Comment classification method, device, equipment, storage medium and program product

Abstract

The present disclosure provides a comment classification method, device, equipment, storage medium and program product, which belong to the technical field of computers. The method comprises: performing feature extraction processing on a plurality of obtained comment data respectively to obtain a feature vector corresponding to each comment data; performing theme clustering processing on the feature vectors corresponding to the plurality of comment data to obtain theme information corresponding to each comment data; performing frequency statistical processing on the theme data corresponding to the plurality of comment data to obtain the occurrence frequency of each theme information; and determining the demand intensity classification corresponding to each theme information based on the theme information corresponding to each comment data, the occurrence frequency of each theme information and a classification model. By using the present disclosure, human and material resources are saved, and the relatively intense after-sales problems can be processed in time and quickly, thereby improving the user experience.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to a comment classification method, apparatus, device, storage medium, and program product. Background Technology

[0002] With increasing competition in the automotive market, after-sales service has become a key factor influencing customer satisfaction and brand loyalty. High-quality after-sales service not only improves customer satisfaction but also enhances a brand's market competitiveness. Among the data collected in after-sales service, customer reviews of vehicle usage are an important reference for reflecting user experience. For example, reviews might include complaints about weak dynamism, unclear photos taken by the vehicle's main unit camera, or malfunctioning brakes.

[0003] Typically, staff will manually filter user reviews and select those that require priority after-sales processing to improve the vehicle user experience.

[0004] However, the above methods require manual screening of comment data, which is a waste of manpower and resources. Furthermore, the accuracy of manual screening is not high, and it cannot respond to user needs in a timely manner, thereby reducing the user experience. Summary of the Invention

[0005] This disclosure provides a comment classification method that can solve the technical problems existing in related technologies. The technical solution is as follows:

[0006] Firstly, a comment classification method is provided, the method comprising:

[0007] The acquired multiple comment data are processed by feature extraction to obtain the feature vector corresponding to each comment data.

[0008] Perform topic clustering on the feature vectors corresponding to the multiple comment data to obtain topic information corresponding to each comment data;

[0009] Frequency statistics are performed on the topic data corresponding to the multiple comment data to obtain the frequency of occurrence of each topic information;

[0010] Based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, and the classification model, the demand intensity classification corresponding to each topic information is determined, wherein the demand intensity classification corresponding to the topic information is used to indicate the after-sales processing priority of the topic information.

[0011] In one possible implementation, the step of performing feature extraction processing on the acquired multiple comment data to obtain a feature vector corresponding to each comment data includes:

[0012] Based on the multiple comment data and the topic feature extraction model, the topic feature vector corresponding to each comment data is determined;

[0013] Based on the multiple comment data and the semantic feature extraction model, the semantic feature vector corresponding to each comment data is determined;

[0014] Based on the topic feature vector and the semantic feature vector corresponding to each comment data, the feature vector corresponding to each comment data is determined.

[0015] In one possible implementation, obtaining the feature vector corresponding to each comment data based on the topic feature vector and the semantic feature vector corresponding to each comment data includes:

[0016] The topic feature vector and the semantic feature vector corresponding to each comment data are concatenated to obtain the feature vector corresponding to each comment data.

[0017] In one possible implementation, the method further includes:

[0018] Based on the multiple comment data and the sentiment intensity model, a sentiment intensity value corresponding to each comment data is determined, wherein the sentiment intensity value corresponding to the comment data is used to indicate the degree of sentiment intensity and sentiment polarity of the comment data;

[0019] The step of determining the demand intensity classification corresponding to each topic information based on the multiple comment data, the topic information corresponding to each comment data, the frequency of occurrence of each topic information, and a classification model includes:

[0020] Based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, the emotional intensity value corresponding to each comment data, and the classification model, the demand intensity classification corresponding to each topic information is determined.

[0021] In one possible implementation, before performing feature extraction processing on the acquired multiple comment data to obtain the feature vector corresponding to each comment data, the method further includes:

[0022] Obtain multiple initial comment data;

[0023] The initial comment data is preprocessed to obtain the comment data.

[0024] In one possible implementation, the text preprocessing includes at least one of the following: text deduplication, compression and word removal, short sentence deletion, pure symbol deletion, pure English deletion, text segmentation, and stop word removal.

[0025] Secondly, a comment classification device is provided, the device comprising:

[0026] The feature extraction module is used to perform feature extraction processing on the acquired multiple comment data respectively, and obtain the feature vector corresponding to each comment data;

[0027] The clustering module is used to perform topic clustering processing on the feature vectors corresponding to the multiple comment data to obtain topic information corresponding to each comment data.

[0028] The statistics module is used to perform frequency statistics processing on the topic data corresponding to the multiple comment data to obtain the frequency of occurrence of each topic information;

[0029] The classification module is used to determine the demand intensity classification corresponding to each topic information based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, and the classification model. The demand intensity classification corresponding to the topic information is used to indicate the after-sales processing priority of the topic information.

[0030] In one possible implementation, the feature extraction module is used to:

[0031] Based on the multiple comment data and the topic feature extraction model, the topic feature vector corresponding to each comment data is determined;

[0032] Based on the multiple comment data and the semantic feature extraction model, the semantic feature vector corresponding to each comment data is determined;

[0033] Based on the topic feature vector and the semantic feature vector corresponding to each comment data, the feature vector corresponding to each comment data is determined.

