Data processing methods, devices, equipment, media and products

Abstract

This application discloses a data processing method, apparatus, device, medium, and product. The method includes: obtaining a complaint message text; determining at least one candidate website name corresponding to the complaint message text; and determining whether each of the at least one candidate website name is an illegal website name. This allows accurate identification of illegal website names from the complaint message text.

Description

Technical Field

[0001] The present application belongs to the field of data processing technology, and in particular relates to a data processing method, device, equipment, medium and product. Background Art

[0002] With the development and popularization of the Internet, illegal websites have emerged one after another. Such websites are exploited by criminals, causing economic losses and life troubles to Internet users, and have become one of the factors of social instability.

[0003] After discovering an illegal website or suffering losses from it, users often file a complaint with the service provider. Complaints are typically in natural language. After obtaining the complaint text, a computer is typically used to determine the website's Uniform Resource Locator (URL) from the complaint text. However, users may enter the website name instead of the URL in the complaint text. Existing methods cannot accurately determine the illegal website name from the complaint text.

[0004] Therefore, there is an urgent need for a method that can accurately determine the names of illegal websites from complaint information text. Summary of the Invention

[0005] The embodiments of the present application provide a data processing method, apparatus, device, medium, and product that can accurately determine the name of an illegal website from a complaint information text.

[0006] In a first aspect, an embodiment of the present application provides a data processing method, the method comprising:

[0007] Get complaint information text;

[0008] Determine at least one candidate website name corresponding to the complaint information text;

[0009] It is determined whether each candidate website name in the at least one candidate website name is an illegal website name.

[0010] In a second aspect, an embodiment of the present application provides a data processing device, the device comprising:

[0011] Acquisition module, used to obtain complaint information text;

[0012] A determination module, configured to determine at least one candidate website name corresponding to the complaint information text;

[0013] The judging module is used to judge whether each candidate website name in at least one candidate website name is an illegal website name.

[0014] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a processor and a memory storing computer program instructions;

[0015] When the processor executes the computer program instructions, the data processing method as shown in any one of the embodiments of the first aspect is implemented.

[0016] In a fourth aspect, an embodiment of the present application provides a computer storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the data processing method shown in any one of the embodiments of the first aspect is implemented.

[0017] In a fifth aspect, an embodiment of the present application provides a computer program product. When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the data processing method shown in any one of the embodiments of the first aspect.

[0018] The data processing method, apparatus, equipment, medium and product of the embodiments of the present application can obtain a complaint information text, determine at least one candidate website name corresponding to the complaint information text, and then determine whether each candidate website name in the at least one candidate website name is an illegal website name, so as to accurately determine the illegal website name from the complaint information text. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0020] Figure 1 This is a flow chart of a data processing method provided by one embodiment of the present application;

[0021] Figure 2 This is a schematic diagram of a complaint information graph network provided by an embodiment of the present application;

[0022] Figure 3 This is a structural diagram of a data processing device provided by an embodiment of the present application;

[0023] Figure 4 This is a schematic structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0024] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is merely to provide a better understanding of the present application by illustrating the examples of the present application.

[0025] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, the elements defined by the phrase "comprising..." do not exclude the presence of other identical elements in the process, method, article, or device comprising the elements.

[0026] As mentioned in the background, with the development and popularization of the internet, illegal websites have emerged in an endless stream. These websites are exploited by criminals, causing economic losses and life-threatening problems for a large number of internet users, becoming a factor in social instability. Internet users often make payments on illegal websites, either actively or passively, through third-party payment channels. Upon discovering losses, they typically file a complaint through the third-party payment channel, using natural language text to describe the process and / or outcome of the incident. The complaint text may contain the URL or website name of the illegal website. While extracting the website URL information using a computer is relatively easy, extracting and identifying the website name is more difficult. Furthermore, the appearance and disappearance of specific illegal websites is highly random.

[0027] In related technologies, there are three methods for determining website URL information or website name:

[0028] 1. Graph-based new word discovery: This method constructs a directed, weighted graph based on the adjacency relationships between words. By searching for paths in the graph structure to find the path with the highest weight, it preserves collocations with high co-occurrence frequencies in the corpus. However, because many new words appear infrequently, searching for the path with the highest weight in the graph structure is not effective for discovering low-frequency new words, especially for identifying illegal website names.

