Word cloud processing method, device and system

Abstract

The present disclosure relates to a word cloud processing method, device and system. The word cloud processing method comprises: obtaining first frequency data of words in a word cloud within a first period and second frequency data of words within multiple second periods, wherein the first period refers to a complete period before the current time, and the multiple second periods refer to a predetermined number of complete periods before the first period; inputting the second frequency data into a frequency prediction model to obtain predicted frequency data of words within the first period; obtaining abnormal difference values ​​of words based on the predicted frequency data and the first frequency data; and performing abnormal processing when the abnormal difference value is greater than a preset alarm threshold.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing, and in particular to a word cloud processing method, device and system. Background Art

[0002] Word cloud is a very important natural language processing technology. As a major method and tool for visualizing text content, it can concisely visualize the main content and information contained in a text set through keywords and keyword size. Figure 1 An example of a word cloud is given.

[0003] At present, word cloud processing mainly involves business personnel checking the words in the word cloud to determine whether there are any anomalies. However, this word cloud processing method has the following disadvantages:

[0004] 1) High labor costs. The judgment of abnormal situations depends entirely on the experience of business personnel. Because user needs are very broad, manual judgment requires a lot of work and is prone to errors;

[0005] 2) Untimely alarms. The word cloud is generally a statistics of user needs in the past period of time, such as 10 minutes, 30 minutes, etc. When abnormal problems are manually discovered, the abnormalities have already existed for a period of time, resulting in untimely problem solving. Summary of the invention

[0006] The present disclosure provides a word cloud processing method, device and system to at least solve the problems of high labor cost and untimely alarm in the word cloud processing method in the related art.

[0007] According to a first aspect of an embodiment of the present disclosure, a word cloud processing method is provided, comprising: obtaining first frequency data of words in a word cloud within a first period and second frequency data of words within multiple second periods, wherein the first period refers to a complete period before a current time, and the multiple second periods refer to a predetermined number of complete periods before the first period; inputting the second frequency data into a frequency prediction model to obtain predicted frequency data of the words within the first period; obtaining abnormal difference values ​​of the words based on the predicted frequency data and the first frequency data; and performing abnormal processing when the abnormal difference value is greater than a preset alarm threshold.

[0008] Optionally, when the abnormal difference is greater than a preset alarm threshold, before performing abnormal processing, it also includes: fitting the second frequency data and the first frequency data to obtain the fitted frequency; obtaining the residual frequency based on the second frequency data, the first true frequency and the fitted frequency; obtaining the normal fluctuation value based on the standard deviation of the residual frequency; and obtaining the preset alarm threshold of the word based on the normal fluctuation value and a preset multiplier parameter.

[0009] Optionally, before obtaining the first frequency data of the words in the word cloud in the first period and the second frequency data of the words in multiple second periods, it also includes: storing the frequency data of the words in corresponding storage units, wherein each word in the word cloud corresponds to a storage unit, and the frequency data includes the first frequency data and the second frequency data.

[0010] Optionally, inputting the second frequency data into the frequency prediction model to obtain the predicted frequency data of the word in the first period includes: inputting the second frequency data in the storage unit corresponding to the word into the frequency prediction model to obtain the predicted frequency data of the word in the first period.

[0011] Optionally, when the abnormal difference is greater than a preset alarm threshold, after performing abnormal processing, it also includes: receiving a word segmentation processing result sent by the word segmentation module; and updating a storage unit corresponding to the word based on the word segmentation processing result.

[0012] Optionally, updating the storage unit corresponding to the word includes: storing the frequency data of the word in the new period in the word segmentation processing result into the storage unit corresponding to the word, and when the length of the storage unit exceeds the preset maximum length, deleting the frequency data of the word in the front period in the storage unit in sequence until the length of the storage unit meets the preset maximum length.

[0013] According to a second aspect of an embodiment of the present disclosure, a word cloud processing device is provided, comprising: a frequency data acquisition unit, configured to acquire first frequency data of words in a word cloud within a first period and second frequency data of words within multiple second periods, wherein the first period refers to a complete period before a current time, and the multiple second periods refer to a predetermined number of complete periods before the first period; a predicted frequency data acquisition unit, configured to input the second frequency data into a frequency prediction model to obtain predicted frequency data of words within the first period; an abnormal difference acquisition unit, configured to obtain abnormal differences of words based on the predicted frequency data and the first frequency data; and a processing unit, configured to perform abnormal processing when the abnormal difference is greater than a preset alarm threshold.

