Intelligent content inspection and proofreading method, device, equipment and readable storage medium

Abstract

The present application belongs to the technical field of media content security monitoring, and discloses an intelligent content inspection and proofreading method, device, equipment and readable storage medium, the method comprising: processing data from different sources and incorporating them into a database; receiving inspection rules, alarm rules and proofreading keyword libraries configured by users; performing inspections based on inspection rules to obtain manuscripts to be reviewed; determining the corresponding target proofreading keyword library based on inspection rules; proofreading the manuscripts to be reviewed based on the target proofreading keyword library to obtain proofreading results; and issuing warnings based on proofreading results and alarm rules. The present application facilitates targeted inspection and proofreading of manuscripts, and improves the inspection efficiency of editing content on the basis of ensuring the safety of editing content during inspection.

Description

Technical Field

[0001] The present application relates to the technical field of media content security monitoring, and in particular to an intelligent content inspection and proofreading method, device, equipment and readable storage medium. Background Art

[0002] Review and proofreading are important steps in the content editing process of book, newspaper and periodical publishing units. In order to ensure the content orientation and quality of publication, the industry currently mainly relies on full-time professional proofreaders who meet the relevant qualifications and job training requirements to review published or issued manuscripts. However, this method consumes a lot of human resources and has high requirements. Proofreaders need to have corresponding search and judgment experience. In the face of huge amounts of manuscript data, the quality and speed of review cannot be guaranteed. For new media such as websites, Weibo, WeChat, and clients run by publishing units, they also conduct preliminary reviews of manuscripts by accessing proofreading services to reduce the reviewers' obvious but common error troubleshooting work, which correspondingly improves the quality and speed of review.

[0003] However, usually publishing units can only allow their own manuscript data to be accessed as a third-party data source to a third-party proofreading service, and proofread in real time using the usual rules of the platform on which the third-party proofreading service relies. Users of publishing units cannot customize proofreading inspections and corresponding proofreading rules, and it is inconvenient to add new data sources. It is impossible to conduct targeted inspections and proofreading of manuscript data from different data sources to meet the needs of personalized review business. Moreover, the entire process of data access, inspection, and proofreading requires a corresponding amount of time each time, and the inspection efficiency for editing content is actually relatively low. Summary of the invention

[0004] The present application provides an intelligent content inspection and proofreading method, device, equipment and readable storage medium, which facilitates targeted inspection and proofreading of manuscripts, and improves the inspection efficiency of editing content on the basis of ensuring the security of editing content during inspection.

[0005] In a first aspect, an embodiment of the present application provides an intelligent content inspection and proofreading method, the method comprising:

[0006] Process data from different sources and merge them into the database;

[0007] Receive inspection rules, alarm rules and proofreading keyword library configured by the user, wherein the inspection rules include target data source, inspection mode, inspection time range and inspection content, and the alarm rules include alarm name, alarm range, alarm trigger threshold and alarm contact person;

[0008] Constructing a first mapping relationship between the content to be inspected in the inspection rule and the proofreading keyword library, and determining a corresponding target proofreading keyword library based on the first mapping relationship and the content to be inspected in the inspection rule;

[0009] Constructing a second mapping relationship between the target data source in the inspection rule and the alarm triggering threshold and the alarm contact person in the alarm rule;

[0010] Find the target data source in the database based on the inspection rules, and inspect the manuscript data within the inspection time range in the target data source in the inspection mode to obtain the manuscripts to be reviewed;

[0011] Proofread the manuscript to be reviewed based on the target proofreading keyword library to obtain the proofreading results;

[0012] When it is determined based on the second mapping relationship and the proofreading result that the number of problematic manuscripts to be verified exceeds the alarm trigger threshold in the alarm rule, an alarm message is sent to the corresponding alarm contact, wherein the alarm message includes an alarm name and an alarm range.

[0013] Furthermore, the step of processing the data from different sources and incorporating them into the database includes:

[0014] Pull data based on API interfaces and set scheduled tasks, or actively pull data from different sources from real-time consumption queues;

[0015] The data from different sources are expanded in terms of fields, non-standard dirty data are cleaned, and indexed to distinguish the sources. Based on the index tags, the data from different sources are distributed to the database.

