A method, system, device and storage medium for detecting network violence

Abstract

The application discloses a network violence detection method, system, device and storage medium. First information published by a target object for a target event is input into a sentiment analysis model to obtain a first sentiment analysis result corresponding to the target object. When it is determined that other objects comment on the first information, second information of the other objects and third information of the target object are obtained. The second information and the third information are input into the sentiment analysis model to obtain a second sentiment analysis result corresponding to the other objects and a third sentiment analysis result corresponding to the target object. Whether the emotion of the target object changes to a negative direction is determined according to the first sentiment analysis result and the third sentiment analysis result. When the second sentiment analysis result is negative and the emotion of the target object changes to a negative direction, it is determined that the other objects have network violence behaviors. The method can effectively improve the accuracy of judging network violence behaviors. The application can be widely applied in the field of sentiment analysis.

Description

Technical Field

[0001] This application relates to the field of sentiment analysis technology, and in particular to a method, system, device, and storage medium for detecting cyberbullying. Background Technology

[0002] In recent years, with the popularity of the internet, detecting users' emotions and identifying cyberbullying incidents has become increasingly valuable, enabling online platforms to intervene and address these issues as early as possible. Traditional methods for identifying human emotions fall into two categories: one involves physically measuring physiological indicators such as electrocardiograms, electroencephalograms, speech, and facial expressions; the other uses non-contact methods to observe physiological characteristics. The development of artificial intelligence technology has led to the application of techniques based on sentiment analysis models to determine users' emotional tendencies.

[0003] In related technologies, sentiment analysis models are mostly applied in scenarios such as shopping websites, travel websites, and movie review websites. They acquire user comments, convert them into corresponding vector representations, and then use sentiment analysis models (usually multi-layer neural networks such as MLP, CNN, and LSTM) to categorize user comments as positive or negative. However, it can be observed that most of these sentiment analysis models only classify comments as positive or negative based on a single comment text, failing to understand the emotional impact of comment responses on users. This results in low accuracy in identifying cyberbullying and poor application effectiveness. Summary of the Invention

[0004] The purpose of this application is to at least partially solve one of the technical problems existing in the related art.

[0005] Therefore, one objective of the embodiments of this application is to provide a method, system, device, and storage medium for detecting cyberbullying.

[0006] To achieve the above-mentioned technical objectives, the technical solutions adopted in the embodiments of this application include:

[0007] On the one hand, embodiments of this application provide a method for detecting cyberbullying, the method comprising:

[0008] Retrieve the first information published by the target object in response to the target event;

[0009] The first information is input into the sentiment analysis model to perform sentiment analysis, and the first sentiment analysis result corresponding to the target object is obtained.

[0010] When it is determined that other objects have commented on or replied to the first information, the second information of the other objects' comments and replies, as well as the third information published by the target object in response to the target event, are obtained.

[0011] Each of the second and third pieces of information is input into the sentiment analysis model to perform sentiment analysis, thereby obtaining the second sentiment analysis result corresponding to the other objects and the third sentiment analysis result corresponding to the target object;

[0012] Based on the first sentiment analysis result and the third sentiment analysis result, determine whether the target object's emotion has shifted to a negative direction;

[0013] When the second sentiment analysis result is negative, and the target object's emotion shifts to negative, it is determined that the other objects have engaged in cyberbullying.

[0014] In addition, the method for detecting cyberbullying according to the above embodiments of this application may also have the following additional technical features:

[0015] Furthermore, in one embodiment of this application, the step of inputting the first information into a sentiment analysis model for sentiment analysis to obtain a first sentiment analysis result corresponding to the target object includes:

[0016] The first information is input into the sentiment analysis model to extract the first feature data of the first information;

[0017] Based on the first feature data, the first sentiment analysis result corresponding to the target object is obtained.

[0018] Furthermore, in one embodiment of this application, the extraction of the first feature data of the first information includes:

[0019] The first information is segmented to obtain word groups corresponding to the first information, and word feature data corresponding to each word in the word group is determined; wherein, the word group includes several words;

[0020] Based on the word feature data corresponding to each word in the phrase, determine the first feature data corresponding to the first information.