[0029] 2. URL-based feature recognition: This method uses URL features to construct an illegal website identification model. This method analyzes the structure and lexical characteristics of illegal website URLs, extracts sensitive features from the URLs, and constructs an illegal website classifier. Alternatively, it identifies illegal websites based on the Hamming distance between legitimate and illegal websites. However, because website URL features are easy to mimic and their number is limited, identifying illegal websites based solely on URL features presents significant uncertainty. Furthermore, this method is ineffective when only the website name is provided in the complaint text.

[0030] 3. Feature recognition based on web page content and structure: This method completes the identification of illegal websites by extracting effective features from the title, keywords, description information and other content of the web page. Features are extracted using the Head tag and Body tag in the Hypertext Markup Language (HTML) script of the page, such as extracting effective features such as keywords and icons in the title to identify illegal websites; or detecting the similarity of the spatial layout features of the page and using the correlation between them. When the similarity of the pages exceeds a threshold, it can be determined that the web page to be detected is an illegal website. This method needs to obtain the URL of the website and obtain the website content through crawler technology to perform feature recognition. Therefore, this method is invalid when only the website name is provided in the complaint information text.

[0031] Furthermore, graph-based new word discovery methods, which construct graph networks and traverse them to find the path with the highest weight, are not suitable for discovering low-frequency new words. Because illegal websites often appear suddenly and then disappear due to being blocked or banned, the frequency of illegal website names in complaint information is even lower. Therefore, existing methods cannot effectively identify illegal website names in complaint information.

[0032] Existing illegal website identification methods rely on model training and feature extraction based on existing illegal websites. With the development of web technology and network regulation, illegal websites are constantly changing, emerging and disappearing, and knowledge related to illegal websites also undergoes unpredictable changes over time. Existing identification technologies are unable to effectively adapt to the latest illegal websites, resulting in poor prediction and decision-making results for existing illegal website identification models when dealing with newly emerging illegal websites. Furthermore, both methods require the known website URL, making them inapplicable when the complaint text only contains the illegal website name.

[0033] Therefore, there is an urgent need for a method that can accurately determine the names of illegal websites from complaint information text.

[0034] The embodiments of the present application provide a data processing method, apparatus, device, medium and product, which can obtain complaint information text and determine at least one candidate website name corresponding to the complaint information text, and then determine whether each candidate website name in the at least one candidate website name is an illegal website name, so as to accurately determine the illegal website name from the complaint information text.

[0035] Figure 1 A flow chart of a data processing method provided by an embodiment of the present application is shown. It should be noted that the execution subject of the data processing method may be a data processing device, such as Figure 1 As shown, the data processing method may include the following steps:

[0036] S110, obtaining complaint information text;

[0037] S120, determining at least one candidate website name corresponding to the complaint information text;

[0038] S130: Determine whether each candidate website name in at least one candidate website name is an illegal website name.

[0039] In this way, the complaint information text can be obtained, and at least one candidate website name corresponding to the complaint information text can be determined, and then it can be judged whether each candidate website name in the at least one candidate website name is an illegal website name, so that the illegal website name can be accurately determined from the complaint information text.

[0040] Regarding S110 , the complaint information text may be a natural language text input by the user.

[0041] Regarding S120 , the candidate website name may be a word that is suspected to be an illegal website name.

[0042] In some implementations, to obtain more accurate candidate website names, the above S120 may include S121-S124, as follows:

[0043] S121, deleting the first segmented word that is not a website name in the first segmented word sequence corresponding to the complaint information text, to obtain multiple first segmented word segments;

[0044] S122, respectively determining target segmentation nodes corresponding to the multiple target segmentations included in each first segmentation segment in the complaint information graph network and target edges between different target segmentation nodes;

[0045] S123, when the target edge meets the preset conditions, deleting the target segmentation corresponding to the target segmentation node connected to the target edge to obtain multiple second segmentation fragments;

[0046] S124: Determine at least one candidate website name based on the plurality of second word segmentation segments.

[0047] In step S121, the first segmented word that is not a website name in the first segmented word sequence corresponding to the complaint information text of the illegal website complaint category can be deleted to obtain multiple first segmented word segments. Specifically, the complaint information text of the illegal website complaint category can be identified by a machine or directly input manually.