[0016] Further, the step of finding the target data source in the database based on the inspection rule, inspecting the manuscript data within the inspection time range in the target data source in the inspection mode, and obtaining the manuscripts to be reviewed includes:

[0017] Find the target data source in the inspection rule based on the corresponding index tag;

[0018] Determine the type of inspection mode to be inspected in the inspection rule. The types of inspection mode to be inspected include one-time inspection and automatic inspection.

[0019] If it is a one-time inspection, the time range to be inspected is the preset batch and preset time period of the target data source, and the manuscript data within the preset batch and preset time period in the target data source are inspected once to obtain the manuscripts to be reviewed;

[0020] If it is an automatic inspection, the inspection time range is the preset inspection cycle of the target data source, and the manuscript data updated by the target data source in each cycle is continuously inspected in the preset inspection cycle to obtain the manuscripts to be reviewed.

[0021] Furthermore, after the step of receiving the inspection rules, alarm rules and proofreading keyword library configured by the user, the method further includes:

[0022] Determine whether inspection rules, alarm rules, proofreading keyword libraries, or data from different sources in the database have changed;

[0023] If the inspection rules, alarm rules or proofreading keyword library are changed, the first mapping relationship is changed accordingly;

[0024] If the data from different sources in the database changes, the second mapping relationship will be changed accordingly.

[0025] Furthermore, the manuscripts to be reviewed include text, pictures and video manuscripts.

[0026] Furthermore, the proofreading keyword library is based on manual entry by users or batch import of files.

[0027] Furthermore, the content to be inspected includes review types such as errors, violence, terrorism or pornography-related issues in the manuscript, and different content to be inspected corresponds to different proofreading keyword libraries.

[0028] In a second aspect, the present invention further provides an intelligent content inspection and proofreading device, the device comprising:

[0029] The data processing module is used to process data from different sources and then merge them into the database;

[0030] A receiving module, configured to receive inspection rules, alarm rules and proofreading keyword library configured by a user, wherein the inspection rules include target data source, inspection mode, inspection time range and inspection content, and the alarm rules include alarm name, alarm range, alarm trigger threshold and alarm contact person;

[0031] A word library determination module is used to construct a first mapping relationship between the content to be inspected in the inspection rule and the proofreading keyword library, and determine a corresponding target proofreading keyword library based on the first mapping relationship and the content to be inspected in the inspection rule;

[0032] A mapping relationship building module, used to build a second mapping relationship between the target data source in the inspection rule and the alarm triggering threshold and the alarm contact in the alarm rule;

[0033] The inspection module is used to find the target data source in the database based on the inspection rules, and inspect the manuscript data within the inspection time range in the target data source in the inspection mode to obtain the manuscripts to be reviewed;

[0034] The proofreading module is used to proofread the manuscript to be reviewed based on the target proofreading keyword library to obtain the proofreading results;

[0035] The alarm module is used to send an alarm message to a corresponding alarm contact when the number of manuscripts with problems to be verified exceeds the alarm trigger threshold in the alarm rule based on the second mapping relationship and the proofreading result, wherein the alarm message includes an alarm name and an alarm range.

[0036] Furthermore, the data processing module is specifically used for:

[0037] Pull data based on API interfaces and set scheduled tasks, or actively pull data from different sources from real-time consumption queues;

[0038] The data from different sources are expanded in terms of fields, non-standard dirty data are cleaned, and indexed to distinguish the sources. Based on the index tags, the data from different sources are distributed to the database.

[0039] Furthermore, the inspection module is specifically used for:

[0040] Find the target data source in the inspection rule based on the corresponding index tag;

[0041] Determine the type of inspection mode to be inspected in the inspection rule. The types of inspection mode to be inspected include one-time inspection and automatic inspection.

[0042] If it is a one-time inspection, the time range to be inspected is the preset batch and preset time period of the target data source, and the manuscript data within the preset batch and preset time period in the target data source are inspected once to obtain the manuscripts to be reviewed;

[0043] If it is an automatic inspection, the inspection time range is the preset inspection cycle of the target data source, and the manuscript data updated by the target data source in each cycle is continuously inspected in the preset inspection cycle to obtain the manuscripts to be reviewed.