[0021] Furthermore, in one embodiment of this application, the word feature data includes word embedding vectors; determining the word feature data corresponding to each word in the word group includes:

[0022] Map the words to a vector space to obtain the word embedding vectors corresponding to the words;

[0023] The step of determining the first feature data corresponding to the first information based on the word feature data corresponding to each word in the phrase includes:

[0024] The word embedding vectors corresponding to each word in the phrase are summed to obtain the first vector;

[0025] The first vector is normalized to obtain the first feature data.

[0026] Furthermore, in one embodiment of this application, determining whether the target object's emotion has shifted negatively based on the first sentiment analysis result and the third sentiment analysis result includes:

[0027] If the first sentiment analysis result is positive and the third sentiment analysis result is negative, it is determined that the target object's emotion has shifted to a negative direction.

[0028] Furthermore, in one embodiment of this application, determining whether the target object's emotion has shifted negatively based on the first sentiment analysis result and the third sentiment analysis result includes:

[0029] If both the first sentiment analysis result and the third sentiment analysis result are negative, obtain the first negative probability corresponding to the first sentiment analysis result and the second negative probability corresponding to the third sentiment analysis result;

[0030] Calculate the difference between the second negative probability and the first negative probability;

[0031] If the difference is greater than a first preset threshold, it is determined that the target object's emotion has shifted to a negative direction.

[0032] Furthermore, in one embodiment of this application, the method further includes:

[0033] Based on the second sentiment analysis results, calculate the proportion of the second information that is negative information;

[0034] If the ratio is greater than the second preset threshold, the second sentiment analysis result is determined to be negative.

[0035] On the other hand, embodiments of this application provide a cyberbullying detection system, the system comprising:

[0036] The first acquisition module is used to acquire the first information published by the target object in response to the target event;

[0037] The first analysis module is used to input the first information into the sentiment analysis model to perform sentiment analysis and obtain the first sentiment analysis result corresponding to the target object;

[0038] The second acquisition module is used to acquire, when it is determined that other objects have commented on or replied to the first information, second information of the other objects' comment replies and third information published by the target object in response to the target event;

[0039] The second analysis module is used to input each of the second information and the third information into the sentiment analysis model to perform sentiment analysis, and obtain the second sentiment analysis result corresponding to the other objects and the third sentiment analysis result corresponding to the target object.

[0040] The comparison module is used to determine whether the target object's emotion has shifted to a negative direction based on the first sentiment analysis result and the third sentiment analysis result;

[0041] The processing module is used to determine that the other objects have engaged in cyberbullying behavior when the second sentiment analysis result is negative and the target object's sentiment shifts to negative.

[0042] On the other hand, embodiments of this application provide a terminal device, including:

[0043] At least one processor;

[0044] At least one memory for storing at least one program;

[0045] When the at least one program is executed by the at least one processor, the at least one processor implements the above-described method for detecting cyberbullying.

[0046] On the other hand, embodiments of this application also provide a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement the aforementioned method for detecting cyberbullying.

[0047] The advantages and beneficial effects of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application:

[0048] This application discloses a method for detecting cyberbullying. The method includes: acquiring first information posted by a target object in response to a target event; inputting the first information into a sentiment analysis model for sentiment analysis to obtain a first sentiment analysis result corresponding to the target object; when it is determined that other objects comment on or reply to the first information, acquiring second information of the other objects' comment replies and third information posted by the target object in response to the target event; inputting each of the second and third information into the sentiment analysis model for sentiment analysis to obtain a second sentiment analysis result corresponding to the other objects and a third sentiment analysis result corresponding to the target object; determining whether the target object's emotion has shifted to a negative direction based on the first and third sentiment analysis results; when the second sentiment analysis result is negative and the target object's emotion has shifted to a negative direction, determining that the other objects have engaged in cyberbullying behavior. This method can effectively improve the accuracy of judging cyberbullying behavior, which is conducive to enabling network platforms to intervene and rectify the situation as early as possible, thereby improving the network environment. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following description is provided with accompanying drawings of the relevant technical solutions in the embodiments of this application or the prior art. It should be understood that the accompanying drawings described below are only for the purpose of clearly illustrating some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0050] Figure 1 This is a flowchart illustrating a method for detecting cyberbullying provided in an embodiment of this application;

[0051] Figure 2 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. Detailed Implementation

[0052] The present application will be further described below with reference to the accompanying drawings and specific embodiments. The described embodiments should not be considered as limitations on the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0053] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0055] Before introducing the methods provided in the embodiments of this application, a brief description of some background technologies involved in this application will be given first.