[0048] For complaint information texts that are not classified as illegal website complaints, they generally do not contain the names of illegal websites and do not require subsequent processing.

[0049] The first participle may be any participle in the first participle sequence. The first participle that is not a website name may be the first participle with the target part of speech or the target syntax.

[0050] In some implementations, the above S121 may include:

[0051] The first participle whose part of speech is the target part of speech or whose syntax is the target syntax is deleted from the first participle sequence to obtain multiple first participle fragments.

[0052] Here, the target part of speech may be a part of speech other than the website name, and the target syntax may be a syntax other than the website name.

[0053] In some implementations, the target part of speech may include verbs, prepositions, time words, non-morpheme words, and pronouns, and the target syntax may include predicates and adverbials.

[0054] Of course, the target part of speech may also include other parts of speech, and the target syntax may also include other syntax, which are not limited here.

[0055] In some implementations, to obtain a more accurate first word segmentation sequence, before S121, the method may further include:

[0056] The complaint information text is segmented to obtain a first segmentation sequence corresponding to the complaint information text.

[0057] Here, the complaint information text input by the user may be obtained first, and then the complaint information text may be segmented to obtain a first segmentation sequence corresponding to the complaint information text.

[0058] In addition, after performing word segmentation processing on the complaint information text to obtain an initial word segmentation sequence, a start word segmentation and an end word segmentation can be added to the initial word segmentation sequence to obtain a first word segmentation sequence.

[0059] For example, the initial word segmentation sequence TS=(T1, T2, ..., T n ), add T0 = [starting participle] and T n+1 =[end segmentation] to TS, and get TS=(T0,T1,T2,…,Tn ,T n+1 ), T i is the word segmentation, i=0,1,2,…,n,n+1.

[0060] For example, the initial word segmentation sequence may be as shown in Table 1.

[0061] Table 1 - Initial word segmentation sequence table

[0062]

[0063] In this way, the first word segmentation sequence can be obtained more accurately by performing word segmentation processing on the complaint information text.

[0064] In some implementations, in order to more accurately determine the complaint category corresponding to the complaint information text, before S121, the method may further include:

[0065] The complaint information text is input into the complaint information text classification model, the complaint information text is classified, and the complaint category corresponding to the complaint information text is output.

[0066] Here, a large number of complaint information texts can be pre-labeled with labels such as "target complaint" and "general complaint." The labeled complaint information texts are used to train a convolutional neural network classification model to obtain a complaint information text classification model.

[0067] In this way, the complaint category corresponding to the complaint information text can be quickly and accurately determined through the pre-trained complaint information text classification model.

[0068] In some implementations, to more accurately determine the part of speech and syntax of each first word in the first word segmentation sequence, before S121, the method may further include:

[0069] The part of speech of each first word in the first word sequence is tagged by a part-of-speech tagging algorithm, and the syntax of each first word in the first word sequence is tagged by a natural language processing algorithm.

[0070] Illustratively, the part of speech of each first segmented word in the first segmented word sequence may be as shown in Table 2.

[0071] Table 2 - Part-of-speech tagging table

[0072]

[0073] Among them, n is a name, v is a verb, p is a preposition, m is a numeral, vn is a noun verb, x is a non-morpheme word, t is a time word, zg is a state word, ul is a tense auxiliary word, nz is other proper nouns, nr is a personal name, and r is a pronoun.

[0074] For example, for each first word T i Add a deletable tag and initialize the "delete" tag to "unmarked", such as: T i = (word text, part of speech, deletable tag), initialized to T i (EF, n, unmarked). If T i If the part of speech of belongs to the target part of speech, then update the "deletable mark" to "delete". If T i If the syntax of belongs to the target syntax, then the “deletable mark” is updated to “delete”. The first word sequence after updating the “deletable mark” can be shown in Table 3. Then, each first word T can be determined i If the "delete mark" is "delete", then the T i Replace with a separator, so that the T i , get multiple consecutive first word segment segments TS1 ′ ,…,TS ′ m . m can be the number of first word segmentation fragments.

[0075] Table 3 - Table of deletable tags for multiple first word segments

[0076]

[0077] Regarding S122, the complaint information graph network can be obtained by adding the first word segmentation sequence to the preset graph network. The complaint information graph network may include multiple nodes and multiple edges. The multiple nodes may include word segmentation nodes. Each first word segmentation in the first word segmentation sequence may correspond to a word segmentation node. Each edge may be used to connect two adjacent word segmentation nodes in the word segmentation sequence. Each edge may correspond to an edge information. The edge information may include the first frequency of the edge being used and the time when the edge was created.