[0044] Furthermore, the intelligent content inspection and proofreading device also includes a change module, which is used to:

[0045] Determine whether inspection rules, alarm rules, proofreading keyword libraries, or data from different sources in the database have changed;

[0046] If the inspection rules, alarm rules or proofreading keyword library are changed, the first mapping relationship is changed accordingly;

[0047] If the data from different sources in the database changes, the second mapping relationship will be changed accordingly.

[0048] Furthermore, the manuscripts to be reviewed include text, pictures and video manuscripts.

[0049] Furthermore, the proofreading keyword library is based on manual entry by users or batch import of files.

[0050] Furthermore, the content to be inspected includes review types such as errors, violence, terrorism or pornography-related issues in the manuscript, and different content to be inspected corresponds to different proofreading keyword libraries.

[0051] In a third aspect, an embodiment of the present application further provides an intelligent content inspection and proofreading device, which includes a processor, a memory, and an intelligent content inspection and proofreading program stored in the memory and executable by the processor, wherein when the intelligent content inspection and proofreading program is executed by the processor, the steps of the intelligent content inspection and proofreading method described above are implemented.

[0052] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which an intelligent content inspection and proofreading program is stored, wherein when the intelligent content inspection and proofreading program is executed by a processor, the steps of the intelligent content inspection and proofreading method as described above are implemented.

[0053] In summary, compared with the prior art, the technical solution provided in the embodiment of the present application has at least the following beneficial effects:

[0054] The embodiments of the present application provide an intelligent content inspection and proofreading method, device, equipment and readable storage medium. In the present application, the intelligent content inspection supports customized inspection rules and customized keyword library proofreading, so as to carry out targeted proofreading of manuscripts in a more flexible and convenient manner, and can flexibly configure different alarm rules for multiple inspection sources, so that different regulatory groups can grasp the latest status of inspection and proofreading. On the basis of ensuring the safety of editing content through inspection, the efficiency of editing content inspection is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A flowchart of an intelligent content inspection and proofreading method provided by one embodiment of the present application;

[0056] Figure 2 A schematic diagram of the functional modules of an intelligent content inspection and proofreading device provided in one embodiment of the present application;

[0057] Figure 3 This is a schematic diagram of the hardware structure of an intelligent content inspection and proofreading device involved in an embodiment of the present application. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0059] See also Figure 1 The embodiment of the present application provides an intelligent content inspection and proofreading method, which is applied to an intelligent content inspection and proofreading platform. The method specifically includes:

[0060] Step S10, processing data from different sources and incorporating them into a database;

[0061] Step S20, receiving inspection rules, alarm rules and proofreading keyword library configured by the user, wherein the inspection rules include target data source, inspection mode, inspection time range and inspection content, and the alarm rules include alarm name, alarm range, alarm trigger threshold and alarm contact person;

[0062] Step S30, constructing a first mapping relationship between the content to be inspected in the inspection rule and the proofreading keyword library, and determining a corresponding target proofreading keyword library based on the first mapping relationship and the content to be inspected in the inspection rule;

[0063] Step S40, constructing a second mapping relationship between the target data source in the inspection rule and the alarm triggering threshold and the alarm contact in the alarm rule;

[0064] Step S50, based on the inspection rule, find the target data source in the database, and inspect the manuscript data within the inspection time range in the target data source in the inspection mode to obtain the manuscripts to be reviewed;

[0065] Step S60, proofreading the manuscript to be reviewed based on the target proofreading keyword library to obtain a proofreading result;

[0066] Step S70, when it is determined based on the second mapping relationship and the proofreading result that the number of manuscripts with problems to be verified exceeds the alarm triggering threshold in the alarm rule, an alarm message is sent to the corresponding alarm contact, wherein the alarm message includes an alarm name and an alarm range.