[0056] Artificial Intelligence (AI) is the theory, methods, technology, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results. In other words, AI is a comprehensive technology within computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can react in a way similar to human intelligence. AI studies the design principles and implementation methods of various intelligent machines, enabling them to have perception, reasoning, and decision-making capabilities. AI technology is a comprehensive discipline involving a wide range of fields, encompassing both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly include computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0057] Machine Learning (ML) is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to endow computers with intelligence. Its applications span all areas of artificial intelligence. Machine learning (deep learning) typically includes techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and instructional learning.

[0058] In recent years, with the popularity of the internet, detecting users' emotions and identifying cyberbullying incidents has become increasingly valuable, enabling online platforms to intervene and address these issues as early as possible. Traditional methods for identifying human emotions fall into two categories: one involves physically measuring physiological indicators such as electrocardiograms, electroencephalograms, speech, and facial expressions; the other uses non-contact methods to observe physiological characteristics. The development of artificial intelligence technology has led to the application of techniques based on sentiment analysis models to determine users' emotional tendencies.

[0059] In related technologies, sentiment analysis models are mostly applied in scenarios such as shopping websites, travel websites, and movie review websites. They acquire user comments, convert them into corresponding vector representations, and then use sentiment analysis models (usually multi-layer neural networks such as MLP, CNN, and LSTM) to categorize user comments as positive or negative. However, it can be observed that most of these sentiment analysis models only classify comments as positive or negative based on a single comment text, failing to understand the emotional impact of comment responses on users. This results in low accuracy in identifying cyberbullying and poor application effectiveness.

[0060] In view of this, this application provides a method for detecting cyberbullying. This method applies a sentiment analysis model to the cyberbullying detection scenario, and comprehensively analyzes the impact of other objects on the target object by examining the emotional changes of users before and after a comment event, as well as the emotional tendency of the comment replies themselves, thereby detecting whether other objects have engaged in cyberbullying behavior. This method can effectively improve the accuracy of judging cyberbullying behavior, and is conducive to enabling network platforms to intervene and rectify the situation as early as possible, thus improving the network environment.

[0061] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for detecting cyberbullying provided in an embodiment of this application. (Refer to...) Figure 1 This method for detecting cyberbullying includes, but is not limited to:

[0062] Step 110: Obtain the first information published by the target object in response to the target event;

[0063] In this step, when detecting and identifying cyberbullying, the object being detected as a potential target can be designated as the target. The target can be a user of any software or website; this application does not impose specific restrictions. After identifying the target, information posted by the target in response to a target event can be obtained and designated as the first information. Specifically, the target event here can be a current trending event. For example, if the target posts a comment under a news article, the event corresponding to that news article is the target event, and the comment posted by the target is the first information. Of course, the specific type of target event and the format of the first information posting can be set according to the software or website settings; this application will not elaborate on this.

[0064] Step 120: Input the first information into the sentiment analysis model to perform sentiment analysis and obtain the first sentiment analysis result corresponding to the target object;

[0065] In this embodiment, after obtaining the first information, it can be input into a sentiment analysis model for sentiment analysis to obtain the first sentiment analysis result corresponding to the target object. Here, the sentiment analysis model can be built using any existing machine learning model structure, such as a multilayer neural network (MLP), LSTM, CNN, etc., and its specific training algorithm is not limited. Here, the first sentiment analysis result can be a sentiment tendency result represented by probabilities, for example, it can be a vector (a, b), where a can represent the probability of a positive sentiment, b can represent the probability of a negative sentiment, and the sum of the two can be 1. Specifically, the sentiment analysis model can use multiple linear layers and a sigmoid function as a classifier. Adding a sigmoid function after the linear layers of the classifier can transform the values ​​of the linear layers into values ​​between 0 and 1, thus ensuring that the output result conforms to the range constraint of the probability value.

[0066] Step 130: When it is determined that other objects have commented on or replied to the first information, obtain the second information of the other objects' comments and replies, as well as the third information published by the target object in response to the target event;

[0067] In this step, when other objects comment on or reply to the first message, it is possible to detect whether cyberbullying has occurred during their comment responses. Specifically, at this time, the second message from other objects' comments and replies, and the third message posted by the target object at a target time after the other objects' replies, can be obtained. The third message can be either a repost by the target object or a reply by the target object to the second message; this application does not limit this. Furthermore, the number of second messages here can be one or multiple.

