negative reviews:

[0276] Negative reviews usually express user dissatisfaction or negative emotions, and their grammatical structure includes:

[0277] Subject: subject of the user or specific function (such as "I", "system").

[0278] Verb phrase: verbs expressing negative emotions or problems (such as "hate", "crash", "slow").

[0279] Noun Phrase: A noun phrase describes a specific function or characteristic (e.g., "interface," "performance").

[0280] Sentiment Word: An adjective or adverb that expresses negative emotion (e.g., "terrible," "very slow").

[0281] Example:

[0282] "I hate that the system always crashes."

[0283] "The interface is very slow and the user experience is terrible."

[0284] (2). Differences in grammatical structure between demanding and negative reviews

[0285] Verb Usage:

[0286] Demanding reviews often use verbs that express requests, such as "hope," "suggest," and "need."

[0287] Negative reviews often use verbs that express dissatisfaction or problems, such as "hate," "crash," and "slow."

[0288] Emotional Expression:

[0289] Demanding reviews usually contain constructive suggestions and have neutral or positive emotions.

[0290] Negative reviews usually contain strong negative emotions and have clearly negative sentiment words.

[0291] Target Object:

[0292] Demanding reviews usually target functions or services that the user hopes to improve.

[0293] Negative reviews usually target functions or features that the user believes have problems.

[0294] (3). Methods to distinguish between demanding and negative reviews

[0295] 1. Sentiment Analysis

[0296] Use sentiment analysis algorithms in NLP technology (such as VADER, TextBlob, or deep learning models) to calculate sentiment scores for reviews. According to the scores, distinguish between positive, neutral and negative reviews.

[0297] 2. Keyword Detection

[0298] Define a list of keywords for demanding and negative reviews, and use keyword matching to assist in distinguishing.

[0299] 3. Syntax Parsing

[0300] Use NLP tools like spaCy for syntax parsing to identify the syntactic structure in the comments. Determine the comment type based on the syntactic structure.

[0301] Sentiment analysis can be used as the most common preliminary classification method to roughly classify comments into positive, neutral, and negative categories. For comments with scores close to positive and neutral or neutral and negative, we can choose to customize keywords for secondary classification to improve classification granularity. Syntax parsing can further verify the comments classified by the above two methods to improve classification accuracy.

[0302] 4. Processing of demand and negative comment information

[0303] Demand comment processing:

[0304] Priority assessment:

[0305] According to the importance and frequency of the functions or features mentioned in the comments, assess the priority.

[0306] Improvement plan development:

[0307] For high-priority demand comments, develop appropriate product improvement plans.

[0308] User feedback loop:

[0309] Feedback the improvement plan and progress to the user and establish a good user communication channel.

[0310] Negative comment processing:

[0311] Problem positioning:

[0312] Analyze the problems mentioned in negative comments to determine the specific manifestations and frequency of the problems.

[0313] Root cause analysis:

[0314] Through further data analysis and user surveys, find the root cause of the problem.

[0315] Solution measures:

[0316] Develop solution measures for the root cause to optimize the product or service.

[0317] User satisfaction improvement:

[0318] After solving the problem, give positive feedback to the user to improve user satisfaction.

[0319] In addition, further feature extraction in the evaluation data is performed using a convolutional neural network.

[0320] Convolutional Neural Networks (CNNs) are deep learning models that have achieved tremendous success in the field of computer vision. Their design is inspired by the biological visual system, aiming to mimic how humans process visual information. In the past few years, CNNs have made significant progress in image recognition, object detection, image generation, and many other areas, becoming an important part of computer vision and deep learning research.

[0321] A convolution kernel is a weight matrix that generates a feature map by sliding summation with the input text data, thereby extracting the specified features we need.

[0322] Mathematical model formula:

[0323]

[0324] Among them, y ij It is the value of the i-th row and j-th column of the output feature map, x ik It is the value in the i-th row and k-th column of the input text data, w kj It is the value of the convolution kernel in the k-th row and j-th column, b j is the bias term, and K is the width of the convolution kernel.

[0325] import spacy

[0326] nlp=spacy.load('zh_core_web_sm')

[0327] defextract_information(sentence):

[0328] doc = nlp(sentence)

[0329] For token in the document:

[0330] iftoken.dep_=='ROOT':

[0331] print(f'Subject:{token.text},Predicate:{token.head.text}')

[0332] The above methods can extract specified features.