[0034] In one possible implementation, the feature extraction module is used to:

[0035] The topic feature vector and the semantic feature vector corresponding to each comment data are concatenated to obtain the feature vector corresponding to each comment data.

[0036] In one possible implementation, the apparatus further includes a determining module for:

[0037] Based on the multiple comment data and the sentiment intensity model, a sentiment intensity value corresponding to each comment data is determined, wherein the sentiment intensity value corresponding to the comment data is used to indicate the degree of sentiment intensity and sentiment polarity of the comment data;

[0038] The classification module is used for:

[0039] Based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, the emotional intensity value corresponding to each comment data, and the classification model, the demand intensity classification corresponding to each topic information is determined.

[0040] In one possible implementation, the apparatus further includes a preprocessing module for:

[0041] Obtain multiple initial comment data;

[0042] The initial comment data is preprocessed to obtain the comment data.

[0043] In one possible implementation, the text preprocessing includes at least one of the following: text deduplication, compression and word removal, short sentence deletion, pure symbol deletion, pure English deletion, text segmentation, and stop word removal.

[0044] Thirdly, a computer device is provided, the computer device including a processor and a memory, the memory storing at least one instruction, the instruction being loaded and executed by the processor to implement the operation performed by the comment classification method.

[0045] Fourthly, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, the instruction being loaded and executed by a processor to implement the operations performed by the comment classification method.

[0046] Fifthly, a computer program product is provided, the computer program product including at least one instruction, the at least one instruction being loaded and executed by a processor to implement the operations performed by the comment classification method.

[0047] The beneficial effects of the technical solution provided in this disclosure are as follows: The solution mentioned in this disclosure can determine the topic information corresponding to each comment data through feature extraction and topic clustering. Furthermore, it can determine the demand intensity category corresponding to each topic information through a classification model. This allows staff to prioritize each topic information based on its demand intensity category and then handle the corresponding after-sales issues accordingly. This eliminates the need for manual screening of comment data, saving human and material resources. Moreover, staff can directly handle after-sales issues corresponding to each topic information based on demand intensity category, enabling timely and rapid handling of high-demand after-sales issues, thereby improving the user experience. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart illustrating an evaluation and classification method provided in an embodiment of this disclosure;

[0050] Figure 2 This is a flowchart illustrating an evaluation and classification method provided in an embodiment of this disclosure;

[0051] Figure 3 This is a schematic diagram of the structure of an evaluation and classification device provided in an embodiment of this disclosure;

[0052] Figure 4 This is a structural block diagram of a terminal provided in an embodiment of this disclosure;

[0053] Figure 5 This is a structural block diagram of a server provided in an embodiment of this disclosure. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this disclosure clearer, the embodiments of this disclosure will be described in further detail below with reference to the accompanying drawings.

[0055] This disclosure provides a comment classification method, which can be implemented by a computer device. The computer device can be a terminal or server, and the terminal can be a desktop computer, laptop computer, tablet computer, mobile phone, etc.

[0056] Computer equipment may include processors, memory, communication components, etc.

[0057] The processor can be a central processing unit (CPU). The processor can be used to read instructions and process data. For example, it can perform feature extraction on multiple sets of review data to obtain the feature vector corresponding to each review data; perform topic clustering on the feature vectors corresponding to multiple review data to obtain the topic information corresponding to each review data; perform frequency statistics on the topic data corresponding to multiple review data to obtain the frequency of occurrence of each topic information; determine the demand intensity classification corresponding to each topic information, and so on.

[0058] The memory can be various types of volatile or non-volatile memory, such as solid-state drives (SSDs) and dynamic random access memory (DRAM). Memory can be used for data storage, for example, storing multiple sets of acquired comment data, storing the feature vector corresponding to each set of comment data, storing the topic information corresponding to each set of comment data, storing the demand intensity classification corresponding to each determined topic information, and so on.

[0059] Communication components can be wired network connectors, wireless fidelity (WiFi) modules, Bluetooth modules, cellular communication modules, etc. These components can be used to transmit data with other devices.

[0060] The comment classification method provided in this disclosure can be applied to vehicles. A host computer is installed in the vehicle, and users can comment on the vehicle's usage through a comment application installed on the host computer. The host computer obtains the comment data (or initial comment data) entered by the user and sends it to a backend server. The backend server classifies the multiple comment data based on the obtained data, obtaining a demand intensity classification for different topics corresponding to the multiple comment data. Then, based on this demand intensity classification, it provides a priority ranking of each topic and corresponding after-sales improvement methods, thereby achieving timely vehicle optimization and improving the user's driving experience.

[0061] Of course, the comment classification method can be applied not only to vehicles, but also to other reasonable comment scenarios, such as classifying comments on the use of applications, mechanical equipment, electrical appliances, etc. This disclosure does not limit this.

[0062] The following section provides a detailed description of the comment classification method provided in this embodiment, using a vehicle application scenario as an example.

[0063] Figure 1 and Figure 2 This is a flowchart illustrating a comment classification method provided in an embodiment of this disclosure. See also... Figure 1 and Figure 2 This embodiment includes:

[0064] 101. Perform feature extraction processing on the multiple obtained comment data to obtain the feature vector corresponding to each comment data.

