News production system and method based on big data and artificial intelligence

Abstract

The invention provides a news production system and method based on big data and artificial intelligence, and the system comprises a clue collection module which captures multi-source data information; the news clue generation module is used for generating a plurality of news clues from the captured multi-source data information; the evaluating and sorting module is used for evaluating the news value of the news clues and sorting the news clues according to priorities; the reporter scheduling and interview task distribution module is used for matching the reporter to go to the news scene of the news clue with the highest priority for interview based on the priority sequence and according to the background factors of the reporter; the storage module is used for uploading and storing interview manuscripts of journalists; the manuscript editing module is used for acquiring the interview manuscripts in the storage module and editing and processing the interview manuscripts by utilizing an artificial intelligence technology to obtain news manuscripts; the auditing module is used for auditing the edited news manuscripts; the manuscript distribution module is used for distributing the news manuscripts which are audited to be qualified to each platform; and the intelligent tracking module is used for tracking the distributed news manuscripts.

Description

Technical Field

[0001] The present invention relates to the technical field of news media, and particularly relates to a news production system based on big data and artificial intelligence, and a news production method based on big data and artificial intelligence. Background Art

[0002] The production process of news manuscripts includes multiple links such as collecting leads, interviewing and data collection, manuscript writing, editing and modification, review and proofreading, release and promotion, feedback and improvement, etc. Each link requires the collaboration of roles such as journalists, editors, and review personnel to ensure the quality and compliance of news manuscripts.

[0003] However, in the prior art, during the process of collecting leads, there is a lack of a systematic standard to judge the value of news leads (such as timeliness, influence, significance, etc.), mainly relying on manual experience, with strong subjectivity and low efficiency. Summary of the Invention

[0004] The present invention provides a news production system and method based on big data and artificial intelligence to solve the above technical problems.

[0005] The technical solution adopted by the present invention is as follows:

[0006] A news production system based on big data and artificial intelligence, characterized by including: a lead collection module for real-time capturing multi-source data information through big data technology; a news lead generation module for generating multiple news leads from the captured multi-source data information through classification, clustering, and topic modeling technologies; an evaluation and sorting module for evaluating the news value of news leads and sorting the priorities of news leads; a journalist scheduling and interview task assignment module for matching journalists to go to the news scene of the news lead with the highest priority for interviews based on the priority order of the news value of news leads and according to the background factors of journalists; a storage module for uploading and storing the interview manuscripts of journalists; a manuscript editing module for obtaining the interview manuscripts in the storage module and using artificial intelligence technology to edit and process them to obtain news manuscripts; a review module for reviewing the edited news manuscripts; a manuscript distribution module for distributing the qualified news manuscripts to each platform; an intelligent tracking module for tracking the distributed news manuscripts to obtain tracking data.

[0007] The data types of the multi-source data information include text data, image data, and video data.

[0008] The method for generating news leads includes the following steps: extracting the features of text data, image data, and video data respectively; fusing the features of text, image, and video through multi-modal data fusion; classifying the fused features using a machine learning model; using a clustering algorithm to cluster similar data into news leads; extracting the theme from the text data; and generating multiple news leads based on the classification, clustering, and theme modeling results.

[0009] The evaluation and ranking module includes an evaluation unit for evaluating the news value of news leads. The evaluation unit evaluates the news value based on six factors: newsworthiness factor, importance factor, audience attention factor, verifiability factor, reporting feasibility factor, and potential impact factor.

[0010] The news value is expressed by the following formula:

[0011]

[0012] In the formula, V S represents the news value of the S-th news lead, N represents the newsworthiness score, I represents the importance score, A represents the audience attention score, C represents the verifiability score, F represents the reporting feasibility score, P represents the potential impact score; K N 、K I 、K A 、K C 、K F 、K P represent the power exponents of the newsworthiness score, importance score, audience attention score, verifiability score, reporting feasibility score, and potential impact score respectively.