[0333] like Figure 7 As shown, the present invention also provides an interface design device, comprising:

[0334] The collection module 71 is used to automatically collect evaluation data on interface design from the Internet using web crawling tools.

[0335] The distinguishing module 72 is configured to distinguish the collected evaluation data into positive evaluation data, negative evaluation data and neutral evaluation data based on text classification.

[0336] The extraction module 73 is configured to identify the demand evaluation and the negative evaluation in the evaluation data through a predefined syntax structure, extract the key information in the evaluation data, and generate a visual report.

[0337] The calculation module 74 is configured to set a weighting coefficient of each function based on the number and score of user comments, and calculate the function priority.

[0338] The verification module 75 is configured to verify the feasibility of the design process through A / B testing.

[0339] It is worth noting that, although only some basic function modules are disclosed in the embodiments of the present application, it does not mean that the composition of the system is limited to the above-mentioned basic function modules. On the contrary, the meaning expressed by the embodiments is that on the basis of the above-mentioned basic function modules, one or more function modules can be added by those skilled in the art in combination with the prior art to form infinite embodiments or technical solutions. That is, the system is open rather than closed, and the protection scope of the claims of the present application cannot be limited to the disclosed basic function modules because only individual basic function modules are disclosed in the embodiments. At the same time, in order to facilitate description, the above device is described as various units and modules. Of course, the functions of the units and modules can be realized in the same software and / or hardware in the implementation of the present application.

[0340] As shown in Figure 8 , the present application also provides an electronic device, comprising: a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus; the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the interface design method.

[0341] Figure 8 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. As shown in the structure Figure 8 , the electronic device provided in the embodiment of the present application comprises one or more processors 810 and a memory 820; the processor 810 in the electronic device can be one or more, Figure 8 , taking a processor 810 as an example; the memory 820 is configured to store one or more programs; the one or more programs are executed by the one or more processors 810, so that the one or more processors 810 implement the interface design method according to any one of the embodiments of the present application.

[0342] The electronic device can further include an input device 830 and an output device 840.

[0343] The processor 810, the memory 820, the input device 830 and the output device 840 in the electronic device can be connected by a bus or other means, Figure 8 The bus connection is taken as an example.

[0344] The memory 820 in the electronic device can be used to store one or more programs as a kind of computer readable storage medium, the program can be a software program, a computer executable program and a module, such as the program instruction / module corresponding to the full link test method of the voice vehicle control provided in the embodiment of the application. The processor 810 executes the software program, instruction and module stored in the memory 820, thereby performing various function applications and data processing of the electronic device, that is, realizing the interface design method in the above method embodiment.

[0345] The memory 820 can include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; the data storage area can store data created according to the use of the electronic device, etc. In addition, the memory 820 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state memory device. In some examples, the memory 820 can further include a memory remotely arranged relative to the processor 810, which can be connected to the device through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0346] The input device 830 can be used to receive input digital or character information, and generate key signal input related to user settings and function control of the electronic device. The output device 840 can include a display device such as a display screen.

[0347] The application further provides a computer readable storage medium, which stores a computer program executable by an electronic device, when the computer program runs on the electronic device, so that the electronic device executes the steps of the interface design method.

[0348] In particular, the computer storage medium of the embodiments of the present application can adopt any combination of one or more computer readable mediums. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples (non-exhaustive list) of the computer readable storage medium include: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present embodiment, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus.

[0349] The present application also provides a vehicle provided with the interface design device described above.