[0068] Step 140: Input each of the second information and the third information into the sentiment analysis model to perform sentiment analysis, and obtain the second sentiment analysis result corresponding to the other objects and the third sentiment analysis result corresponding to the target object;

[0069] In this step, after obtaining the second and third information, both information can be input into the sentiment analysis model for analysis, thereby obtaining the second sentiment analysis results corresponding to other objects and the third sentiment analysis results corresponding to the target object. Here, the third sentiment analysis result is the sentiment tendency of the target object after communicating with other objects, which can reflect the emotional impact of other objects on the target object.

[0070] Step 150: Based on the first sentiment analysis result and the third sentiment analysis result, determine whether the target object's emotion has shifted to a negative direction;

[0071] In this step, as mentioned earlier, since the third sentiment analysis result reflects the emotional tendency of the target object after interacting with other objects, it can reflect the emotional influence of other objects on the target object. Therefore, after obtaining the third sentiment analysis result, it is possible to determine whether the target object's emotions have shifted towards a negative direction based on the first and third sentiment analysis results.

[0072] Specifically, in some embodiments, determining whether the target object's emotion has shifted negatively based on the first sentiment analysis result and the third sentiment analysis result includes:

[0073] If the first sentiment analysis result is positive and the third sentiment analysis result is negative, it is determined that the target object's emotion has shifted to a negative direction.

[0074] In the embodiments of this application, it can be understood that if the target object's original emotion was positive, but became negative after communicating with other objects, it indicates that the target object's emotion has changed to negative.

[0075] Specifically, in some embodiments, determining whether the target object's emotion has shifted negatively based on the first sentiment analysis result and the third sentiment analysis result includes:

[0076] If both the first sentiment analysis result and the third sentiment analysis result are negative, obtain the first negative probability corresponding to the first sentiment analysis result and the second negative probability corresponding to the third sentiment analysis result;

[0077] Calculate the difference between the second negative probability and the first negative probability;

[0078] If the difference is greater than a first preset threshold, it is determined that the target object's emotion has shifted to a negative direction.

[0079] In this embodiment, it is understood that, assuming the target object's original emotion is negative, and remains negative after communicating with other objects, the probabilities corresponding to the two negative emotions fluctuate significantly. For example, the first negative probability corresponding to the first sentiment analysis result is relatively low, indicating that the target object's negative emotion is relatively mild; the second negative probability corresponding to the third sentiment analysis result is relatively high, indicating that the target object's negative emotion is relatively severe, or that the target object's emotion has shifted towards a negative direction. Specifically, in this embodiment, a first preset threshold can be set, for example, the threshold value could be 20%. Then, the difference between the second negative probability and the first negative probability can be calculated. If this difference is greater than the first preset threshold, it indicates that the target object's emotion has further shifted towards a negative direction. Of course, it should be noted that in this embodiment, the specific size of the first preset threshold is not limited, and it can be flexibly set as needed.

[0080] Step 160: When the second sentiment analysis result is negative and the target object's sentiment shifts to negative, it is determined that the other objects have engaged in cyberbullying.

[0081] In this step, if step 150 determined that the target's emotions shifted negatively, and the second sentiment analysis results for other objects were also negative, it indicates that the comments and replies made by other objects to the target object likely involved cyberbullying, and the negative information caused the target's emotions to shift negatively. Therefore, it can be determined that other objects engaged in cyberbullying.

[0082] In some embodiments, inputting the first information into a sentiment analysis model for sentiment analysis to obtain a first sentiment analysis result corresponding to the target object includes:

[0083] The first information is input into the sentiment analysis model to extract the first feature data of the first information;

[0084] Based on the first feature data, the first sentiment analysis result corresponding to the target object is obtained.

[0085] In this embodiment, when analyzing the first sentiment analysis result corresponding to the target object using the first information, the first information can be input into the sentiment analysis model to extract the feature data of the first information, which is denoted as the first feature data. Then, based on the first feature data, the sentiment tendency of the target object can be analyzed using a classifier to obtain the first sentiment analysis result corresponding to the target object. Of course, similarly, the determination process of other sentiment analysis results can be implemented by referring to the first sentiment analysis result, which will not be elaborated upon in this application.