[0013] The calculation formula for the newsworthiness score N is:

[0014]

[0015] In the formula, N1 represents the timeliness score, N2 represents the novelty score, N3 represents the exclusivity score; W N2 represents the weight coefficient

[0016] The calculation formula for the importance score I is:

[0017]

[0018] In the formula, I1 represents the geographical scope score, I2 represents the population involved score, I3 represents the industry impact score, I4 represents the public interest relevance score, I5 represents the historical significance score; W I2 、W I3 represent the weight coefficients;

[0019] The calculation formula for the audience attention score A is as follows:

[0020]

[0021] In the formula, A1 represents the target audience range score, A2 represents the broad interest score, and A3 represents the social media popularity score; represents the weight coefficient;

[0022] The calculation formula for the verifiability score C is as follows:

[0023]

[0024] In the formula, C1 represents the information source reliability score, C2 represents the sufficiency of evidence support score, and C3 represents the verification difficulty score; and represent the weight coefficient;

[0025] The calculation formula for the report feasibility score F is as follows:

[0026]

[0027] In the formula, F1 represents the rationality of resource requirement score, F2 represents the legal compliance score, and F3 represents the operation difficulty score; represents the weight coefficient;

[0028] The calculation formula for the potential impact score P is as follows:

[0029]

[0030] In the formula, P1 represents the significance of positive impact score, P2 represents the controllability of negative impact score, and P3 represents the expected social response score; represents the weight coefficient.

[0031] The evaluation and ranking module further includes a ranking unit, which is used to rank the news value of news leads evaluated by the evaluation unit, and prioritize them. For the news lead with the highest priority, multiple reporters can be dispatched simultaneously to form a temporary interview team for in-depth reporting.

[0032] A news production method based on big data and artificial intelligence, comprising the following steps: real-time capturing multi-source data information through big data technology; generating multiple news leads from the captured multi-source data information through classification, clustering, and topic modeling techniques; evaluating the news value of the news leads and sorting them according to the priority of the news leads; matching reporters to go to the news scene of the news lead with the highest priority for interviews based on the priority order of the news value of the news leads and according to the background factors of the reporters; obtaining the interview manuscripts of the reporters and using artificial intelligence technology to edit and process them to obtain news manuscripts; reviewing the edited news manuscripts; distributing the qualified news manuscripts to various platforms; and tracking the distributed news manuscripts to obtain tracking data.

[0033] Advantages of the present invention:

[0034] Through the news lead generation module provided in the present invention, and through classification, clustering, and topic modeling techniques, multiple news leads are generated from the captured multi-source data information. Then, through the evaluation and sorting module provided, the news value of the news leads is evaluated, and the news leads are sorted according to the priority. The interview scheduling mechanism based on value evaluation in the present invention focuses limited reporter resources on high-value news leads, avoiding wasting time and energy on low-value leads. High-value leads usually involve important events or hot topics and can produce more influential news reports. By real-time evaluating the timeliness of the leads, reporters can be quickly dispatched to follow up on hot events, ensuring the timely release of news. High-quality reports generated from high-value leads can also attract more readers, enhancing the influence and competitiveness of the media. Description of the drawings

[0035] Figure 1 It is a block diagram of the news production system based on big data and artificial intelligence according to an embodiment of the present invention;

[0036] Figure 2 It is a flowchart of the news production method based on big data and artificial intelligence according to an embodiment of the present invention. Detailed implementation manners

[0037] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0038] As Figure 1 shown, the news production system based on big data and artificial intelligence according to an embodiment of the present invention includes:

[0039] Clue collection module: Used to capture multi-source data information in real time through big data technology.

[0040] Among them, the multi-source data information includes the Internet (news websites, forums, blogs, etc.), social media (Twitter, Facebook, Weibo, Douyin, etc.), government announcements (government websites, announcement platforms, etc.), etc. Specifically, web crawlers (such as Scrapy, BeautifulSoup, etc.), API interfaces (such as Twitter API, Facebook GraphAPI, etc.), RSS subscriptions (used to capture news websites and blogs), and other capture tools can be used to capture the corresponding data information in real time.

[0041] Among them, the data types of the multi-source data information include text data, image data, and video data.

[0042] News clue generation module: Used to generate multiple news clues from the captured multi-source data information through classification, clustering, and topic modeling techniques. The clue generation specifically includes the following steps:

[0043] Among them, the processing of text data includes the following steps:

[0044] First, clean the text data, including removing HTML tags, special characters, punctuation marks, etc., and removing stop words (such as meaningless words like "de", "shi", etc.). Then perform word segmentation, including word segmentation for Chinese texts (such as using Jieba, HanLP); finally, perform deduplication and filtering, including removing duplicate texts and filtering out texts that are too short or meaningless.