[0014] Optionally, the processing unit is also configured to fit the second frequency data and the first frequency data to obtain a fitted frequency before performing exception processing when the abnormal difference is greater than a preset alarm threshold; obtain a residual frequency based on the second frequency data, the first frequency data and the fitted frequency data; obtain a normal fluctuation value based on the standard deviation of the residual frequency; and obtain a preset alarm threshold for the word based on the normal fluctuation value and a preset multiplier parameter.

[0015] Optionally, the frequency data acquisition unit is also configured to store the frequency data of the words in the corresponding storage unit before acquiring the first frequency data of the words in the word cloud in the first period and the second frequency data of the words in multiple second periods, wherein each word in the word cloud corresponds to a storage unit, and the frequency data includes the first frequency data and the second frequency data.

[0016] Optionally, the predicted frequency data acquisition unit is further configured to input the second frequency data in the storage unit corresponding to the word into the frequency prediction model to obtain the predicted frequency data of the word in the first period.

[0017] Optionally, the processing unit is further configured to receive the word segmentation processing result sent by the word segmentation module after performing exception processing when the abnormal difference is greater than a preset alarm threshold; and update the storage unit corresponding to the word based on the word segmentation processing result.

[0018] Optionally, the processing unit is also configured to store the frequency data of the new period words in the word segmentation processing result into the storage unit corresponding to the word, and when the length of the storage unit exceeds the preset maximum length, delete the frequency data of the words in the front period in the storage unit in sequence until the length of the storage unit meets the preset maximum length.

[0019] According to a third aspect of an embodiment of the present disclosure, a word cloud processing system is provided, including: a word segmentation module and a word cloud processing module, the word segmentation module being configured to receive a demand text fed back by a user and perform word segmentation processing on the demand text; the word cloud processing module being configured to execute the above word cloud processing method.

[0020] According to a fourth aspect of an embodiment of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing processor executable instructions; wherein the processor is configured to execute instructions to implement a word cloud processing method according to the present disclosure.

[0021] According to a fifth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor is prompted to execute the word cloud processing method according to the present disclosure as described above.

[0022] According to a sixth aspect of an embodiment of the present disclosure, a computer program product is provided, comprising computer instructions, which implement the word cloud processing method according to the present disclosure when the computer instructions are executed by a processor.

[0023] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0024] According to the word cloud processing method, device and system disclosed in the present invention, the predicted frequency data of the words in the first period before the current time are predicted by the frequency prediction model, and then based on the difference between the predicted frequency data and the first frequency data, the abnormality processing is performed in combination with the preset alarm threshold, which can realize automatic alarm, without the participation of business personnel, saving labor costs, and compared with the manual discovery of abnormalities in the related art, the present invention can also shorten the discovery of abnormalities through automatic alarm, and can timely alarm processing. Therefore, the present invention solves the problems of high labor costs and untimely alarms in the word cloud processing method in the related art.

[0025] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0027] Figure 1 is a schematic diagram showing a word cloud in the related art;

[0028] Figure 2 is a schematic diagram showing an implementation scenario of a word cloud processing method according to an exemplary embodiment of the present disclosure;

[0029] Figure 3 is a flow chart of a word cloud processing method according to an exemplary embodiment;

[0030] Figure 4 is a schematic diagram of a bidirectional queue according to an exemplary embodiment;

[0031] Figure 5 is a workflow diagram of a word abnormality alarm according to an exemplary embodiment;

[0032] Figure 6 is a data flow diagram of a word abnormality alarm according to an exemplary embodiment;

[0033] Figure 7 is a block diagram of a word cloud processing device according to an exemplary embodiment;

[0034] Figure 8 is a block diagram of an electronic device 800 according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0035] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0036] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The implementation methods described in the following examples do not represent all implementation methods consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the attached claims.

[0037] It should be noted that the phrase "at least one of the items" in the present disclosure includes three types of parallel situations: "any one of the items", "a combination of any number of the items", and "all of the items". For example, "including at least one of A and B" includes the following three parallel situations: (1) including A; (2) including B; (3) including A and B. Another example is "executing at least one of step 1 and step 2" which means the following three parallel situations: (1) executing step 1; (2) executing step 2; (3) executing step 1 and step 2.

[0038] In response to the above problems, the present disclosure provides a word cloud processing method, which does not require the participation of business personnel, saves manpower costs, and can shorten the time to discover abnormalities and provide timely alarms. The following is an example of a problem scenario where a web page freezes.