[0086] Specifically, in this embodiment, when extracting the first feature data of the first information, it is understood that the first information itself is generally text information, which is composed of words, and words are the most basic unit in natural language processing technology. Therefore, in this embodiment, the first feature data corresponding to the first information can be determined based on each word of the first information. For example, the first information can be segmented first to obtain multiple words that make up the first information, and these words can be recorded as a word group, with each piece of first information corresponding to a word group. Specifically, in this embodiment, various word segmentation algorithms can be used. For example, in some embodiments, a dictionary-based word segmentation algorithm can be used, which first segments the first information into words according to the dictionary, and then finds the best combination of words; in some embodiments, a character-based word segmentation algorithm can also be used, which first segments the first information into individual characters, and then combines the characters into words to find the optimal combination. In this embodiment, after obtaining the word group corresponding to each statement, the first statement feature data corresponding to the statement can be determined based on the word feature data corresponding to each word in the word group.

[0087] In the embodiments of the present application, the word feature data is a data form used to represent words, and its data structure can also be any one of numerical values, vectors, matrices, or tensors. For example, in some embodiments, numerical values can be used to represent each word. For example, the numerical value "1" can be used to represent the Chinese word "I". In some embodiments, a dictionary can be established for all the words in the corpus, and then each word in the dictionary can be represented by a one-hot vector. For example, if there are 1000 words in the dictionary, the dimension of the one-hot vector is 1000. The corresponding one-hot vectors can be assigned to each word according to the order in which the words appear in the dictionary. For example, if the word "I" is the first word in the dictionary, the one-hot vector corresponding to "I" can be represented as (1, 0, 0...0, 0), where the first element in the vector is 1 and the remaining 999 elements are all 0. In some embodiments, in order to reduce the dimension of the vector representing the word, a word embedding vector can also be used to represent the word. The word embedding vector can be obtained by mapping the words in the dictionary to a vector space with a unified lower dimension. The strategies for generating this mapping include neural networks, dimensionality reduction of word co-occurrence matrices, probability models, and interpretable knowledge base methods, etc. For example, taking the use of a neural network to obtain the word embedding vector as an example, a trained word2vec (word to vector) model can be used to map the word embedding vectors of each word in the dictionary. Compared with the aforementioned one-hot vector, the word embedding vector also has a certain semantic representativeness, that is, words with similar semantics are also relatively close in the vector space, so it is more widely used. Of course, the above implementation manners are only used to illustrate the determination method of the word feature data in the present application, and the actual implementation process is not limited to the above specific examples.

[0088] In the embodiments of the present application, after obtaining the word feature data corresponding to each word in the phrase, the first feature data corresponding to the first information can be determined. For example, the word feature data can be weighted or concatenated to obtain the first feature data. In some embodiments, for example, the word feature data includes a word embedding vector, and the word embedding vector can be obtained by mapping the word to a vector space. When determining the first feature data, the word embedding vectors corresponding to each word in the phrase can be accumulated to obtain a first vector. For the convenience of normalization and data processing, the first vector can be normalized and used as the first feature data. The normalization method here can be to limit the norm of the first vector to a specific value. For example, each element in the first vector can be scaled proportionally so that the norm of the first vector is 1.

[0089] In some embodiments, the method further includes:

[0090] Calculating the proportion of the second information being negative information according to the second sentiment analysis result;

[0091] If the ratio is greater than the second preset threshold, the second sentiment analysis result is determined to be negative.

[0092] In this embodiment of the application, when determining whether the second sentiment analysis result is negative, in some cases, there may be multiple pieces of second information. In this case, the sentiment analysis result corresponding to each piece of second information can be determined, with positive ones recorded as positive information and negative ones recorded as negative information. Then, the proportion of second information that is negative information can be counted, and a second preset threshold can be set, for example, 70%. If this proportion is greater than the second preset threshold, it indicates that the second information as a whole is negative information, and in this case, the second sentiment analysis result can be determined to be negative.