[0065] During implementation, when preset conditions are met, multiple comment data can be obtained. This comment data represents users' evaluations of the vehicle's performance, such as weak engine power, a comprehensive music library in the car, and difficult-to-control windows. These comment data can reflect users' dissatisfaction or satisfaction after using the vehicle.

[0066] After obtaining multiple comment data, features can be extracted from each of these comment data to obtain a feature vector corresponding to each comment data. This feature vector is the semantic representation of the corresponding comment data.

[0067] In one possible implementation, the aforementioned preset conditions may be that the time elapsed since the last comment categorization reaches a preset period, and / or that the current time reaches the preset comment categorization time for the target project. The preset period may be one month, 20 days, etc., and the preset comment categorization time for the target project may be one month, two months, three months, etc., after the target project has started. This disclosure does not impose specific limitations on these embodiments.

[0068] Understandably, when the preset condition "the time elapsed since the last comment category reaches a preset period" is triggered, the retrieved comment data will be the user-entered comment data since the last comment category. When the preset condition "the current time reaches the target project's preset comment category time" is triggered, the retrieved comment data will be the user-entered comment data since the target project was launched.

[0069] In one possible implementation, the method for performing feature extraction processing on multiple comment data separately can be: based on multiple comment data and a feature extraction model, determine the feature vector corresponding to each comment data.

[0070] In practice, multiple comment data can be input into a pre-trained feature extraction model to obtain the feature vector corresponding to each comment data.

[0071] In this embodiment of the disclosure, the feature extraction model can be any reasonable deep learning model, such as a semantic feature extraction model, etc.

[0072] In another possible implementation, the method of performing feature extraction processing on multiple comment data separately can also be as follows: based on multiple comment data and a topic feature extraction model, determine the topic feature vector corresponding to each comment data; based on multiple comment data and a semantic feature extraction model, determine the semantic feature vector corresponding to each comment data; based on the topic feature vector and semantic feature vector corresponding to each comment data, determine the feature vector corresponding to each comment data.

[0073] In practice, multiple comment data can be input into a pre-trained topic feature extraction model to obtain a topic feature vector corresponding to each comment data. The topic feature vector corresponding to the comment data focuses on representing the topic significance of the comment data.

[0074] Multiple comment data can also be input into a pre-trained semantic feature extraction model to obtain a semantic feature vector corresponding to each comment data. The semantic feature vector corresponding to the comment data focuses on representing the meaning of the entire sentence in the comment data.

[0075] Then, for each comment data, the topic feature vector and the semantic feature vector corresponding to the comment data can be fused to obtain the feature vector corresponding to the comment data. This feature vector contains both the topic meaning and the semantic meaning of the comment data, which can more accurately reflect the problem pointed out by the comment data, thereby improving the accuracy of comment classification.

[0076] In this embodiment, the topic feature extraction model can be an LDA (Latent Dirichlet Allocation) topic model. LDA is a classic method for document modeling in natural language processing. Its core principle is to treat each document as a mixture of multiple topics, and each topic is composed of a probability distribution of a set of words. LDA assumes that each word in the document is generated by its respective topic, and each topic is generated by words in the vocabulary set according to a certain probability distribution. By iteratively optimizing the document-topic distribution and the topic-vocabulary distribution, LDA can maximize the generation probability of words in the document, thereby revealing the potential topic structure in the document set.

[0077] In the LDA topic model, comment data can be composed of word sequences w = {w1, w2, ..., w...} n The entire corpus consists of a series of comment texts, D = {d1, d2, ..., d...}. m}, where n represents the number of words in each comment data, and m represents the number of comment texts in the corpus. The probability relationship between latent topics and words w calculated by the LDA topic model can be expressed as:

[0078]

[0079] Among them, z i Indicates the word "w" in the document i The corresponding theme.

[0080] Of course, the topic feature extraction model can also be other structures and their learning models, and this disclosure does not limit this.

[0081] In this embodiment of the disclosure, the semantic feature extraction model may be a BERT (Bidirectional Encoder Representations from Transformers) model or a SBert (Sentence Bidirectional Encoder Representations from Transformers, a feature extraction model based on an improved BERT model), etc. Of course, the semantic feature extraction model may also be any other reasonable feature extraction model, and this embodiment of the disclosure does not limit it.

[0082] When training the semantic feature extraction model, a Siamese network architecture can be used. This involves using two semantic feature extraction models with shared weights to encode the input comment data, and performing pooling on the output of all tokens in the last layer to obtain the predicted semantic feature vectors corresponding to the comment data. Then, the vector similarity (e.g., cosine similarity) between the predicted semantic feature vectors of the two comment data is calculated. Finally, based on the actual similarity between the two comment data, the calculated vector similarity, and the loss function, the loss value is calculated to train the semantic feature extraction model.

[0083] Furthermore, based on the topic feature vector and semantic feature vector corresponding to each comment data, the fusion method for determining the feature vector corresponding to each comment data can be: concatenating the topic feature vector and semantic feature vector corresponding to each comment data separately to obtain the feature vector corresponding to each comment data.

[0084] In practice, for each comment data, the topic feature vector and semantic feature vector corresponding to that comment data can be directly concatenated to obtain the feature vector corresponding to each comment data.