[0039] Figure 2 is a schematic diagram showing an implementation scenario of a word cloud processing method according to an exemplary embodiment of the present disclosure, such as Figure 2 The implementation scenario includes a server 100, a user terminal 110 and a user terminal 120, wherein the user terminals are not limited to two, and include but are not limited to mobile phones, personal computers and other devices. The user terminal may be installed with an application that can obtain user needs. The server may be a single server, or a server cluster composed of several servers, or a cloud computing platform or a virtualization center.

[0040] The user terminal 110 and the user terminal 120 obtain the user demand text and transmit the user demand text to the server 100. The server 100 can segment the received user demand text and put the segmentation result into the word cloud. The server 100 can obtain the first frequency data of the words in the word cloud in the first period and the second frequency data of the words in multiple second periods, wherein the first period refers to a complete period before the current time, and the multiple second periods refer to a predetermined number of complete periods before the first period; then, the server 100 inputs the second frequency data into the frequency prediction model to obtain the predicted frequency data of the words in the first period; based on the predicted frequency data and the first frequency data, obtain the abnormal difference of the words; when the abnormal difference is greater than the preset alarm threshold, perform abnormal processing.

[0041] Next, we will refer to Figures 3 to 7 A word cloud processing method, device, and system according to exemplary embodiments of the present disclosure are described in detail.

[0042] Figure 3 is a flowchart of a word cloud processing method according to an exemplary embodiment. Figure 3 As shown, the word cloud processing method includes the following steps:

[0043] In step S301, first frequency data of words in the word cloud in a first period and second frequency data of words in multiple second periods are obtained, wherein the first period refers to a complete period before the current time, and the multiple second periods refer to a predetermined number of complete periods before the first period.

[0044] According to an exemplary embodiment of the present disclosure, before obtaining the first frequency data of the words in the word cloud in the first period and the second frequency data of the words in multiple second periods, it also includes: storing the frequency data of the words in the corresponding storage unit, wherein each word in the word cloud corresponds to a storage unit, and the frequency data includes the first frequency data and the second frequency data. According to this embodiment, each word has a corresponding storage unit to store the frequency of the corresponding word within a certain number of periods, which is convenient for subsequent acquisition of the frequency and for subsequent abnormal processing, such as alarm processing.

[0045] For example, the historical frequency of each word can be saved. For example, each word maintains a bidirectional queue (i.e., the above storage unit), and the bidirectional queue stores the frequency data of the corresponding word over a period of time. Figure 4As shown, timestamp represents the current time, predict_idx is the time index closest to timestap, that is, the time index idx required for alarm judgment, retain_idx represents the oldest data required for alarm judgment, and the data before retain_idx will be cleared to save space. The present disclosure can use the frequency data corresponding to the time index between retain_idx and predict_idx as the data inside the bidirectional queue corresponding to the word during alarm prediction. It should be noted that the frequency of words is not limited to the storage method of the bidirectional queue, and other methods with the same function can also be used, which is not limited by the present disclosure.

[0046] return Figure 3 In step S302, the second frequency data is input into the frequency prediction model to obtain the predicted frequency data of the words in the first period. For example, the frequency prediction model can use the Holt-Winters algorithm or a deep learning algorithm such as LSTM, which is not limited in the present disclosure.

[0047] In step S303, an abnormal difference value of the word is obtained based on the predicted frequency data and the first frequency data.

[0048] In step S304, when the abnormal difference is greater than the preset alarm threshold, abnormal processing is performed.

[0049] According to an exemplary embodiment of the present disclosure, when the abnormal difference is greater than the preset alarm threshold, before performing the abnormal processing, it also includes: obtaining the preset alarm threshold of the word based on the second frequency data and the first frequency data. According to this embodiment, the corresponding preset alarm threshold can be obtained according to the frequency data corresponding to the current word, that is, the threshold is dynamically set at different time points, and different words use different preset alarm thresholds, which can effectively reduce the false alarm and missed alarm of the problem and improve the accuracy of the alarm.

[0050] According to an exemplary embodiment of the present disclosure, when the abnormal difference value is greater than the preset alarm threshold, before performing abnormal processing, it also includes: fitting the second frequency data and the first frequency data to obtain the fitted frequency; obtaining the residual frequency based on the second frequency data, the first true frequency and the fitted frequency; obtaining the normal fluctuation value based on the standard deviation of the residual frequency; and obtaining the preset alarm threshold of the word based on the normal fluctuation value and the preset multiplier parameter. According to this embodiment, the preset alarm threshold of the current word can be obtained conveniently and quickly.