[0093] It is understood that the cyber violence detection method provided in this application uses a sentiment analysis model for cyber violence detection and utilizes the sigmoid function to calculate the changes and degree of user emotions before and after the implementation of the behavior, analyzes the user's tendency to be affected, and fully considers the new viewpoints and information that users constantly generate as the event develops. This can effectively improve the accuracy of judging cyber violence behavior and help network platforms to intervene and rectify the situation as early as possible, thereby improving the network environment.

[0094] This application also provides a cyberbullying detection system, the system comprising:

[0095] The first acquisition module is used to acquire the first information published by the target object in response to the target event;

[0096] The first analysis module is used to input the first information into the sentiment analysis model to perform sentiment analysis and obtain the first sentiment analysis result corresponding to the target object;

[0097] The second acquisition module is used to acquire, when it is determined that other objects have commented on or replied to the first information, second information of the other objects' comment replies and third information published by the target object in response to the target event;

[0098] The second analysis module is used to input each of the second information and the third information into the sentiment analysis model to perform sentiment analysis, and obtain the second sentiment analysis result corresponding to the other objects and the third sentiment analysis result corresponding to the target object.

[0099] The comparison module is used to determine whether the target object's emotion has shifted to a negative direction based on the first sentiment analysis result and the third sentiment analysis result;

[0100] The processing module is used to determine that the other objects have engaged in cyberbullying behavior when the second sentiment analysis result is negative and the target object's sentiment shifts to negative.

[0101] Understandable Figure 1 The content of the embodiment of the cyberbullying detection method shown is applicable to the embodiment of this cyberbullying detection system. The specific functions implemented by the embodiment of this cyberbullying detection system are the same as those in this embodiment. Figure 1 The embodiment of the method for detecting cyberbullying shown is the same, and the beneficial effects achieved are the same. Figure 1 The beneficial effects achieved by the embodiment of the cyberbullying detection method shown are also the same.

[0102] Reference Figure 2 This application also discloses a terminal device, including:

[0103] At least one processor 201;

[0104] At least one memory 202 is used to store at least one program;

[0105] When at least one program is executed by at least one processor 201, such that at least one processor 201 performs as follows: Figure 1 This is an example of a method for detecting cyberbullying.

[0106] It is understandable that, such as Figure 1 The content of the embodiment of the method for detecting cyberbullying shown is applicable to the embodiment of this terminal device. The specific functions implemented by the embodiment of this terminal device are the same as those shown below. Figure 1 The embodiment of the method for detecting cyberbullying shown is the same, and the beneficial effects achieved are the same as those described above. Figure 1 The beneficial effects achieved by the embodiment of the cyberbullying detection method shown are also the same.

[0107] This application also discloses a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement, for example... Figure 1 This is an example of a method for detecting cyberbullying.

[0108] It is understandable that, such as Figure 1 The content of the embodiment of the method for detecting cyberbullying shown is applicable to the embodiment of this computer-readable storage medium. The specific functions implemented by the embodiment of this computer-readable storage medium are the same as those shown below. Figure 1 The embodiment of the method for detecting cyberbullying shown is the same, and the beneficial effects achieved are the same as those described above. Figure 1 The beneficial effects achieved by the embodiment of the cyberbullying detection method shown are the same.

[0109] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this application are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0110] Furthermore, although this application is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the functions and / or features may be integrated into a single physical system and / or software module, or one or more functions and / or features may be implemented in a separate physical system or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding this application. Rather, given the properties, functions, and internal relationships of the various functional modules in the system disclosed herein, the actual implementation of the module will be understood within the scope of ordinary skill of an engineer. Therefore, those skilled in the art can implement the application set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of this application, which is determined by the full scope of the appended claims and their equivalents.

[0111] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0112] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, system, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, system, or device). For the purposes of this specification, "computer-readable medium" can mean any system that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, system, or device.