[0067] In this embodiment, data from different sources include electronic manuscript data (internal data sources) from distribution channels such as internal paper media and internal websites of publishing units, and electronic manuscript data (third-party data sources) from distribution channels such as WeChat, Weibo, and other third-party platforms. Due to the large amount of data and complexity, in order to facilitate the subsequent inspection and proofreading of data from different sources, the data connected to the intelligent content inspection and proofreading platform will be uniformly incorporated into the cloud native database corresponding to the platform for management. Before being uniformly incorporated into the intelligent content inspection platform, considering the different ways in which data from different sources access the platform and in order to specifically distinguish different data sources for inspection and processing, it is necessary to process data from different sources separately and then incorporate them into the database. After the data is uniformly managed in this way, it is also convenient to add new data sources in the future, and it will not interfere with the inspection and proofreading of existing data sources that have been connected.

[0068] In addition, before inspecting and proofreading data from different sources, this embodiment supports platform users to flexibly customize inspection rules, proofreading keyword libraries and alarm rules. Among them, the inspection rules include the target data source (corresponding to the source of the data source that needs to be inspected), the inspection mode (one-time inspection or scheduled inspection), the inspection time range (determine the manuscripts to be reviewed in the source of the data source that needs to be inspected) and the content to be inspected (corresponding to the review categories when proofreading and reviewing the manuscripts to be reviewed, including the review of errors, violent terrorism or pornographic issues). The alarm rules include the alarm name (to facilitate users to distinguish different alarms), the alarm range (which inspection sources need to be alarmed), the alarm trigger threshold (when the number of problematic manuscripts to be verified is more than the alarm trigger threshold, an alarm will be triggered) and the alarm contact (when the alarm is triggered, an alarm will be sent to the users in the alarm contact list). Therefore, before the inspection, the inspection rules, proofreading keyword library and alarm rules customized by the user on the platform will be received and read first, so as to conduct inspections and proofreading on data from different sources based on business needs, and facilitate different groups of people to understand the latest status of the inspection.

[0069] After receiving the inspection rules, the target data source in the database can be found based on the inspection rules, and the manuscript data within the inspection time range in the target data source can be inspected in the inspection mode to obtain the manuscript to be reviewed. Since multiple inspections and proofreading may be set at the same time, and considering that there will be current and historical inspections and proofreading, in order to reduce the confusion of multiple setting rules and improve proofreading efficiency, the first mapping relationship between the content to be inspected in the inspection rules and the proofreading keyword library will be constructed first, and then the corresponding target proofreading keyword library will be determined based on the first mapping relationship and the content to be inspected in the inspection rules. After that, the manuscript to be reviewed is proofread in a targeted manner based on the target proofreading keyword library to obtain the proofreading results, which can improve the accuracy of the review and comparison content and expand the application coverage. Among them, the deep learning proofreading algorithm model used for proofreading is an algorithm model developed by the platform itself. The training is based on the proofreading keyword library of all audit types. The proofreading keyword library will change. The algorithm model can also be continuously updated and optimized using the keyword library after the change and update to improve the error detection and correction capabilities of the proofreading manuscript text. During proofreading, the proofreading vocabulary and proofreading rules accumulated by the media industry are mainly used for matching, and corresponding warnings are marked when a match is found. The existing proofreading through semantic model proofreading is highly universal, but the proofreading results will be more segmented and accurate when the proofreading vocabulary accumulated by the media industry is used.

[0070] The alarm rules corresponding to different data sources are also set differently, so it is necessary to construct a second mapping relationship between the target data source in the inspection rule and the alarm trigger threshold and alarm contact in the alarm rule. After calling the second mapping relationship to obtain the proofreading result, the proofreading result and the alarm rule will be compared to determine whether the number of manuscripts with problems to be verified in the manuscripts to be reviewed exceeds the alarm trigger threshold in the alarm rule. If it exceeds, an alarm message will be sent to the corresponding alarm contact, so that different groups of people can grasp the latest status of the inspection in time and take corresponding measures. Among them, the alarm information includes the alarm name and the alarm range, so as to accurately issue an alarm.