[0051] For example, a dynamic threshold algorithm for obtaining a preset alarm threshold can use a bidirectional queue as input. If the preset alarm threshold is greater than the abnormal difference, an alarm message is output for an abnormal alarm. The above dynamic threshold algorithm can use the holt-winters algorithm to fit the frequency data yi of the input bidirectional queue (the frequency data of the input bidirectional queue can be the data corresponding to the time index idx2-idx5) to obtain the fitting value sequence ypi, and use the true value of each input time index minus the fitting value sequence value to obtain the residual sequence di, and then calculate the standard deviation of the residual sequence as the normal fluctuation value δ, and input an alarm multiple parameter r. The alarm threshold K (that is, the above preset alarm threshold) is the product of the normal fluctuation value and the alarm multiple parameter. The detailed algorithm is as follows:

[0052] Residual di = yi - ypi (1)

[0053] Residual standard deviation δ=std(di) (2)

[0054] Threshold K = δ*r (3)

[0055] According to an exemplary embodiment of the present disclosure, the second frequency data is input into the frequency prediction model to obtain the predicted frequency data of the word in the first period, including: inputting the second frequency data in the storage unit corresponding to the word into the frequency prediction model to obtain the predicted frequency data of the word in the first period. According to this embodiment, the second frequency data to be input into the frequency prediction model can be quickly obtained from the storage unit, thereby quickly obtaining the predicted frequency data.

[0056] According to an exemplary embodiment of the present disclosure, when the abnormal difference is greater than a preset alarm threshold, after performing abnormal processing, the method further includes: receiving a word segmentation processing result sent by the word segmentation module; and updating a storage unit corresponding to the word based on the word segmentation processing result. According to this embodiment, the storage unit corresponding to the word can be updated in real time based on the word segmentation processing result.

[0057] For example, the word segmentation module can receive the demand text fed back by the user in real time, perform word segmentation on the demand text, and send the word segmentation processing result to the server. When the server receives the word segmentation processing result transmitted by the word segmentation module, it will update the frequency data corresponding to idx in the bidirectional queue in real time according to the word segmentation processing result.

[0058] According to an exemplary embodiment of the present disclosure, updating the storage unit corresponding to a word includes: storing the frequency data of the word in the new cycle in the word segmentation processing result into the storage unit corresponding to the word, and when the length of the storage unit corresponding to the word exceeds the preset maximum length, deleting the frequency data of the word in the first cycle in the storage unit until the length of the storage unit meets the preset maximum length. According to this embodiment, when the updated bidirectional queue exceeds the allowed length, it is deleted from the front to maintain the latest data.

[0059] In order to better understand the above embodiments, Figure 5 and Figure 6 For detailed explanation, Figure 5 is a workflow diagram of a word abnormality alarm according to an exemplary embodiment. Figure 6 The present invention is a data flow diagram of a word abnormality alarm according to an exemplary embodiment.

[0060] like Figure 5 As shown, a time series alarm model and a dynamic alarm threshold model can be built into the word cloud monitoring system, and the workflow of word abnormality alarm can be as follows:

[0061] 1. The user needs of different users are fed back to the analysis system (the system includes the word segmentation module in the above embodiment) through various channels such as mobile phones and web pages;

[0062] 2. The analysis system performs word segmentation processing on the received user needs, and the word segmentation processing results are sent to the word cloud monitoring system (which can be set in the server);

[0063] 3. Divide the words in the word cloud into continuous time indexes idx according to a fixed period (for example, 5 minutes as a period). That is, in the above embodiment, each word maintains a bidirectional queue, such as Figure 6 As shown, the frequency of a single word at each time index is used to judge abnormality using a time series alarm algorithm, and abnormal difference is obtained. Specifically, the bidirectional queue of words is input into the time series alarm algorithm to obtain the predicted frequency of the words, and the predicted frequency is compared with the actual frequency of the words to obtain the abnormal difference. For example, the time series alarm algorithm may be the Holt-Winters algorithm, but the present disclosure is not limited to this.