[0113] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections with one or more wires (electronic systems), portable computer disk drives (magnetic systems), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic systems, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0114] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0115] In the foregoing description of this specification, the references to terms such as "one embodiment," "another embodiment," or "some embodiments," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0116] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

[0117] The foregoing detailed description of preferred embodiments of this application is not limited to these embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application, and all such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

[0118] In the description of this specification, the references to terms such as "one embodiment," "another embodiment," or "some embodiments," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0119] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for detecting cyberbullying, characterized in that, The method includes: Retrieve the first information published by the target object in response to the target event; The first information is input into the sentiment analysis model for sentiment analysis to obtain the first sentiment analysis result corresponding to the target object; When it is determined that other objects have commented on or replied to the first information, second information about the comments and replies of the other objects and third information published by the target object are obtained; wherein, the third information is information republished by the target object in response to the target event or information in response to the second information by the target object; Each of the second and third pieces of information is input into the sentiment analysis model to perform sentiment analysis, thereby obtaining the second sentiment analysis result corresponding to the other objects and the third sentiment analysis result corresponding to the target object; Based on the first sentiment analysis result and the third sentiment analysis result, determine whether the target object's emotion has shifted to a negative direction; When the second sentiment analysis result is negative, and the target object's emotion shifts to negative, it is determined that the other objects have engaged in cyberbullying.

2. The method for detecting cyberbullying according to claim 1, characterized in that, The step of inputting the first information into a sentiment analysis model for sentiment analysis to obtain a first sentiment analysis result corresponding to the target object includes: The first information is input into the sentiment analysis model to extract the first feature data of the first information; Based on the first feature data, the first sentiment analysis result corresponding to the target object is obtained.

3. The method for detecting cyberbullying according to claim 2, characterized in that, The first feature data for extracting the first information includes: The first information is segmented to obtain word groups corresponding to the first information, and word feature data corresponding to each word in the word group is determined; wherein, the word group includes several words; Based on the word feature data corresponding to each word in the phrase, determine the first feature data corresponding to the first information.

4. The method for detecting cyberbullying according to claim 3, characterized in that, The word feature data includes word embedding vectors; determining the word feature data corresponding to each word in the word group includes: Map the words to a vector space to obtain the word embedding vectors corresponding to the words; The step of determining the first feature data corresponding to the first information based on the word feature data corresponding to each word in the phrase includes: The word embedding vectors corresponding to each word in the phrase are summed to obtain the first vector; The first vector is normalized to obtain the first feature data.

5. The method for detecting cyberbullying according to claim 1, characterized in that, The step of determining whether the target object's emotion has shifted towards a negative shift based on the first sentiment analysis result and the third sentiment analysis result includes: If the first sentiment analysis result is positive and the third sentiment analysis result is negative, it is determined that the target object's emotion has shifted to a negative direction.

6. The method for detecting cyberbullying according to claim 1, characterized in that, The step of determining whether the target object's emotion has shifted towards a negative shift based on the first sentiment analysis result and the third sentiment analysis result includes: If both the first sentiment analysis result and the third sentiment analysis result are negative, obtain the first negative probability corresponding to the first sentiment analysis result and the second negative probability corresponding to the third sentiment analysis result; Calculate the difference between the second negative probability and the first negative probability; If the difference is greater than a first preset threshold, it is determined that the target object's emotion has shifted to a negative direction.

7. The method for detecting cyberbullying according to claim 1, characterized in that, The method further includes: Based on the second sentiment analysis results, calculate the proportion of the second information that is negative information; If the ratio is greater than the second preset threshold, the second sentiment analysis result is determined to be negative.

8. A system for detecting cyberbullying, characterized in that, The system includes: The first acquisition module is used to acquire the first information published by the target object in response to the target event; The first analysis module is used to input the first information into the sentiment analysis model to perform sentiment analysis and obtain the first sentiment analysis result corresponding to the target object; The second acquisition module is used to acquire, when it is determined that other objects have commented on or replied to the first information, second information of the other objects' comments and replies, and third information published by the target object; wherein, the third information is information republished by the target object in response to the target event or information in response to the second information by the target object; The second analysis module is used to input each of the second information and the third information into the sentiment analysis model to perform sentiment analysis, and obtain the second sentiment analysis result corresponding to the other objects and the third sentiment analysis result corresponding to the target object. The comparison module is used to determine whether the target object's emotion has shifted to a negative direction based on the first sentiment analysis result and the third sentiment analysis result; The processing module is used to determine that the other objects have engaged in cyberbullying behavior when the second sentiment analysis result is negative and the target object's sentiment shifts to negative.

9. A terminal device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a method for detecting cyberbullying as described in any one of claims 1-7.

10. A computer-readable storage medium storing a processor-executable program, characterized in that: The processor-executable program, when executed by the processor, is used to implement a method for detecting cyberbullying as described in any one of claims 1-7.

Images