[0071] The intelligent content inspection and proofreading platform of this embodiment supports the inspection of internal resources and third-party data sources, and also supports customized inspection rules and keyword library proofreading. It can be more flexible and convenient to perform targeted proofreading of manuscripts, and supports flexible configuration of alarm rules for different inspection sources, so that different regulatory groups can grasp the latest status of inspections. On the basis of ensuring the security of editing content through inspections, the inspection efficiency of editing content is further improved.

[0072] Furthermore, in one embodiment, step S10 includes:

[0073] Pull data based on API interfaces and set scheduled tasks, or actively pull data from different sources from real-time consumption queues;

[0074] The data from different sources are expanded in terms of fields, non-standard dirty data are cleaned, and indexed to distinguish the sources. Based on the index tags, the data from different sources are distributed to the database.

[0075] In this embodiment, for internal data sources, the internal data is synchronized to the cloud native database of the platform, and can be checked in real time when used. Therefore, it can be synchronized to the inspection platform by consuming mq data in real time. The third-party data source can set a scheduled task and pull it regularly through the API. Among them, due to the different sources of different data sources and the possible existence of unnecessary dirty data, it is necessary to expand the fields of the manuscript data, clean the non-standard dirty data, and adapt the data coming in from different data sources, and mark them with corresponding index tags for distinction. Finally, the data is distributed and incorporated into the cloud native database for use by the platform for inspection and proofreading. Among them, after a series of operations such as data pulling, data preprocessing, data adaptation and data distribution are performed on the manuscript data, it is necessary to provide data services based on index tags. The tag corresponds to a field to distinguish manuscript data from different sources.

[0076] Furthermore, in one embodiment, step S30 includes:

[0077] Find the target data source in the inspection rule based on the corresponding index tag;

[0078] Determine the type of inspection mode to be inspected in the inspection rule. The types of inspection mode to be inspected include one-time inspection and automatic inspection.

[0079] If it is a one-time inspection, the time range to be inspected is the preset batch and preset time period of the target data source, and the manuscript data within the preset batch and preset time period in the target data source are inspected once to obtain the manuscripts to be reviewed;

[0080] If it is an automatic inspection, the inspection time range is the preset inspection cycle of the target data source, and the manuscript data updated by the target data source in each cycle is continuously inspected in the preset inspection cycle to obtain the manuscripts to be reviewed.

[0081] In this embodiment, when the manuscript to be reviewed is found according to the inspection rules, it is necessary to locate and search the data from different sources in the database based on the corresponding data index until the target data source in the inspection rules is found. After finding the target data source, it can be determined whether the type of inspection mode in the inspection rules is a one-time inspection or an automatic inspection. If it is a one-time inspection, it is only necessary to inspect the manuscripts within the selected batch and time range, and it is only executed once. The one-time inspection can support the inspection of historical data. The user can conduct a one-time inspection of the historical data in batches and time periods. When the content in the inspection rules changes, or when a custom keyword is added, the user can modify the entry to re-inspect the historical manuscript data. If it is an automatic inspection, the time range to be inspected is the preset inspection cycle of the target data source. By default, the manuscript data updated by the target data source in each cycle will be continuously inspected with the preset inspection cycle to obtain the manuscript to be reviewed. For example, if the preset inspection cycle is one hour, the manuscript data added within this hour will be inspected every hour.

[0082] Furthermore, in one embodiment, after step S20, the method further includes:

[0083] Determine whether inspection rules, alarm rules, proofreading keyword libraries, or data from different sources in the database have changed;

[0084] If the inspection rules, alarm rules or proofreading keyword library are changed, the first mapping relationship is changed accordingly;

[0085] If the data from different sources in the database changes, the second mapping relationship will be changed accordingly.

[0086] In this embodiment, the inspection rules, alarm rules, proofreading keyword library configured by the user and the data from different sources in the database will all be changed, and the changes include additions, deletions and modifications. When the inspection rules, alarm rules or proofreading keyword library are changed, the first mapping relationship is changed accordingly, and the target proofreading keyword library is determined by the new first mapping relationship. This can be used to deal with the situation when the content in the inspection rules changes, or when new custom keywords are added, the user can modify the entry by himself to re-inspect the historical manuscript data. When the data from different sources in the database are changed, the second mapping relationship is changed accordingly. This can be used to deal with the situation where the inspection data source of the new channel is freely added, and the alarm rules and alarm persons of the inspection data source are newly mapped, so that the newly added inspection data source can continue to be inspected, proofread and alarmed in a targeted manner.