[0064] 4. If Figure 6 As shown, the bidirectional queue of words is input into the dynamic threshold algorithm to obtain the alarm threshold of the words (i.e., the preset alarm threshold mentioned above), and then the abnormal difference obtained in the third step is compared with the alarm threshold. If the abnormal difference is greater than the alarm threshold, an abnormal alarm is issued through the alarm path 2;

[0065] 5. Business personnel 2 intervenes in and handles the alarm information.

[0066] In summary, in view of the current problems of high labor cost, insufficient timeliness, easy false alarms and missed alarms, and difficult threshold setting in word cloud monitoring and alarm, the present disclosure proposes an automatic word cloud monitoring and alarm solution based on time series prediction. The solution has the following advantages: 1) Automatic alarm, no need for business personnel to participate, that is, no need to rely on the experience of business personnel to find abnormalities, so that user problems can be found in time, and the related technology manually finds abnormalities, and the automatic alarm of the present disclosure can also shorten the time to find user problems, thereby improving user experience. 2) The automatic alarm has high accuracy and recall rate, and the alarm is comprehensive. Even if the frequency of an abnormal problem itself is not high enough, it is not easy to be found in the word cloud. It can be detected by the method of the present disclosure, and then the alarm threshold is modified in real time by the dynamic threshold algorithm, which can effectively reduce false alarms and missed alarms. 3) Modeling the frequency characteristics of each word, different words will calculate different alarm thresholds, and the alarm is not affected by the frequency of the word itself. 4) The alarm threshold changes with time. The alarm algorithm based on the time series can timely capture the situation where the frequency of each word changes with time. 5) Timely alarm. The alarm algorithm based on the time series can find abnormal problems and alarm within a single time window, shortening the time to find abnormalities.

[0067] Figure 7 is a block diagram of a word cloud processing device according to an exemplary embodiment. Figure 7 The device includes a frequency data acquisition unit 70, a predicted frequency data acquisition unit 72, an abnormal difference acquisition unit 74 and a processing unit 76.

[0068] The frequency data acquisition unit 70 is configured to acquire the first frequency data of the words in the word cloud within a first period and the second frequency data of the words within multiple second periods, wherein the first period refers to a complete period before the current time, and the multiple second periods refer to a predetermined number of complete periods before the first period; the predicted frequency data acquisition unit 72 is configured to input the second frequency data into the frequency prediction model to obtain the predicted frequency data of the words within the first period; the abnormal difference acquisition unit 74 is configured to obtain the abnormal difference of the words based on the predicted frequency data and the first frequency data; the processing unit 76 is configured to perform abnormal processing when the abnormal difference is greater than the preset alarm threshold.

[0069] According to an exemplary embodiment of the present disclosure, the processing unit 76 is further configured to obtain a preset alarm threshold of the word based on the second frequency data and the first frequency data before performing abnormal processing when the abnormal difference is greater than the preset alarm threshold.

[0070] According to an exemplary embodiment of the present disclosure, the processing unit 76 is also configured to fit the second frequency data and the first frequency data to obtain a fitted frequency; obtain a residual frequency based on the second frequency data, the first frequency data and the fitted frequency data; obtain a normal fluctuation value based on the standard deviation of the residual frequency; obtain a preset alarm threshold of the word based on the normal fluctuation value and a preset multiplier parameter.

[0071] According to an exemplary embodiment of the present disclosure, the frequency data acquisition unit 70 is also configured to store the frequency data of the words into corresponding storage units before acquiring the first frequency data of the words in the word cloud in the first period and the second frequency data of the words in multiple second periods, wherein each word in the word cloud corresponds to a storage unit, and the frequency data includes the first frequency data and the second frequency data.

[0072] According to an exemplary embodiment of the present disclosure, the predicted frequency data acquisition unit 72 is further configured to input the second frequency data in the storage unit corresponding to the word into the frequency prediction model to obtain the predicted frequency data of the word in the first period.

[0073] According to an exemplary embodiment of the present disclosure, the processing unit 76 is also configured to receive the word segmentation processing result sent by the word segmentation module after performing exception processing when the abnormal difference is greater than a preset alarm threshold; and update the storage unit corresponding to the word based on the word segmentation processing result.

[0074] According to an exemplary embodiment of the present disclosure, the processing unit 76 is also configured to store the frequency data of the new periodic words in the word segmentation processing result into the storage unit corresponding to the word, and when the length of the storage unit exceeds the preset maximum length, delete the frequency data of the words in the front period in the storage unit in sequence until the length of the storage unit meets the preset maximum length.