[0085] Alternatively, for each comment data, the topic feature vector and semantic feature vector corresponding to that comment data can be added together positionally to obtain the feature vector corresponding to each comment data, and so on. Of course, other reasonable fusion methods can also be used, and this disclosure does not limit them.

[0086] In one possible implementation, the comment data can be initial comment data input by the user into the vehicle's host system, which may also include text data corresponding to text input data and / or voice input data.

[0087] In another possible implementation, the comment data can be preprocessed data obtained from the initial comment data input by the user into the vehicle's main unit. Before performing feature extraction on the acquired comment data, the following processing can also be performed:

[0088] Obtain multiple initial comment data sets; perform text preprocessing on the multiple initial comment data sets to obtain the final comment data.

[0089] In practice, the initial comment data may contain a lot of meaningless text content. For example, when a user inputs a sentence, they may repeat it multiple times. Or, when a user outputs a comment by voice, the corresponding text data may contain a lot of repeated interjections, and so on.

[0090] Therefore, after obtaining the initial comment data, it can be preprocessed to remove meaningless text, thereby obtaining comment data that more accurately expresses the comment issues.

[0091] In one possible implementation, text preprocessing may include at least one of text deduplication, compression and word removal, sentence deletion, pure symbol deletion, pure English deletion, text segmentation, and stop word removal. Of course, it may also include other text preprocessing methods, which are not limited in this embodiment.

[0092] Text deduplication refers to deleting parts of multiple initial comment data that are completely repeated by the same user within a short period of time, and keeping only one of the initial comment data.

[0093] Word compression refers to the process of compressing words to remove meaningless words. Since vehicle users use both voice and text input modes when using the comment application, repeated words may appear in the same user feedback sentence, such as "my my my..." or "car map map map...", which affects the semantic expression of the whole sentence. Therefore, meaningless repeated words such as "my" and "map" are compressed.

[0094] Short sentence deletion refers to removing excessively short initial comment data. Extremely short sentences can convey little information and are often caused by user "incomplete input" or "misoperation". Therefore, a lower limit for sentence length is set during text preprocessing, and initial comment data with sentence lengths less than this lower limit is deleted to ensure the integrity of the semantic expression of the corpus.

[0095] Pure symbol removal refers to using regular expression matching to delete meaningless short comments consisting solely of pure symbols and pure English from the initial comment data. Since pure symbolic garbled text is meaningless, the initial comment data containing pure symbols is deleted.

[0096] The deletion of purely English text refers to the following: If the comment classification method provided in this embodiment targets the domestic car market, then all the algorithms used are in Chinese, so purely English text will be deleted.

[0097] Text segmentation processing refers to the ability to segment initial comment data to obtain segmentation information for each initial text data, meaning that the comment data includes segmentation information.

[0098] Stop word removal refers to removing some meaningless words such as auxiliary words and interjections from the initial comment data. These words can interfere with subsequent frequency statistics and other analytical operations.

[0099] Typically, one or more of the above text preprocessing techniques can enable the obtained comment data to more accurately express the problems reflected by users, thereby simplifying subsequent calculations and improving classification accuracy.

[0100] 102. Perform topic clustering on the feature vectors corresponding to multiple comment data to obtain topic information corresponding to each comment data.

[0101] In practice, after obtaining the feature vector corresponding to each comment data in multiple comment data, topic clustering can be performed based on these multiple feature vectors. That is, these multiple feature vectors are matched to the corresponding topic information respectively, so as to obtain the topic information corresponding to each feature vector, and then obtain the topic information corresponding to each comment data.

[0102] The topic information can include host hardware, vehicle body hardware, ecosystem software, host interconnection, cameras, powertrain, fuel consumption, fuel tank, etc.

[0103] The main unit hardware can include audio systems, central control unit, passenger-side screen, view-activated screen, return button, chips, black screen, etc. The vehicle body hardware can include the body, trunk, door buttons, physical buttons, roof rack, tires, brakes, windshield wipers, airbags, seats, rearview mirrors, etc. The ecosystem software can include navigation applications, various video applications, various music applications, various shopping applications, scene modes, application management, personal center, map applications, etc. The main unit interconnectivity can include various interconnected application software. Cameras can include panoramic cameras, radar, cameras, facial recognition devices, panoramic imaging, etc. The powertrain system can include the engine, coolant, transmission, etc. Fuel consumption and fuel tank functionality can include the fuel tank, accelerator, fuel consumption, etc.

[0104] In addition to the various topics listed above, other topics may also be included, but this disclosure does not limit the scope of such topics.

[0105] In this embodiment of the disclosure, the topic clustering process described above can be any reasonable clustering algorithm, such as the K-means algorithm, etc.

[0106] 103. Perform frequency statistics on the topic data corresponding to multiple comment data to obtain the frequency of occurrence of each topic information.

[0107] In practice, after classifying each comment data into topics through the above 102 steps, the frequency of occurrence of these multiple topics can be calculated, that is, the frequency of occurrence of each topic in these multiple comment data can be calculated.

[0108] For example, for a certain topic, we can first calculate the number of comments containing that topic, and denote it as the first number. Then, we can calculate the ratio between the first number and the total number of comments. This ratio is the frequency of the topic.

[0109] 104. Based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, and the classification model, determine the demand intensity classification corresponding to each topic information.