[0078] In some implementations, in order to obtain a more accurate complaint information graph network, before S122, the method may further include:

[0079] For each first participle in the first participle sequence, determining whether there is a first participle node corresponding to the first participle in the preset graph network;

[0080] When the first word segmentation node exists in the preset graph network, the second frequency corresponding to the first word segmentation node is increased by 1;

[0081] If the first word segmentation node does not exist in the preset graph network, create a first word segmentation node corresponding to the first word segmentation node in the preset graph network, and set the second frequency corresponding to the first word segmentation node to 1;

[0082] For every two adjacent first participles in the first participle sequence, determining whether there is a first edge connecting the participle nodes corresponding to the two adjacent first participles in the preset graph network;

[0083] If there is a first edge connected in the preset graph network, the first frequency corresponding to the first edge is increased by 1;

[0084] When there is no first connecting edge in the preset graph network, a first edge is created between the segmentation nodes corresponding to two adjacent first segmentations in the preset graph network, and the first frequency corresponding to the first edge is set to 1.

[0085] Here, each node in the preset graph network can correspond to a node information, and the node information can include the second frequency of the node's use. The preset graph network can be a graph network to which word segmentation sequences corresponding to other complaint information texts have been added. The first word segmentation sequences corresponding to complaint information texts of any complaint category can be added to the preset graph network to obtain a complaint information graph network.

[0086] Specifically, for each first word segmentation in the first word segmentation sequence, we can first determine whether there is a first word segmentation node corresponding to the first word segmentation in the preset graph network. If so, we can add 1 to the second frequency corresponding to the first word segmentation node already existing in the preset graph network, without creating the first word segmentation node again in the preset graph network; if not, we need to create a first word segmentation node corresponding to the first word segmentation in the preset graph network, and set the second frequency corresponding to the first word segmentation node to 1.

[0087] For every two adjacent first participles in the first participle sequence, it is possible to first determine whether there is a first edge in the preset graph network connecting the participle nodes corresponding to the two adjacent first participles. If so, the first frequency corresponding to the first edge already existing in the preset graph network can be increased by 1, without the need to create the first edge again in the preset graph network. If not, it is necessary to create the first edge in the preset graph network. Specifically, the first edge can be created between the two participle nodes corresponding to the above-mentioned two adjacent first participles in the preset graph network, and the first frequency corresponding to the first edge needs to be set to 1.

[0088] After the above process, the first word segmentation sequence can be added to the preset graph network to obtain the complaint information graph network.

[0089] In some implementations, a preset graph network G(V, E) may be predefined.

[0090] Node V = (label, name, frequency), where: the label can be the start node / end node / word node / punctuation node, the name can be [start node] / [end node] / word text / punctuation text, and the frequency can be the number of times the node V is used in the preset graph network G.

[0091] Edge E = (frequency, creation time, last update time), where frequency can be the number of times edge E is used in the preset graph network G, creation time can be the time when edge E is created in the preset graph network G, and last update time can be the time when the frequency of edge E is last updated so far.

[0092] Then, the first word segmentation sequence can be added to the preset graph network.

[0093] Specifically, each word in the first word segmentation sequence can be added to the preset graph network by the following method:

[0094] Construct word bigrams (T i ,T i+1 );

[0095] Use left and right word query nodes: For word bigrams (T i ,T i+1 ), for each word in, if its corresponding V does not exist, create V = (label, name, frequency = 1), such as: V (word node, CD, 1); otherwise update V = (label, name, frequency = frequency + 1), such as: V (word node, EF, 5 + 1);

[0096] Use left and right nodes to query edges: For word bigrams (T i ,T i+1 ) If E between the corresponding V does not exist, create E = (frequency = 1, creation time = current time, last update time = current time), such as: E(1, 2021 / 11 / 210:10:30, 2021 / 11 / 2 10:10:30); otherwise update E = (frequency = frequency + 1, creation time, last update time = current time), such as: E(8+1, 2021 / 10 / 29 17:45:21, 2021 / 11 / 2 10:10:30).