[0087] Furthermore, in one embodiment, the manuscripts to be reviewed include text, pictures and video manuscripts.

[0088] In this embodiment, the proofreading algorithm model through deep learning can support the review and proofreading of text manuscripts. In addition, the API interface of third-party AI capabilities can be introduced to review and proofread image manuscripts and video manuscripts to further support the proofreading of various manuscript forms.

[0089] Furthermore, in one embodiment, the proofreading keyword library is based on manual entry by the user or batch import of files.

[0090] In this embodiment, the proofreading keyword library can be manually entered by the user or imported in batches of files. Through these two methods, the user can import the proofreading keyword library based on business needs and customization.

[0091] Furthermore, in one embodiment, the content to be inspected includes the review types of errors, violence, terrorism or pornography-related issues in the manuscript, and different content to be inspected corresponds to different proofreading keyword libraries and different alarm information display contents.

[0092] In this embodiment, the content to be inspected corresponds to the audit category when the manuscript to be audited is proofread and audited, including the audit of errors, violence, terrorism or pornography. For different audit categories, the corresponding proofreading keywords and the content displayed by the alarm are also different, so that targeted proofreading and accurate alarming can be carried out.

[0093] The embodiment of the present application also provides an intelligent content inspection and proofreading device.

[0094] Reference Figure 2 , a schematic diagram of the functional modules of the first embodiment of the intelligent content inspection and proofreading device.

[0095] In this embodiment, the intelligent content inspection and proofreading device includes:

[0096] The data processing module 10 is used to process the data from different sources and then merge them into the database;

[0097] The receiving module 20 is used to receive the inspection rules, alarm rules and proofreading keyword library configured by the user, wherein the inspection rules include the target data source, the inspection mode, the inspection time range and the inspection content, and the alarm rules include the alarm name, the alarm range, the alarm trigger threshold and the alarm contact person;

[0098] The word library determination module 30 is used to construct a first mapping relationship between the content to be inspected in the inspection rule and the proofreading keyword library, and determine the corresponding target proofreading keyword library based on the first mapping relationship and the content to be inspected in the inspection rule;

[0099] A mapping relationship building module 40 is used to build a second mapping relationship between the target data source in the inspection rule and the alarm triggering threshold and the alarm contact in the alarm rule;

[0100] Inspection module 50, used to find the target data source in the database based on the inspection rules, and inspect the manuscript data within the inspection time range in the target data source in the inspection mode to obtain the manuscripts to be reviewed;

[0101] The proofreading module 60 is used to proofread the manuscript to be reviewed based on the target proofreading keyword library to obtain a proofreading result;

[0102] The alarm module 70 is used to send an alarm message to a corresponding alarm contact when the number of manuscripts with problems to be verified exceeds the alarm trigger threshold in the alarm rule based on the second mapping relationship and the proofreading result, wherein the alarm message includes an alarm name and an alarm range.

[0103] Furthermore, in one embodiment, the data processing module 10 is specifically used for:

[0104] Pull data based on API interfaces and set scheduled tasks, or actively pull data from different sources from real-time consumption queues;

[0105] The data from different sources are expanded in terms of fields, non-standard dirty data are cleaned, and indexed to distinguish the sources. Based on the index tags, the data from different sources are distributed to the database.

[0106] Furthermore, in one embodiment, the inspection module 30 is specifically used for:

[0107] Find the target data source in the inspection rule based on the corresponding index tag;

[0108] Determine the type of inspection mode to be inspected in the inspection rule. The types of inspection mode to be inspected include one-time inspection and automatic inspection.

[0109] If it is a one-time inspection, the time range to be inspected is the preset batch and preset time period of the target data source, and the manuscript data within the preset batch and preset time period in the target data source are inspected once to obtain the manuscripts to be reviewed;

[0110] If it is an automatic inspection, the inspection time range is the preset inspection cycle of the target data source, and the manuscript data updated by the target data source in each cycle is continuously inspected in the preset inspection cycle to obtain the manuscripts to be reviewed.