[0045] Then use natural language processing technology (NLP) to perform sentiment analysis, topic modeling, entity recognition, and text feature extraction on the text data.

[0046] Among them, sentiment analysis is used to judge the sentiment tendency of the text data, such as positive, negative, or neutral. Specifically, pre-trained models (such as BERT, TextBlob, VADER, etc.) or machine learning (such as SVM, LSTM) can be used to train a sentiment classifier for sentiment analysis.

[0047] Topic modeling is used to extract topics from the text data. Specifically, LDA (Latent Dirichlet Allocation) or BERTopic tools can be used.

[0048] Entity analysis is used to identify named entities in the text data, such as person names, place names, organization names, etc. Specifically, tools such as SpaCy, StanfordNLP, BERT-based NER models, etc. can be used.

[0049] Extract text features using TF-IDF, Word2Vec, BERT, etc.

[0050] In the embodiments of the present invention, the processing of image data and video data includes the following steps:

[0051] Among them, the steps of image data processing are:

[0052] First, perform image annotation. You can use an image annotation tool (such as LabelImg) to manually annotate the objects in the image; then perform image enhancement. You can use OpenCV or PIL to perform operations such as cropping, scaling, and rotating on the image.

[0053] Use a pre-trained model (such as ResNet, VGG) to extract image features.

[0054] The steps of video data processing are:

[0055] First, extract key frames. For example, use FFmpeg or OpenCV to extract key frames from the video; then extract video captions. Use speech recognition technology (such as Google Speech-to-Text, Whisper) to extract the speech in the video and convert it into text.

[0056] Adopt the image data processing method to extract the image features of the key frames, and adopt the text data processing method to extract the text features after converting the video captions into text.

[0057] Through multi-modal data fusion, fuse the features of text, images, and videos to form a unified representation. Specifically, concatenate the text features, image features, and video features into a vector; use a multi-modal model (such as CLIP, ViLT) to directly process text and image data.

[0058] First, generate multiple news leads through classification, clustering, and topic modeling techniques. Specifically,

[0059] Then, use a machine learning model (such as SVM, random forest) to classify the fused features.

[0060] Next, use a clustering algorithm (such as K-Means, DBSCAN) to cluster similar data into news leads.

[0061] Then, use LDA or BERTopic to extract topics from the text data.

[0062] Finally, generate multiple news leads based on the classification, clustering, and topic modeling results.

[0063] Evaluation and sorting module: used to evaluate the news value of news leads and sort the priorities of news leads.

[0064] The evaluation and ranking module includes an evaluation unit for evaluating the news value of news leads.

[0065] Evaluating the news value of news leads is a multi-dimensional and comprehensive process involving multiple considerations. The present invention evaluates news value from six factors: newsworthiness factors, importance factors, audience attention factors, verifiability factors, reportability factors, and potential impact factors, ranks the news value according to the evaluation, and updates it in real time.

[0066] Specifically, the newsworthiness factors mainly include:

[0067] Timeliness: Whether the news lead is timely, that is, whether the event has recently occurred or is about to occur, and whether it needs to be reported in a short time.

[0068] Novelty: Whether the lead contains new information or a unique perspective, and whether it can provide unprecedented insights to the public.

[0069] Exclusivity: Whether the news lead is an exclusive report, that is, whether only a few or only one media has access to the information.

[0070] The importance factors mainly include:

[0071] Social impact: The long-term impact that the event involved in the news lead may have on aspects such as public life, social order, and national policies.

[0072] Public interest: Whether the lead is related to public interests, such as issues of public safety, health, education, environment, etc.

[0073] Historical significance: Whether the event has historical significance, or whether it may become an important node for future historical research.

[0074] The audience attention factors mainly include:

[0075] Target audience: Whether the news lead targets a specific audience group and the potential interest of this group in the event.

[0076] Broad interest: Whether the lead is likely to attract the attention of a wide audience, including people of different ages, genders, regions, occupations, etc.

[0077] Social media potential: The potential of the event to trigger discussions and sharing on social media and the possible viral spread effect.