[0110] In implementation, the topic information corresponding to these multiple comment data and the frequency of occurrence of multiple topic information can be input into the classification model to obtain the demand intensity classification corresponding to each topic information. The demand intensity classification corresponding to the topic information is used to indicate the priority of after-sales processing of the topic information.

[0111] The intensity of demand can be categorized into basic demand, expected demand, attractive demand, and irrelevant demand.

[0112] The topics that correspond to basic needs are those that users pay close attention to. While the fulfillment of these needs may not inspire users to agree or share, users are likely to generate and express strong negative emotions if these needs are not met.

[0113] The topics that correspond to the expected needs are also topics that users pay close attention to. The fulfillment of these needs will inspire users to agree or share. If these needs are not met, the intensity of the emotions expressed by users will be lower than that of basic needs.

[0114] The topics corresponding to attractive needs are those with average user attention and fluctuating levels of user feedback controversy, making them a type of need that is in a state of flux.

[0115] The topics related to irrelevant needs are those that users pay little attention to and that are comments that are not closely related to or have no relation to the actual product.

[0116] The above classification model is used to obtain the demand intensity classification for each topic information. Then, these comment data are distributed to each after-sales professional department of the car company according to the category of the topic information. The professional department will formulate improvement strategies based on the urgency of the demand intensity classification and the category of the demand, so as to realize an "end-to-end" solution from user comment data to after-sales issues.

[0117] After achieving end-to-end requirement identification, the above automated methods can be used periodically to track newly added comment data, achieving targeted problem tracking. Regular statistical analysis of each topic and requirement intensity category enables automated evaluation of the after-sales department's problem-solving effectiveness.

[0118] In this embodiment of the disclosure, the classification model can be any reasonable algorithm model or machine learning model, such as the Kano model, etc., and this embodiment of the disclosure does not limit it.

[0119] In one possible implementation, before determining the demand intensity classification, the following processing can be performed: based on multiple comment data and the sentiment intensity model, determine the sentiment intensity value corresponding to each comment data, wherein the sentiment intensity value corresponding to the comment data is used to indicate the degree of sentiment intensity and sentiment polarity of the comment data.

[0120] In practice, multiple comment data can be input into the sentiment intensity model to obtain the sentiment intensity value corresponding to each comment data. The value of the sentiment intensity value can range from 0 to 1.

[0121] In this system, 0 represents extremely negative emotion, 0.5 represents neutral emotion, and 1 represents extremely positive emotion. Values ​​between 0 and 0.5 represent negative emotions, with the closer to 0, the stronger the negative sentiment. Values ​​between 0.5 and 1 represent positive emotions, with the closer to 1, the stronger the positive sentiment.

[0122] Furthermore, when determining the intensity of demand classification, the following processing can be performed: based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, the sentiment intensity value corresponding to each comment data, and the classification model, determine the intensity of demand classification corresponding to each topic information.

[0123] The specific processing method can be as follows: the topic information corresponding to the comment data, the frequency of occurrence of multiple topic information, and the emotional intensity value corresponding to multiple comment data can be input into the classification model for classification, and the demand intensity classification corresponding to each topic information can be obtained.

[0124] Alternatively, the specific processing method could be as follows: Based on the sentiment intensity value corresponding to each comment and the topic information corresponding to each comment, calculate the average sentiment score for each topic. The average sentiment score for each topic is used to characterize the degree of user emotion towards that topic in the comment data. If the average sentiment score is low, it indicates that the user's emotion towards the topic is not strong, and the priority for after-sales processing can be lower; if the average sentiment score is high, it indicates that the user's emotion towards the topic is strong, and the priority for after-sales processing can be higher. In this way, issues that have generated significant user feedback can be resolved promptly.

[0125] The average sentiment score can be calculated as follows: For each topic, add the sentiment intensity values ​​of one or more comment data corresponding to that topic to obtain a first value. Then, calculate the ratio of the first value to the number of comment data corresponding to that topic. This ratio is the average sentiment score corresponding to that topic.

[0126] After obtaining the average sentiment score corresponding to each topic information based on the above method, the average sentiment score corresponding to each topic information can be determined by the Kano model. The processing method of the Kano formula is shown in the following formula (2).

[0127]

[0128] Where t is the topic information, K(t) is the demand intensity category corresponding to topic information t, f(t) is the occurrence frequency corresponding to topic information t, S(t) is the average sentiment score corresponding to topic information t, and α, β, δ, γ and θ are all custom values, which can be set according to needs or actual conditions. This disclosure does not impose specific limitations on the embodiments.

[0129] All of the above-mentioned optional technical solutions can be combined in any way to form optional embodiments of this disclosure, and will not be described in detail here.

[0130] The solution mentioned in this embodiment can determine the topic information corresponding to each comment data through feature extraction and topic clustering. Furthermore, it can determine the demand intensity category corresponding to each topic information through a classification model. This allows staff to prioritize each topic information based on its demand intensity category and then handle the corresponding after-sales issues accordingly. This eliminates the need for manual filtering of comment data, saving manpower and resources. Moreover, staff can directly handle after-sales issues based on demand intensity categories, enabling timely and rapid resolution of more pressing after-sales problems, thereby improving the user's driving experience.