[0111] Furthermore, in one embodiment, the intelligent content inspection and proofreading device further includes a change module, which is used to:

[0112] Determine whether inspection rules, alarm rules, proofreading keyword libraries, or data from different sources in the database have changed;

[0113] If the inspection rules, alarm rules or proofreading keyword library are changed, the first mapping relationship is changed accordingly;

[0114] If the data from different sources in the database changes, the second mapping relationship will be changed accordingly.

[0115] Furthermore, in one embodiment, the manuscripts to be reviewed include text, pictures and video manuscripts.

[0116] Furthermore, in one embodiment, the proofreading keyword library is based on manual entry by the user or batch import of files.

[0117] Furthermore, in one embodiment, the content to be inspected includes the types of review of errors, violence, terrorism, or pornography-related issues in the manuscript, and different content to be inspected corresponds to different proofreading keyword libraries.

[0118] Among them, the functional implementation of each module in the above-mentioned intelligent content inspection and proofreading device corresponds to the various steps in the above-mentioned intelligent content inspection and proofreading method embodiment, and its functions and implementation processes will not be repeated here one by one.

[0119] An embodiment of the present application provides an intelligent content inspection and proofreading device, which may be a device with data processing function, such as a personal computer (PC), a laptop computer, or a server.

[0120] Reference Figure 3 , Figure 3The hardware structure diagram of the intelligent content inspection and proofreading device involved in the embodiment of the present application. In the embodiment of the present application, the intelligent content inspection and proofreading device may include a processor 1001 (e.g., a central processing unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components; the user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard); the network interface 1004 may optionally include a standard wired interface, a wireless interface (such as wireless fidelity WIreless-FIdelity, WI-FI interface); the memory 1005 may be a high-speed random access memory (random access memory, RAM), or a stable memory (non-volatile memory), such as a disk memory, and the memory 1005 may optionally be a storage device independent of the aforementioned processor 1001. Those skilled in the art will appreciate that the hardware structure shown in the diagram of intelligent content inspection does not constitute a limitation of the present invention, and may include more or fewer components than shown in the diagram, or a combination of certain components, or a different arrangement of components.

[0121] Continue to refer to Figure 3 , Figure 3 The memory 1005 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and an intelligent content inspection and proofreading program. The processor 1001 may call the intelligent content inspection and proofreading program stored in the memory 1005 and execute the steps of the intelligent content inspection and proofreading method provided in the embodiment of the present application.

[0122] Among them, the method implemented when the intelligent content inspection and proofreading program is executed can refer to the various embodiments of the intelligent content inspection and proofreading method of the present application, and will not be repeated here.

[0123] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0124] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. An intelligent content inspection and proofreading method, characterized in that: The method comprises: Process data from different sources and merge them into the database; Receive inspection rules, alarm rules and proofreading keyword library configured by the user, wherein the inspection rules include target data source, inspection mode, inspection time range and inspection content, and the alarm rules include alarm name, alarm range, alarm trigger threshold and alarm contact person; Constructing a first mapping relationship between the content to be inspected in the inspection rule and the proofreading keyword library, and determining a corresponding target proofreading keyword library based on the first mapping relationship and the content to be inspected in the inspection rule; Constructing a second mapping relationship between the target data source in the inspection rule and the alarm triggering threshold and the alarm contact person in the alarm rule; Find the target data source in the database based on the inspection rule, and inspect the manuscript data within the inspection time range in the target data source in the inspection mode to obtain the manuscript to be reviewed; specifically, find the target data source in the inspection rule based on the corresponding index mark; determine the type of the inspection mode in the inspection rule, and the types of the inspection mode include one-time inspection and automatic inspection; if it is a one-time inspection, the inspection time range is the preset batch and preset time period of the target data source, and the manuscript data within the preset batch and preset time period in the target data source is inspected once to obtain the manuscript to be reviewed; if it is an automatic inspection, the inspection time range is the preset inspection cycle of the target data source, and the manuscript data updated by the target data source in each cycle is continuously inspected in the preset inspection cycle to obtain the manuscript to be reviewed; Proofread the manuscript to be reviewed based on the target proofreading keyword library to obtain the proofreading results; When it is determined based on the second mapping relationship and the proofreading result that the number of problematic manuscripts to be verified exceeds the alarm trigger threshold in the alarm rule, an alarm message is sent to the corresponding alarm contact, wherein the alarm message includes an alarm name and an alarm range.