[0078] The verifiability factors mainly include:

[0079] Information source: Whether the source of the news lead is reliable and whether the information provider has authority or credibility.

[0080] Evidential Support: Whether there is sufficient evidence or data to support the statements or claims in the lead.

[0081] Verification Difficulty: The ease or difficulty of verifying the authenticity of the lead, and whether additional resources or time are required for verification.

[0082] The main factors for reporting feasibility include:

[0083] Resource Requirements: Whether the resources (such as manpower, material resources, and financial resources) required for reporting the news lead are feasible.

[0084] Legal Compliance: Whether the report complies with the requirements of laws and regulations, and whether there are legal risks such as privacy infringement and defamation.

[0085] Operational Difficulty: The technical difficulty and practical feasibility in the reporting process, such as the difficulty of interviews and shooting conditions.

[0086] The main potential impact factors include:

[0087] Positive Impact: The positive social effects that the news report may produce, such as raising public awareness and promoting the solution of problems.

[0088] Negative Impact: The negative impacts that the report may bring, such as causing panic, misunderstanding or controversy, and whether these impacts are controllable.

[0089] In summary, evaluating the news value of a news lead requires comprehensive consideration of multiple factors such as newsworthiness, importance, audience attention, verifiability, reporting feasibility, and potential impact. Through a comprehensive and objective evaluation, the news value can be more accurately judged, providing strong support for news reporting.

[0090] Among them, the news value is expressed by the following formula:

[0091] V S =N KN ×I KI ×A KA ×C KC ×F KF ×P KP

[0092] In the formula, V S represents the news value of the S-th news lead, N represents the newsworthiness score, I represents the importance score, A represents the audience attention score, C represents the verifiability score, F represents the reporting feasibility score, P represents the potential impact score; K N 、K I 、K A 、K C 、K F 、K PPower exponents representing news score, importance score, audience attention score, verifiability score, report feasibility score, and potential impact score respectively.

[0093] Among them, the calculation formula for the news score N is:

[0094]

[0095] In the formula, N1 represents the timeliness score, N2 represents the novelty score, and N3 represents the exclusivity score; W N2 represents the weight coefficient

[0096] The calculation formula for the importance score I is:

[0097]

[0098] In the formula, I1 represents the geographical scope score, I2 represents the population involved score, I3 represents the industry impact score, I4 represents the public interest relevance score, and I5 represents the historical significance score; W I2 、W I3 represent the weight coefficients;

[0099] The calculation formula for the audience attention score A is:

[0100]

[0101] In the formula, A1 represents the target audience scope score, A2 represents the broad interest score, and A3 represents the social media popularity score; represent the weight coefficient;

[0102] The calculation formula for the verifiability score C is:

[0103]

[0104] In the formula, C1 represents the information source reliability score, C2 represents the evidence support sufficiency score, and C3 represents the verification difficulty score; and represent the weight coefficients;

[0105] The calculation formula for the report feasibility score F is:

[0106]

[0107] In the formula, F1 represents the resource demand rationality score, F2 represents the legal compliance score, and F3 represents the operation difficulty score; represent the weight coefficient;

[0108] The calculation formula for the potential impact score P is:

[0109]

[0110] In the formula, P1 represents the significance score of positive impact, P2 represents the controllability score of negative impact, and P3 represents the expected score of social response; represents the weight coefficient.

[0111] In an embodiment of the present invention, the evaluation and ranking module further includes a ranking unit for ranking the news value of news leads evaluated by the evaluation unit in terms of priority.

[0112] Journalist scheduling and interview task assignment module: Based on the priority order of the news value of news leads and according to the background factors of journalists (such as geographical location, professional field, work experience, etc.), first match the most suitable journalist to go to the news scene of the news lead with the highest priority for an interview. At the same time, update the location and status information of the journalist in real time to ensure that the actual situation of the journalist can be accurately reflected, improving the scheduling efficiency and accuracy.

[0113] In addition, for the news lead with the highest priority, multiple journalists can be scheduled simultaneously to form a temporary interview team for in-depth reporting.

[0114] Storage module: Used to upload and store the interview manuscripts of journalists.

[0115] Manuscript editing module: Used to obtain the interview manuscripts from the storage module and edit the interview manuscripts to obtain news manuscripts. Specifically, use artificial intelligence technology to edit and process them to obtain high-quality news manuscripts.