[0097] The word segmentation nodes in the embodiment of the present application may include the above-mentioned word nodes and / or punctuation nodes, and the word segmentation may include word text and / or punctuation text.

[0098] For example, the complaint information graph network obtained by adding the first word segmentation sequences corresponding to the first two complaint information texts in Table 1 to the preset graph network can be as follows: Figure 2 As shown in Figure 2, the complaint information graph network includes edges pointing from the start node to the end node.

[0099] In this way, through the above process, the first word segmentation sequence can be added to the preset graph network to obtain a more accurate complaint information graph network.

[0100] Regarding S123, if the edge information corresponding to the target edge satisfies a preset condition, the target segmentation corresponding to the target segmentation node connected to the target edge is deleted from the first segmentation segment to obtain multiple second segmentation segments. The preset condition may be that the first frequency of use of the target edge is not less than a preset frequency or that the target edge is created no later than a preset time. The preset frequency and preset time can be set according to actual needs and are not limited here.

[0101] That is to say, for each target edge, if the target edge is used frequently or has been created for a long time, it means that the target segmentations corresponding to the two target segmentation nodes connected by the target edge are not illegal website names, so these two target segmentations can be deleted.

[0102] For example, the plurality of first word segmentation segments include words whose “deletable mark” is “unmarked”. Each two adjacent target words included in each first word segmentation segment are respectively taken as word bigrams (T i ,T i+1 ), if T i and T i+1 If the first frequency of the target edge between the corresponding word segmentation nodes is less than the preset frequency and the creation time of the target edge is later than the preset time, then T i and T i+1 Update the "deletable mark" to "retain", otherwise T i and T i+1 The “deletable mark” of is updated to “delete”, and the multiple first segmentation fragments after updating the “deletable mark” can be shown in Table 4. Then, each target segmentation T can be determined i If the "delete mark" is "delete", then the T i Replace with a separator, so that the T i , get multiple consecutive second word segment segments TS″1,…,TS″ p . p can be the number of second word segment fragments.

[0103] Table 4 - Table of deletable tags for multiple second word segments

[0104]

[0105] In S124 , a set of candidate website names may be determined based on the multiple second word segmentation segments corresponding to each complaint information text, which may include at least one candidate website name.

[0106] In some implementations, in order to obtain a more accurate candidate website name, the above S124 may include:

[0107] The plurality of second word segmentation segments are processed through a sliding window to obtain at least one candidate website name corresponding to the complaint information text.

[0108] Here, the sliding window can be set according to actual needs and is not limited here.

[0109] For example, the size of the sliding window may be 1, 2, or 3. By processing the deletable words marked as "reserved" in Table 4 through the sliding window, the obtained candidate website name set may be as shown in Table 5.

[0110] Table 5 - Candidate website name set

[0111]

[0112] In this way, a more accurate candidate website name can be obtained by using a sliding window.

[0113] Thus, through the above process, a more accurate candidate website name can be obtained.

[0114] In step S130 , it may be determined whether each candidate website name in the at least one candidate website name is an illegal website name, thereby determining the illegal website name in the complaint information text.

[0115] In some implementations, in order to more accurately determine whether each candidate website name in the at least one candidate website name is an illegal website name, the above S130 may include:

[0116] Input each candidate website name into the website name discrimination model respectively, determine the probability that each candidate website name is a website name, and output the target probability corresponding to each candidate website name;

[0117] The target candidate website name corresponding to the first probability is determined to be an illegal website name.

[0118] The target probability can be used to represent the probability that the candidate website name is a website name. The first probability can be the highest target probability corresponding to at least one candidate website name and higher than a preset probability. The preset probability can be set according to actual needs and is not limited here.

[0119] Here, a large amount of website name corpus can be annotated in advance with labels such as "is a website name" and "is not a website name". The annotated text is used to train a convolutional neural network classification model to obtain a website name discrimination model.

[0120] In this way, the website name discrimination model can be used to more accurately determine whether each candidate website name in at least one candidate website name is an illegal website name.

[0121] For example, the illegal website names included in the complaint information text may be as shown in Table 6.