[0131] This disclosure provides an apparatus for classifying comments, which may be a computer device as described in the above embodiments, such as... Figure 3 As shown, the device includes:

[0132] The feature extraction module 320 is used to perform feature extraction processing on the acquired multiple comment data respectively to obtain the feature vector corresponding to each comment data;

[0133] Clustering module 330 is used to perform topic clustering processing on the feature vectors corresponding to the multiple comment data to obtain topic information corresponding to each comment data.

[0134] The statistics module 340 is used to perform frequency statistics processing on the topic data corresponding to the multiple comment data to obtain the frequency of occurrence of each topic information;

[0135] The classification module 360 ​​is used to determine the demand intensity classification corresponding to each topic information based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, and the classification model. The demand intensity classification corresponding to the topic information is used to indicate the after-sales processing priority of the topic information.

[0136] In one possible implementation, the feature extraction module 320 is used for:

[0137] Based on the multiple comment data and the topic feature extraction model, the topic feature vector corresponding to each comment data is determined;

[0138] Based on the multiple comment data and the semantic feature extraction model, the semantic feature vector corresponding to each comment data is determined;

[0139] Based on the topic feature vector and the semantic feature vector corresponding to each comment data, the feature vector corresponding to each comment data is determined.

[0140] In one possible implementation, the feature extraction module 320 is used for:

[0141] The topic feature vector and the semantic feature vector corresponding to each comment data are concatenated to obtain the feature vector corresponding to each comment data.

[0142] In one possible implementation, the device further includes a determining module 350, configured to:

[0143] Based on the multiple comment data and the sentiment intensity model, a sentiment intensity value corresponding to each comment data is determined, wherein the sentiment intensity value corresponding to the comment data is used to indicate the degree of sentiment intensity and sentiment polarity of the comment data;

[0144] The classification module 360 ​​is used for:

[0145] Based on the topic information corresponding to each comment data, the frequency of occurrence of each topic information, the emotional intensity value corresponding to each comment data, and the classification model, the demand intensity classification corresponding to each topic information is determined.

[0146] In one possible implementation, the apparatus further includes a preprocessing module 310, for:

[0147] Obtain multiple initial comment data;

[0148] The initial comment data is preprocessed to obtain the comment data.

[0149] In one possible implementation, the text preprocessing includes at least one of the following: text deduplication, compression and word removal, short sentence deletion, pure symbol deletion, pure English deletion, text segmentation, and stop word removal.

[0150] It should be noted that the comment classification device provided in the above embodiments is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the comment classification device and the comment classification method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0151] Figure 4 A structural block diagram of a terminal 400 provided in an exemplary embodiment of this disclosure is shown. The terminal can be a computer device as described in the above embodiments. The terminal 400 can be a smartphone, tablet computer, MP3 player (Moving Picture Experts Group Audio Layer III), MP4 player (Moving Picture Experts Group Audio Layer IV), laptop computer, or desktop computer. The terminal 400 may also be referred to as a user device, portable terminal, laptop terminal, desktop terminal, or other names.

[0152] Typically, terminal 400 includes a processor 401 and a memory 402.

[0153] Processor 401 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 401 may be implemented using at least one hardware form selected from DSP (digital signal processing), FPGA (field-programmable gate array), and PLA (programmable logic array). Processor 401 may also include a main processor and a coprocessor. The main processor, also known as a CPU (central processing unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 401 may integrate a GPU (graphics processing unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, processor 401 may also include an AI (artificial intelligence) processor, which is used to handle computational operations related to machine learning.

[0154] Memory 402 may include one or more computer-readable storage media, which may be non-transitory. Memory 402 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in memory 402 are used to store at least one instruction, which is executed by processor 401 to implement the comment classification method provided in the method embodiments of this disclosure.

[0155] In some embodiments, the terminal 400 may also optionally include a peripheral device interface 403 and at least one peripheral device. The processor 401, memory 402, and peripheral device interface 403 can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface 403 via a bus, signal line, or circuit board. Specifically, the peripheral device includes at least one of the following: a radio frequency circuit 404, a display screen 405, a camera 406, an audio circuit 407, a positioning component 408, and a power supply 409.

[0156] Peripheral device interface 403 can be used to connect at least one I / O (input / output) related peripheral device to processor 401 and memory 402. In some embodiments, processor 401, memory 402 and peripheral device interface 403 are integrated on the same chip or circuit board; in some other embodiments, any one or two of processor 401, memory 402 and peripheral device interface 403 can be implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0157] The radio frequency (RF) circuit 404 is used to receive and transmit RF (radio frequency) signals, also known as electromagnetic signals. The RF circuit 404 communicates with communication networks and other communication devices via electromagnetic signals. The RF circuit 404 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals back into electrical signals. Optionally, the RF circuit 404 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, etc. The RF circuit 404 can communicate with other terminals through at least one wireless communication protocol. This wireless communication protocol includes, but is not limited to: metropolitan area networks (MANs), various generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area networks (WLANs), and / or WiFi (wireless fidelity) networks. In some embodiments, the RF circuit 404 may also include circuitry related to NFC (near field communication), which is not limited in this disclosure.