[0075] According to another aspect of an embodiment of the present disclosure, a word cloud processing system is provided, including: a word segmentation module and a word cloud processing module, the word segmentation module is configured to receive a demand text fed back by a user and perform word segmentation processing on the demand text; the word cloud processing module is configured to execute the above word cloud processing method.

[0076] According to an embodiment of the present disclosure, an electronic device may be provided. Figure 8 It is a block diagram of an electronic device 800 according to an embodiment of the present disclosure, and the electronic device includes at least one memory 801 and at least one processor 802, wherein a set of computer executable instructions is stored in the at least one memory, and when the computer executable instruction set is executed by the at least one processor, the word cloud processing method according to the embodiment of the present disclosure is executed.

[0077] As an example, the electronic device 800 may be a PC, a tablet device, a personal digital assistant, a smart phone, or other device capable of executing the above-mentioned instruction set. Here, the electronic device 1000 is not necessarily a single electronic device, but may also be any device or circuit capable of executing the above-mentioned instructions (or instruction sets) individually or in combination. The electronic device 800 may also be part of an integrated control system or system manager, or may be configured as a portable electronic device interconnected with a local or remote (e.g., via wireless transmission) interface.

[0078] In the electronic device 800, the processor 802 may include a central processing unit (CPU), a graphics processing unit (GPU), a programmable logic device, a dedicated processor system, a microcontroller or a microprocessor. As an example and not limitation, the processor 802 may also include an analog processor, a digital processor, a microprocessor, a multi-core processor, a processor array, a network processor, etc.

[0079] The processor 802 may execute instructions or codes stored in the memory, wherein the memory 801 may also store data. Instructions and data may also be sent and received over a network via a network interface device, wherein the network interface device may employ any known transmission protocol.

[0080] The memory 801 may be integrated with the processor 802, for example, by placing RAM or flash memory within an integrated circuit microprocessor or the like. In addition, the memory 801 may include a separate device, such as an external disk drive, a storage array, or any other storage device that can be used by a database system. The memory 801 and the processor 802 may be operationally coupled, or may communicate with each other, such as through an I / O port, a network connection, etc., so that the processor 802 can read files stored in the memory 801.

[0081] In addition, the electronic device 800 may also include a video display (such as a liquid crystal display) and a user interaction interface (such as a keyboard, a mouse, a touch input device, etc.) All components of the electronic device may be connected to each other via a bus and / or a network.

[0082] According to an embodiment of the present disclosure, a computer-readable storage medium may also be provided, wherein when the instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor is prompted to execute the word cloud processing method of the embodiment of the present disclosure. Examples of computer-readable storage media here include: read-only memory (ROM), random access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD+R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD+R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or optical disk storage, hard disk drive (HDD), solid state drive (SSD), card storage (such as, multimedia card, secure digital (SD) card or extreme digital (XD) card), magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid state disk and any other device, any other device is configured to store computer programs and any associated data, data files and data structures in a non-transitory manner and provide the computer programs and any associated data, data files and data structures to a processor or computer so that the processor or computer can execute the computer program. The computer program in the above-mentioned computer-readable storage medium can be run in an environment deployed in a computer device such as a client, a host, an agent device, a server, etc. In addition, in one example, the computer program and any associated data, data files and data structures are distributed on a networked computer system, so that the computer program and any associated data, data files and data structures are stored, accessed and executed in a distributed manner by one or more processors or computers.

[0083] According to an embodiment of the present disclosure, a computer program product is provided, including computer instructions, and when the computer instructions are executed by a processor, the word cloud processing method of the embodiment of the present disclosure is implemented.

[0084] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0085] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A word cloud processing method, characterized in that: include: Obtaining first frequency data of words in the word cloud in a first period and second frequency data of the words in multiple second periods, wherein the first period refers to a complete period before the current time, and the multiple second periods refer to a predetermined number of complete periods before the first period; Inputting the second frequency data into a frequency prediction model to obtain predicted frequency data of the word in the first period; Based on the predicted frequency data and the first frequency data, obtaining an abnormal difference value of the word; When the abnormal difference is greater than a preset alarm threshold, performing abnormal processing; Wherein, when the abnormal difference is greater than the preset alarm threshold, before performing abnormal processing, the method further includes: Fitting the second frequency data and the first frequency data to obtain a fitted frequency; Obtaining a residual frequency based on the second frequency data, the first frequency data and the fitted frequency data; Based on the standard deviation of the residual frequency, a normal fluctuation value is obtained; Based on the normal fluctuation value and the preset multiplier parameter, a preset alarm threshold value of the word is obtained.