2. The method according to claim 1, characterized in that The step of processing data from different sources and incorporating them into a database comprises: Pull data based on API interfaces and set scheduled tasks, or actively pull data from different sources from real-time consumption queues; The data from different sources are expanded in terms of fields, non-standard dirty data are cleaned, and indexed to distinguish the sources. Based on the index tags, the data from different sources are distributed to the database.

3. The method according to claim 1, characterized in that After the step of receiving the inspection rules, alarm rules and proofreading keyword library configured by the user, the method further includes: Determine whether inspection rules, alarm rules, proofreading keyword libraries, or data from different sources in the database have changed; If the inspection rules, alarm rules or proofreading keyword library are changed, the first mapping relationship is changed accordingly; If the data from different sources in the database changes, the second mapping relationship will be changed accordingly.

4. The method according to claim 1, characterized in that: The manuscripts awaiting review include text, pictures and videos.

5. The method according to claim 1, characterized in that: The proofreading keyword library is based on manual entry by the user or batch import of files.

6. The method according to claim 1, characterized in that: The content to be inspected includes review types such as errors, violence, terrorism or pornography-related issues in the manuscript. Different content to be inspected corresponds to different proofreading keyword libraries.

7. An intelligent content inspection and proofreading device, characterized in that: The device comprises: The data processing module is used to process data from different sources and then merge them into the database; A receiving module, configured to receive inspection rules, alarm rules and proofreading keyword library configured by a user, wherein the inspection rules include target data source, inspection mode, inspection time range and inspection content, and the alarm rules include alarm name, alarm range, alarm trigger threshold and alarm contact person; A word library determination module is used to construct a first mapping relationship between the content to be inspected in the inspection rule and the proofreading keyword library, and determine a corresponding target proofreading keyword library based on the first mapping relationship and the content to be inspected in the inspection rule; A mapping relationship building module, used to build a second mapping relationship between the target data source in the inspection rule and the alarm triggering threshold and the alarm contact in the alarm rule; The inspection module is used to find the target data source in the database based on the inspection rules, and inspect the manuscript data within the inspection time range in the target data source in the inspection mode to obtain the manuscript to be reviewed; specifically, find the target data source in the inspection rule based on the corresponding index mark; determine the type of the inspection mode in the inspection rule, and the type of the inspection mode includes one-time inspection and automatic inspection; if it is a one-time inspection, the inspection time range is the preset batch and preset time period of the target data source, and the manuscript data within the preset batch and preset time period in the target data source is inspected once to obtain the manuscript to be reviewed; if it is an automatic inspection, the inspection time range is the preset inspection cycle of the target data source, and the manuscript data updated by the target data source in each cycle is continuously inspected in the preset inspection cycle to obtain the manuscript to be reviewed; The proofreading module is used to proofread the manuscript to be reviewed based on the target proofreading keyword library to obtain the proofreading results; The alarm module is used to send an alarm message to a corresponding alarm contact when the number of manuscripts with problems to be verified exceeds the alarm trigger threshold in the alarm rule based on the second mapping relationship and the proofreading result, wherein the alarm message includes an alarm name and an alarm range.

8. An intelligent content inspection and proofreading device, characterized in that: The intelligent content inspection and proofreading device includes a processor, a memory, and an intelligent content inspection and proofreading program stored in the memory and executable by the processor, wherein when the intelligent content inspection and proofreading program is executed by the processor, the steps of the intelligent content inspection and proofreading method as described in any one of claims 1 to 7 are implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an intelligent content inspection and proofreading program, wherein when the intelligent content inspection and proofreading program is executed by the processor, the steps of the intelligent content inspection and proofreading method according to any one of claims 1 to 7 are implemented.

Images