[0158] Display screen 405 is used to display a user interface (UI). This UI may include graphics, text, icons, videos, and any combination thereof. When display screen 405 is a touch display screen, it also has the ability to collect touch signals on or above its surface. These touch signals can be input as control signals to processor 401 for processing. In this case, display screen 405 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there may be one display screen 405, which serves as the front panel of terminal 400; in other embodiments, there may be at least two display screens, respectively disposed on different surfaces of terminal 400 or in a folded design; in still other embodiments, display screen 405 may be a flexible display screen, disposed on a curved or folded surface of terminal 400. Furthermore, display screen 405 may be configured as a non-rectangular, irregular shape, i.e., a non-rectangular screen. Display screen 405 may be made of materials such as LCD (liquid crystal display) or OLED (organic light-emitting diode).

[0159] The camera assembly 406 is used to acquire images or videos. Optionally, the camera assembly 406 includes a front-facing camera and a rear-facing camera. Typically, the front-facing camera is located on the front panel of the terminal, and the rear-facing camera is located on the back of the terminal. In some embodiments, there are at least two rear-facing cameras, which are any one of a main camera, a depth-sensing camera, a wide-angle camera, and a telephoto camera, to achieve background blurring by fusion of the main camera and the depth-sensing camera, panoramic shooting by fusion of the main camera and the wide-angle camera, VR (virtual reality) shooting, or other fusion shooting functions. In some embodiments, the camera assembly 406 may also include a flash. The flash can be a single-color temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm-light flash and a cool-light flash, which can be used for light compensation at different color temperatures.

[0160] The audio circuit 407 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, converting them into electrical signals that are input to the processor 401 for processing, or to the radio frequency circuit 404 for voice communication. For stereo sound acquisition or noise reduction purposes, multiple microphones may be used, each positioned at a different location on the terminal 400. The microphone may also be an array microphone or an omnidirectional microphone. The speaker is used to convert electrical signals from the processor 401 or the radio frequency circuit 404 into sound waves. The speaker may be a conventional diaphragm speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can convert electrical signals not only into audible sound waves but also into inaudible sound waves for purposes such as distance measurement. In some embodiments, the audio circuit 407 may also include a headphone jack.

[0161] The positioning component 408 is used to determine the current geographic location of the terminal 400 in order to enable navigation or LBS (location-based service). The positioning component 408 can be a positioning component based on GPS (global positioning system), BeiDou system, Grenadin system, or Galileo system.

[0162] Power supply 409 is used to power the various components in terminal 400. Power supply 409 can be AC ​​power, DC power, a disposable battery, or a rechargeable battery. When power supply 409 includes a rechargeable battery, the rechargeable battery can support wired charging or wireless charging. The rechargeable battery can also be used to support fast charging technology.

[0163] In some embodiments, the terminal 400 further includes one or more sensors 410. The one or more sensors 410 include, but are not limited to: an accelerometer 411, a gyroscope 412, a pressure sensor 413, a fingerprint sensor 414, an optical sensor 415, and a proximity sensor 416.

[0164] Accelerometer 411 can detect the magnitude of acceleration along the three coordinate axes of a coordinate system established by terminal 400. For example, accelerometer 411 can be used to detect the components of gravitational acceleration along the three coordinate axes. Processor 401 can control display screen 405 to display the user interface in either a landscape or portrait view based on the gravitational acceleration signal acquired by accelerometer 411. Accelerometer 411 can also be used for games or for acquiring user motion data.

[0165] The gyroscope sensor 412 can detect the orientation and rotation angle of the terminal 400. The gyroscope sensor 412, in conjunction with the accelerometer sensor 411, can collect 3D motion data from the user on the terminal 400. Based on the data collected by the gyroscope sensor 412, the processor 401 can perform the following functions: motion sensing (e.g., changing the UI based on the user's tilt), image stabilization during shooting, game control, and inertial navigation.

[0166] The pressure sensor 413 can be disposed on the side bezel of the terminal 400 and / or on the lower layer of the display screen 405. When the pressure sensor 413 is disposed on the side bezel of the terminal 400, it can detect the user's grip signal on the terminal 400, and the processor 401 can perform left / right hand recognition or quick operation based on the grip signal collected by the pressure sensor 413. When the pressure sensor 413 is disposed on the lower layer of the display screen 405, the processor 401 can control the operable controls on the UI interface based on the user's pressure operation on the display screen 405. The operable controls include at least one of button controls, scroll bar controls, icon controls, and menu controls.

[0167] The fingerprint sensor 414 is used to collect the user's fingerprint. The processor 401 identifies the user's identity based on the fingerprint collected by the fingerprint sensor 414, or the fingerprint sensor 414 identifies the user's identity based on the collected fingerprint. When the user's identity is identified as trusted, the processor 401 authorizes the user to perform relevant sensitive operations, including unlocking the screen, viewing encrypted information, downloading software, making payments, and changing settings. The fingerprint sensor 414 can be located on the front, back, or side of the terminal 400. When the terminal 400 has physical buttons or a manufacturer's logo, the fingerprint sensor 414 can be integrated with the physical buttons or manufacturer's logo.

[0168] An optical sensor 415 is used to collect ambient light intensity. In one embodiment, the processor 401 can control the display brightness of the display screen 405 based on the ambient light intensity collected by the optical sensor 415. Specifically, when the ambient light intensity is high, the display brightness of the display screen 405 is increased; when the ambient light intensity is low, the display brightness of the display screen 405 is decreased. In another embodiment, the processor 401 can also dynamically adjust the shooting parameters of the camera assembly 406 based on the ambient light intensity collected by the optical sensor 415.