[0122] Table 6- List of illegal websites

[0123]

[0124] The data processing method provided in the embodiments of this application can convert the first word segmentation sequence corresponding to the complaint information text into a graph network, generate word segmentation sequence segments using the characteristics of Chinese parts of speech and syntactic structure, then use graph calculations to generate word segmentation sequence segments, and finally use sliding window technology to obtain a set of valid candidate website names. This method extracts website names based on parts of speech, syntactic structure, and graph calculations. Compared with other statistical-based methods, it improves the accuracy and computational efficiency of the website name recognition model, making the identification of illegal website names more accurate.

[0125] Based on the same inventive concept, the present application embodiment also provides a data processing device. Figure 3 The data processing device provided in the embodiments of the present application is described in detail.

[0126] Figure 3 A structural diagram of a data processing device provided by an embodiment of the present application is shown.

[0127] like Figure 3 As shown, the data processing device may include:

[0128] Acquisition module 301, used to obtain complaint information text;

[0129] Determination module 302, for determining at least one candidate website name corresponding to the complaint information text;

[0130] The judging module 303 is configured to judge whether each candidate website name in at least one candidate website name is an illegal website name.

[0131] In this way, the complaint information text can be obtained, and at least one candidate website name corresponding to the complaint information text can be determined, and then it can be judged whether each candidate website name in the at least one candidate website name is an illegal website name, so that the illegal website name can be accurately determined from the complaint information text.

[0132] In some implementations, to obtain a more accurate candidate website name, the determination module 302 may include:

[0133] A first deletion submodule is configured to delete the first segmented word that is not a website name in the first segmented word sequence corresponding to the complaint information text, thereby obtaining a plurality of first segmented word segments;

[0134] The first determination submodule is used to respectively determine the target segmentation nodes corresponding to the multiple target segmentations included in each first segmentation segment in the complaint information graph network and the target edges between different target segmentation nodes;

[0135] A second deletion submodule is configured to delete the target word corresponding to the target word node connected to the target edge when the target edge meets a preset condition, thereby obtaining multiple second word fragments. The preset condition is that the first frequency of use of the target edge is not less than a preset frequency or the creation time of the target edge is not later than a preset time;

[0136] The second determining submodule is configured to determine at least one candidate website name according to the plurality of second word segmentation segments.

[0137] In some embodiments, in order to more accurately determine the complaint category corresponding to the complaint information text, the apparatus may further include:

[0138] The input module is used to input the complaint information text into the complaint information text classification model before deleting the first word that is not the website name in the first word sequence corresponding to the complaint information text to obtain multiple first word fragments, classify the complaint information text, and output the complaint category corresponding to the complaint information text.

[0139] In some implementations, the first deletion submodule may be specifically configured to:

[0140] The first participle whose part of speech is the target part of speech or whose syntax is the target syntax is deleted from the first participle sequence to obtain multiple first participle fragments, where the target part of speech is the part of speech corresponding to the non-website name and the target syntax is the syntax corresponding to the non-website name.

[0141] In some implementations, the target part of speech may include verbs, prepositions, time words, non-morpheme words, and pronouns, and the target syntax may include predicates and adverbials.

[0142] In some embodiments, in order to obtain a more accurate complaint information graph network, the apparatus may further include:

[0143] A third determination submodule is configured to determine, for each first segmentation in the first segmentation sequence, whether there is a first segmentation node corresponding to the first segmentation in the preset graph network before respectively determining target segmentation nodes corresponding to multiple target segmentations included in each first segmentation segment in the complaint information graph network and target edges between different target segmentation nodes;

[0144] A first counting submodule, configured to, when a first word segmentation node exists in the preset graph network, add 1 to the second frequency corresponding to the first word segmentation node;

[0145] A first creation submodule is configured to create a first word segmentation node corresponding to the first word segmentation in the preset graph network if the first word segmentation node does not exist in the preset graph network, and set the second frequency corresponding to the first word segmentation node to 1;

[0146] A fourth determining submodule is configured to determine, for each two adjacent first participles in the first participle sequence, whether there is a first edge in the preset graph network connecting the participle nodes corresponding to the two adjacent first participles;

[0147] A second counting submodule, configured to increase a first frequency corresponding to the first edge by 1 when there is a first edge connected in the preset graph network;

[0148] The second creation submodule is used to create a first edge between the segmentation nodes corresponding to two adjacent first segmentations in the preset graph network when the first edge does not exist in the preset graph network, and set the first frequency corresponding to the first edge to 1.