[0350] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. An interface design method, characterized in that, The interface design method includes: Use web crawlers to automatically collect evaluation data on interface design from the internet; Based on text classification, the collected evaluation data is divided into positive evaluation data, negative evaluation data, and neutral evaluation data. By using predefined syntactic structures, the system identifies demand-driven and negative evaluations in the evaluation data, extracts key information from the evaluation data, and generates a visual report. Based on the number and ratings of user reviews, weighting coefficients are set for each function, and function priorities are calculated. A / B testing is used to verify the feasibility of the design process; When handling demand and negativity evaluations, priority algorithms need to consider different weight settings and scoring mechanisms; The weighting of demand-based evaluation and negative evaluation differs in the following ways: Demand evaluation weight: The weighting is primarily based on the urgency of the user's needs and the feasibility of the suggestions; Negative evaluation weight: The weighting is primarily based on the severity of the problem and the degree of negative impact on the user experience; Web crawlers were used to automatically collect online evaluation data on interface design, including: Web crawling tools were used to automatically collect evaluation data; The collected batch of evaluations are initially identified and corrected. Among them, web scraping tools include: BeautifulSoup or Selenium; Web crawlers are used to automatically collect data, including the following steps: Read / access web pages; Evaluation data was collected from the website; Extract features from webpage scraping; The extracted webpage collection features are stored in the webpage database; Obtain page features; Based on the extracted webpage collection features and page features, the pages are classified. Specify page characteristics; Parameters for feature recognition calculation; Count the number of queries.

2. The method according to claim 1, characterized in that, Based on text classification, the collected evaluation data is divided into positive evaluation data, negative evaluation data, and neutral evaluation data, including: Get the text that needs to be categorized; Using Word2Vec, the text that needs to be classified is converted into vectors; Train an SVM classifier using vectors; Using a trained SVM classifier, the text to be classified is divided into positive evaluation data, negative evaluation data, and neutral evaluation data.

3. The method according to claim 1, characterized in that, By using predefined syntactic structures, the system identifies demand-driven and negative evaluations in the evaluation data, extracts key information from the data, and generates visual reports, including: Define the grammatical structure of demand and negative evaluation, and extract key information; The extracted key information can be visualized using the Matplotlib tool.

4. The method according to claim 3, characterized in that, Define the grammatical structure for demand and negative evaluations, and extract key information, including: The sentiment analysis algorithm in NLP technology is used to calculate the sentiment score of the evaluation. Define a keyword list for demand and negative evaluations, and use keyword matching to help differentiate them; Use NLP tools to perform syntax parsing and identify syntactic structures in the evaluation; Demand-driven evaluation information and negative evaluation information are processed separately; Convolutional neural networks are used to extract features from the evaluation data.

5. The method according to claim 4, characterized in that, Using sentiment analysis algorithms from NLP technology, sentiment scores are calculated for evaluations, including: Use VADER, TextBlob, or deep learning models to calculate sentiment scores for ratings.

6. The method according to claim 4, characterized in that, Use NLP tools for syntax parsing to identify grammatical structures in the evaluation, including: Use the spaCy tool to identify the syntactic structure in the evaluation.

7. An interface design device, characterized in that, The interface design device includes: The collection module is used to automatically collect evaluation data on interface design from the internet using web crawling tools. The differentiation module is used to classify collected evaluation data into positive evaluation data, negative evaluation data, and neutral evaluation data based on text classification. The extraction module is used to identify demand-oriented and negative evaluations in the evaluation data through predefined syntactic structures, extract key information from the evaluation data, and generate a visual report. The calculation module is used to set weighting coefficients for each function and calculate function priority based on the number and rating of user reviews; The verification module is used to verify the feasibility of the design process through A / B testing; When handling demand and negativity evaluations, priority algorithms need to consider different weight settings and scoring mechanisms; The weighting of demand-based evaluation and negative evaluation differs in the following ways: Demand evaluation weight: The weighting is primarily based on the urgency of the user's needs and the feasibility of the suggestions; Negative evaluation weight: The weighting is primarily based on the severity of the problem and the degree of negative impact on the user experience; Web crawlers were used to automatically collect online evaluation data on interface design, including: Web crawling tools were used to automatically collect evaluation data; The collected batch of evaluations are initially identified and corrected. Among them, web scraping tools include: BeautifulSoup or Selenium; Web crawlers are used to automatically collect data, including the following steps: Read / access web pages; Evaluation data was collected from the website; Extract features from webpage scraping; The extracted webpage collection features are stored in the webpage database; Obtain page features; Based on the extracted webpage collection features and page features, the pages are classified. Specify page characteristics; Parameters for feature recognition calculation; Count the number of queries.

8. An electronic device, characterized in that, include: The system includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, communication interface, and memory communicate with each other via the communication bus; the memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, It stores a computer program executable by an electronic device, which, when run on the electronic device, causes the electronic device to perform the steps of the method according to any one of claims 1 to 6.

Images