[0169] The proximity sensor 416, also known as a distance sensor, is typically located on the front panel of the terminal 400. The proximity sensor 416 is used to detect the distance between the user and the front of the terminal 400. In one embodiment, when the proximity sensor 416 detects that the distance between the user and the front of the terminal 400 is gradually decreasing, the processor 401 controls the display screen 405 to switch from a screen-on state to a screen-off state; when the proximity sensor 416 detects that the distance between the user and the front of the terminal 400 is gradually increasing, the processor 401 controls the display screen 405 to switch from a screen-off state to a screen-on state.

[0170] Those skilled in the art will understand that Figure 4 The structure shown does not constitute a limitation on terminal 400 and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0171] Figure 5 This is a schematic diagram of a server structure provided in an embodiment of this disclosure. The server 500 can vary significantly due to differences in configuration or performance. It may include one or more central processing units (CPUs) 501 and one or more memories 502. The memories 502 store at least one instruction, which is loaded and executed by the processors 501 to implement the methods provided in the various method embodiments described above. Of course, the server may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The server may also include other components for implementing device functions, which will not be elaborated upon here.

[0172] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including instructions that can be executed by a processor in a terminal to perform the comment classification method described above. This computer-readable storage medium may be non-transitory. For example, the computer-readable storage medium may be ROM (read-only memory), RAM (random access memory), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0173] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0174] It should be noted that all information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals (including but not limited to signals transmitted between user terminals and other devices) involved in this disclosure are authorized by the user or fully authorized by all parties, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the "comment data," "initial comment data," etc., involved in this disclosure were obtained with full authorization.

[0175] The above description is merely an optional embodiment of this disclosure and is not intended to limit this disclosure. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the protection scope of this disclosure.

Claims

1. A method for classifying comments, characterized in that, The method includes: Based on the acquired multiple comment data and topic feature extraction model, the topic feature vector corresponding to each comment data is determined; Based on the multiple comment data and the semantic feature extraction model, the semantic feature vector corresponding to each comment data is determined; The topic feature vector and the semantic feature vector corresponding to each comment data are concatenated to obtain the feature vector corresponding to each comment data. Perform topic clustering on the feature vectors corresponding to the multiple comment data to obtain topic information corresponding to each comment data; Frequency statistics are performed on the topic data corresponding to the multiple comment data to obtain the frequency of occurrence of each topic information; Based on the multiple comment data and the sentiment intensity model, determine the sentiment intensity value corresponding to each comment data; Based on the sentiment intensity value and topic information corresponding to each comment data, calculate the average sentiment score corresponding to each topic information; Based on the following formula, the demand intensity category corresponding to each topic information is determined, and the demand intensity category corresponding to the topic information is used to indicate the after-sales processing priority of the topic information; in, It is topic information. This is the topic information. Corresponding demand intensity classification, This is the topic information. Corresponding frequency of occurrence This is the topic information. The corresponding average sentiment score, , , , , All values ​​are user-defined.

2. The method according to claim 1, characterized in that, Before determining the topic feature vector corresponding to each comment data based on the acquired multiple comment data and topic feature extraction model, the method further includes: Obtain multiple initial comment data; The initial comment data is preprocessed to obtain the comment data.

3. The method according to claim 2, characterized in that, The text preprocessing includes at least one of the following: text deduplication, compression and word removal, short sentence deletion, pure symbol deletion, pure English deletion, text segmentation, and stop word removal.

4. A comment classification device, characterized in that, The device includes: The feature extraction module is used to determine the topic feature vector corresponding to each comment data based on the acquired multiple comment data and the topic feature extraction model; to determine the semantic feature vector corresponding to each comment data based on the multiple comment data and the semantic feature extraction model; and to concatenate the topic feature vector and the semantic feature vector corresponding to each comment data to obtain the feature vector corresponding to each comment data. The clustering module is used to perform topic clustering processing on the feature vectors corresponding to the multiple comment data to obtain topic information corresponding to each comment data. The statistics module is used to perform frequency statistics processing on the topic data corresponding to the multiple comment data to obtain the frequency of occurrence of each topic information; The classification module is used to determine the sentiment intensity value corresponding to each comment data based on the multiple comment data and the sentiment intensity model; and to calculate the average sentiment score corresponding to each topic information based on the sentiment intensity value corresponding to each comment data and the topic information. Based on the following formula, the demand intensity category corresponding to each topic information is determined, and the demand intensity category corresponding to the topic information is used to indicate the after-sales processing priority of the topic information; in, It is topic information. This is the topic information. Corresponding demand intensity classification, This is the topic information. Corresponding frequency of occurrence This is the topic information. The corresponding average sentiment score, , , , , All values ​​are user-defined.

5. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing at least one instruction, which is loaded and executed by the processor to perform the operation of the comment classification method as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, which is loaded and executed by a processor to perform the operations of the comment classification method as described in any one of claims 1 to 3.

7. A computer program product, characterized in that, The computer program product includes at least one instruction, which is loaded and executed by a processor to perform the operations performed by the comment classification method as described in any one of claims 1 to 3.

Images