2. The word cloud processing method according to claim 1, characterized in that: Before obtaining the first frequency data of the words in the word cloud in the first period and the second frequency data of the words in the second periods, the method further includes: The frequency data of the words are stored in corresponding storage units, wherein each word in the word cloud corresponds to a storage unit, and the frequency data includes first frequency data and second frequency data.

3. The word cloud processing method according to claim 2, characterized in that: The step of inputting the second frequency data into a frequency prediction model to obtain the predicted frequency data of the words in the first period includes: The second frequency data in the storage unit corresponding to the word is input into the frequency prediction model to obtain the predicted frequency data of the word in the first period.

4. The word cloud processing method according to claim 2, characterized in that: When the abnormal difference is greater than the preset alarm threshold, after the abnormality processing is performed, the method further includes: Receive the word segmentation processing result sent by the word segmentation module; Based on the word segmentation result, the storage unit corresponding to the word is updated.

5. The word cloud processing method according to claim 4, characterized in that: The updating of the storage unit corresponding to the word includes: The frequency data of the word in the new period in the word segmentation processing result is stored in the storage unit corresponding to the word. When the length of the storage unit exceeds the preset maximum length, the frequency data of the word in the frontmost period in the storage unit is deleted in sequence until the length of the storage unit meets the preset maximum length.

6. A word cloud processing device, characterized in that: include: A frequency data acquisition unit is configured to acquire first frequency data of words in the word cloud within a first period and second frequency data of the words within a plurality of second periods, wherein the first period refers to a complete period before a current time, and the plurality of second periods refers to a predetermined number of complete periods before the first period; a predicted frequency data acquisition unit, configured to input the second frequency data into a frequency prediction model to obtain predicted frequency data of the word in the first period; an abnormal difference value obtaining unit, configured to obtain an abnormal difference value of the word based on the predicted frequency data and the first frequency data; A processing unit, configured to perform an exception process when the abnormal difference is greater than a preset alarm threshold; Wherein, the processing unit is also configured to fit the second frequency data and the first frequency data to obtain a fitted frequency before performing abnormal processing when the abnormal difference value is greater than a preset alarm threshold; obtain a residual frequency based on the second frequency data, the first frequency data and the fitted frequency data; obtain a normal fluctuation value based on the standard deviation of the residual frequency; and obtain a preset alarm threshold of the word based on the normal fluctuation value and a preset multiplier parameter.

7. The word cloud processing device according to claim 6, characterized in that: The frequency data acquisition unit is further configured to store the frequency data of the words in the corresponding storage unit before acquiring the first frequency data of the words in the word cloud in the first period and the second frequency data of the words in multiple second periods, wherein each word in the word cloud corresponds to a storage unit, and the frequency data includes the first frequency data and the second frequency data.

8. The word cloud processing device according to claim 7, characterized in that: The predicted frequency data acquisition unit is further configured to input the second frequency data in the storage unit corresponding to the word into the frequency prediction model to obtain the predicted frequency data of the word in the first period.

9. The word cloud processing device according to claim 7, characterized in that: The processing unit is further configured to receive a word segmentation processing result sent by the word segmentation module after performing an exception process when the abnormal difference is greater than a preset alarm threshold; Based on the word segmentation result, the storage unit corresponding to the word is updated.

10. The word cloud processing device according to claim 9, characterized in that: The processing unit is further configured to store the frequency data of the word in the new period in the word segmentation processing result into the storage unit corresponding to the word, and when the length of the storage unit exceeds the preset maximum length, delete the frequency data of the word in the frontmost period in the storage unit in sequence until the length of the storage unit meets the preset maximum length.

11. A word cloud processing system, characterized in that: include: Word segmentation module and word cloud processing module, The word segmentation module is configured to receive a demand text fed back by a user and perform word segmentation processing on the demand text; The word cloud processing module is configured to execute the word cloud processing method according to any one of claims 1 to 5.

12. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the word cloud processing method according to any one of claims 1 to 5.

13. A computer-readable storage medium, characterized in that: When the instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor is prompted to perform the word cloud processing method according to any one of claims 1 to 5.

14. A computer program product comprising computer instructions, characterized in that When the computer instructions are executed by a processor, the word cloud processing method according to any one of claims 1 to 5 is implemented.

Images