[0116] Among them, the specific method of using artificial intelligence technology to edit and process it is as follows:

[0117] First, perform data preprocessing, including text cleaning, word segmentation, removing redundant words, and text segmentation.

[0118] Then, perform text analysis and understanding. Specifically, perform semantic analysis and understanding on the interview manuscripts through NLP technology. First, identify entities such as people, places, and organizations in the text through tools such as SpaCy and StanfordNLP; then extract keywords in the text through tools such as TF-IDF and TextRank.

[0119] Next, perform text generation and text optimization. Among them, the generation of text mainly uses text generation models to edit and process the interview manuscripts. Specifically, use tools such as GPT-3, BERT, and T5 to generate news manuscripts from the interview manuscripts; convert the style of the interview manuscripts into the style of news manuscripts through a style transfer model (such as GPT-3 and T5).

[0120] Text optimization is mainly used to improve the readability and attractiveness of news articles. Among them, text optimization specifically includes readability optimization and sentiment optimization. Readability optimization is used to enhance the readability of the text, and sentiment optimization is used to adjust the emotional tendency of the text.

[0121] Specifically, the Flesch-Kincaid readability formula is used to calculate the readability score of the text, and the sentence length and complexity are adjusted. The Flesch-Kincaid readability formula is as follows:

[0122]

[0123] In the formula, Q represents the readability score, Z1 represents the total number of words, Z2 represents the total number of sentences, Z3 represents the total number of syllables, and 206.835 and 1.015 are constants in the formula obtained through statistical methods.

[0124] Then, the emotional tendency of the text is adjusted through sentiment analysis models (such as TextBlob, VADER).

[0125] Finally, by combining text generation, style transfer, and text optimization, high-quality news articles are generated.

[0126] Review module: Used to review the edited news articles.

[0127] Among them, the review of news articles mainly examines whether they comply with news ethics and laws and regulations. News ethics is the basic criterion of the news industry, ensuring the fairness, objectivity, and social responsibility of news reports. For fact-checking of news articles, fact-checking tools (such as Google Fact Check Tools, FactCheck.org) can be used to verify the facts and data in the news to ensure the reliability of the data and information sources; anonymize the content involving personal privacy to avoid disclosing sensitive information of interviewees (such as addresses, phone numbers), and directly modify the unqualified parts to the qualified state. The review module can also conduct a re-review of news articles manually. For unqualified parts, they can be modified to the qualified state manually.

[0128] Article distribution module: The article distribution module is used to distribute the reviewed news articles to various platforms, including the paper media printing workshop, PC terminals, mobile terminals, etc. This module ensures the wide dissemination and coverage of news content.

[0129] Intelligent tracking module: Used to track the distributed news articles to obtain tracking data. The corresponding data includes information such as the access volume, access area, access terminal, and access population of the news articles. In the present invention, the intelligent tracking module helps to understand the dissemination effect of news content and the feedback from the audience.

[0130] The news production system of the present invention based on big data and artificial intelligence generates multiple news leads through the set news lead generation module and by using classification, clustering, and topic modeling techniques on the multi-source data information captured. Then, through the set evaluation and sorting module, it evaluates the news value of the news leads and sorts them according to the priority of the news leads. The interview scheduling mechanism of the present invention based on value evaluation focuses the limited reporter resources on high-value news leads, avoiding wasting time and energy on low-value leads. High-value leads usually involve important events or hot topics and can produce more influential news reports. By real-time evaluating the timeliness of the leads, it can quickly dispatch reporters to follow up on hot events and ensure the timely release of news. The high-quality reports generated from high-value leads can also attract more readers and enhance the influence and competitiveness of the media.

[0131] Corresponding to the news production system based on big data and artificial intelligence in the above embodiment, the present invention also proposes a news production method based on big data and artificial intelligence.

[0132] As Figure 2 shown, the news production method based on big data and artificial intelligence in the embodiment of the present invention includes the following steps:

[0133] S1: Real-time capture multi-source data information through big data technology.

[0134] S2: Generate multiple news leads from the captured multi-source data information through classification, clustering, and topic modeling techniques.