[0149] In some implementations, in order to obtain a more accurate candidate website name, the second determination submodule may be specifically configured to:

[0150] The plurality of second word segmentation segments are processed through a sliding window to obtain at least one candidate website name corresponding to the complaint information text.

[0151] In some implementations, in order to more accurately determine whether each candidate website name in the at least one candidate website name is an illegal website name, the determination module 303 may include:

[0152] The input submodule is used to input each candidate website name into the website name discrimination model, determine the probability that each candidate website name is a website name, and output the target probability corresponding to each candidate website name. The target probability is used to represent the probability that the candidate website name is a website name;

[0153] A fifth determining submodule is configured to determine that a target candidate website name corresponding to a first probability is an illegal website name, where the first probability is the highest among target probabilities corresponding to at least one candidate website name and is higher than a preset probability.

[0154] Figure 4 A schematic structural diagram of an electronic device provided by an embodiment of the present application is shown.

[0155] like Figure 4 As shown, the electronic device 4 is a block diagram of an exemplary hardware architecture of an electronic device that can implement the data processing method and data processing device according to the embodiments of the present application. The electronic device can refer to the electronic device in the embodiments of the present application.

[0156] The electronic device 4 may include a processor 401 and a memory 402 storing computer program instructions.

[0157] Specifically, the processor 401 may include a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0158] Memory 402 may include a large-capacity memory for data or instructions. By way of example and not limitation, memory 402 may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 402 may include removable or non-removable (or fixed) media. Where appropriate, memory 402 may be internal or external to the integrated gateway disaster recovery device. In certain embodiments, memory 402 is non-volatile solid-state memory. In certain embodiments, memory 402 may include read-only memory (ROM), random access memory (RAM), a magnetic disk storage media device, an optical storage media device, a flash memory device, an electrical, optical, or other physical / tangible memory storage device. Therefore, generally, memory 402 includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of the present application.

[0159] The processor 401 implements any one of the data processing methods in the above embodiments by reading and executing computer program instructions stored in the memory 402 .

[0160] In one example, the electronic device may further include a communication interface 403 and a bus 404. Figure 4 As shown, the processor 401 , the memory 402 , and the communication interface 403 are connected via a bus 404 and communicate with each other.

[0161] The communication interface 403 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0162] Bus 404 comprises hardware, software or both, couples the parts of electronic equipment to each other.For example, and not limitation, bus can comprise accelerated graphics port (AGP) or other graphics bus, enhanced industry standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industry standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations.In suitable cases, bus 404 can comprise one or more buses.Although the present application embodiment describes and shows specific bus, the application considers any suitable bus or interconnection.

[0163] The electronic device can execute the data processing method in the embodiment of the present application, thereby realizing the combination Figures 1 to 3 Described data processing method and device.

[0164] In addition, in conjunction with the data processing methods in the above embodiments, embodiments of the present application may provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any one of the data processing methods in the above embodiments is implemented.

[0165] It should be understood that the present application is not limited to the specific configurations and processes described above and illustrated in the figures. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and illustrated as examples. However, the method process of the present application is not limited to the specific steps described and illustrated. Those skilled in the art can make various changes, modifications, and additions, or change the order of the steps after understanding the spirit of the present application.

[0166] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium that can store or transmit information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memories, erasable ROMs (EROMs), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0167] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps. In other words, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0168] Aspects of the present application have been described above with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed via the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. This processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit. It is also understood that each box in the block diagram and / or the flowchart and the combination of the boxes in the block diagram and / or the flowchart can also be implemented by the dedicated hardware that performs the specified function or action, or can be implemented by the combination of dedicated hardware and computer instructions.

[0169] The above description is only a specific embodiment of the present application. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the scope of protection of the present application.

Claims

1. A data processing method, characterized in that: include: Get complaint information text; Determining at least one candidate website name corresponding to the complaint information text; Determining whether each of the at least one candidate website name is an illegal website name; and determining at least one candidate website name corresponding to the complaint information text includes: Deleting the first segmented word that is not a website name in the first segmented word sequence corresponding to the complaint information text to obtain multiple first segmented word segments; Respectively determine target segmentation nodes corresponding to the multiple target segmentations included in each of the first segmentation segments in the complaint information graph network and target edges between different target segmentation nodes; When the target edge meets a preset condition, deleting the target segmentation corresponding to the target segmentation node connected to the target edge to obtain multiple second segmentation fragments, wherein the preset condition is that the first frequency of use of the target edge is not less than a preset frequency, or the creation time of the target edge is not later than a preset time; The at least one candidate website name is determined according to the multiple second word segmentation segments.

2. The method according to claim 1, characterized in that Before deleting the first segmented word that is not a website name in the first segmented word sequence corresponding to the complaint information text to obtain a plurality of first segmented word segments, the method further includes: The complaint information text is input into a complaint information text classification model, the complaint information text is classified, and the complaint category corresponding to the complaint information text is output.

3. The method according to claim 1, characterized in that The first segmented word that is not a website name in the first segmented word sequence corresponding to the complaint information text is deleted to obtain multiple first segmented word segments, including: The first word segmentation whose part of speech is the target part of speech or whose syntax is the target syntax is deleted from the first word segmentation sequence to obtain multiple first word segmentation fragments, wherein the target part of speech is the part of speech corresponding to the non-website name, and the target syntax is the syntax corresponding to the non-website name.

4. The method according to claim 3, characterized in that The target parts of speech include verbs, prepositions, time words, non-morpheme words and pronouns, and the target syntax includes predicates and adverbials.

5. The method according to claim 1, characterized in that Before respectively determining target segmentation nodes corresponding to the multiple target segmentations included in each of the first segmentation segments in the complaint information graph network and target edges between different target segmentation nodes, the method further includes: For each first participle in the first participle sequence, determining whether there is a first participle node corresponding to the first participle in a preset graph network; If the first word segmentation node exists in the preset graph network, adding 1 to the second frequency corresponding to the first word segmentation node; If the first word segmentation node does not exist in the preset graph network, create a first word segmentation node corresponding to the first word segmentation in the preset graph network, and set the second frequency corresponding to the first word segmentation node to 1; For every two adjacent first participles in the first participle sequence, determining whether there is a first edge in the preset graph network connecting the participle nodes corresponding to the two adjacent first participles; If there is a connection to the first edge in the preset graph network, adding 1 to the first frequency corresponding to the first edge; When the first edge does not exist in the preset graph network, the first edge is created between the segmentation nodes corresponding to the two adjacent first segmentations in the preset graph network, and the first frequency corresponding to the first edge is set to 1.

6. The method according to claim 1, characterized in that Determining the at least one candidate website name according to the plurality of second word segmentation segments includes: The multiple second word segmentation segments are processed through a sliding window to obtain the at least one candidate website name corresponding to the complaint information text.

7. The method according to claim 1, characterized in that The determining whether each of the at least one candidate website name is an illegal website name includes: Input each candidate website name into the website name discrimination model, determine the probability that each candidate website name is a website name, and output a target probability corresponding to each candidate website name, where the target probability is used to represent the probability that the candidate website name is a website name; Determine that a target candidate website name corresponding to a first probability is an illegal website name, where the first probability is the highest among the target probabilities corresponding to the at least one candidate website name and is higher than a preset probability.

8. A data processing device, characterized in that: The device comprises: Acquisition module, used to obtain complaint information text; a determination module, configured to determine at least one candidate website name corresponding to the complaint information text; A judging module, configured to judge whether each of the at least one candidate website name is an illegal website name; The determination module is further configured to determine at least one candidate website name corresponding to the complaint information text, including: Deleting the first segmented word that is not a website name in the first segmented word sequence corresponding to the complaint information text to obtain multiple first segmented word segments; Respectively determine target segmentation nodes corresponding to the multiple target segmentations included in each of the first segmentation segments in the complaint information graph network and target edges between different target segmentation nodes; When the target edge meets a preset condition, deleting the target segmentation corresponding to the target segmentation node connected to the target edge to obtain multiple second segmentation fragments, wherein the preset condition is that the first frequency of use of the target edge is not less than a preset frequency, or the creation time of the target edge is not later than a preset time; The at least one candidate website name is determined according to the multiple second word segmentation segments.

9. An electronic device, characterized in that: The device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the data processing method according to any one of claims 1 to 7 is implemented.

10. A computer storage medium, characterized in that The computer storage medium stores computer program instructions, which, when executed by a processor, implement the data processing method according to any one of claims 1 to 7.

11. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the data processing method according to any one of claims 1 to 7.

Images