[0135] S3: Evaluate the news value of the news leads and sort them according to the priority of the news leads.

[0136] S4: Based on the priority order of the news value of the news leads and according to the background factors of the reporters, match reporters to go to the news scene of the news lead with the highest priority for interviews.

[0137] S5: Obtain the interview manuscripts of the reporters and use artificial intelligence technology to edit and process them to obtain news manuscripts.

[0138] S6: Review the edited news manuscripts.

[0139] S7: Distribute the news manuscripts that pass the review to various platforms.

[0140] S8: Track the distributed news manuscripts to obtain tracking data.

[0141] A news production method based on big data and artificial intelligence according to an embodiment of the present invention evaluates the news value of news leads, sorts the priorities of news leads, and based on a value evaluation-based interview scheduling mechanism, concentrates limited journalist resources on high-value news leads, avoiding wasting time and energy on low-value leads. High-value leads usually involve important events or hot topics and can generate more influential news reports. By real-time evaluating the timeliness of leads, journalists can be quickly dispatched to follow up on hot events to ensure the timely release of news. High-quality reports generated from high-value leads can also attract more readers and enhance the influence and competitiveness of the media.

[0142] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. The meaning of "a plurality" is two or more unless otherwise specifically defined.

[0143] In the present invention, unless otherwise clearly specified and defined, terms such as "installed", "connected", "connected to", "fixed" and the like shall be construed in a broad sense. For example, it may be a fixed connection, a detachable connection, or integrated; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the internal communication of two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0144] In the present invention, unless otherwise clearly specified and defined, the first feature being "on" or "under" the second feature may be that the first and second features are in direct contact, or the first and second features are indirectly in contact through an intermediate medium. Moreover, the first feature being "above", "over" and "on top of" the second feature may be that the first feature is directly above or obliquely above the second feature, or merely indicates that the first feature has a higher horizontal height than the second feature. The first feature being "under", "below" and "beneath" the second feature may be that the first feature is directly below or obliquely below the second feature, or merely indicates that the first feature has a lower horizontal height than the second feature.

[0145] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0146] Any process or method description represented in a flowchart or described in other ways herein can be understood to represent a module, segment, or portion of code including one or more executable instructions for implementing a specific logical function or process, and the scope of the preferred embodiments of the present invention includes additional implementations, where the functions can be executed in a manner not shown or discussed, including in a substantially simultaneous manner according to the functions involved or in a reverse order, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0147] The logic and / or steps represented in a flowchart or described in other ways herein, for example, can be considered as an ordered list of executable instructions for implementing a logical function, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in connection with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of the computer-readable medium include the following: an electrical connection portion having one or more wirings (electronic device), a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other suitable processing as necessary, and then stored in a computer memory.

[0148] It should be understood that each part of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one of the following techniques known in the art or a combination thereof can be used: discrete logic circuits with logic gate circuits for implementing logic functions on data signals, application specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0149] Those of ordinary skill in the art can understand that all or part of the steps carried by the method of the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0150] In addition, each functional unit in various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above integrated module can be implemented in the form of hardware or in the form of a software functional module. When the above integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0151] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A news production system based on big data and artificial intelligence, characterized in that: include: The clue collection module is used to capture multi-source data information in real time through big data technology; The news clue generation module is used to generate multiple news clues from the captured multi-source data information through classification, clustering and topic modeling techniques; Evaluation and sorting module, used to evaluate the news value of news clues and sort the news clues according to their priorities; The reporter scheduling and interview task allocation module matches reporters to the news scene with the highest priority news clues for interviews based on the priority order of news value of news clues and the background factors of reporters; Storage module, used to upload and store journalists’ interview manuscripts; The manuscript editing module is used to obtain the interview manuscripts in the storage module and edit and process them using artificial intelligence technology to obtain news articles; Review module, used to review edited news articles; The manuscript distribution module is used to distribute the news articles that have passed the review to various platforms; The intelligent tracking module is used to track news releases after distribution to obtain tracking data.

2. The news production system based on big data and artificial intelligence according to claim 1 is characterized by: The data types of the multi-source data information include text data, image data, and video data.

3. The news production system based on big data and artificial intelligence according to claim 2 is characterized by: The method for generating news clues includes the following steps: Extract features of text data, image data and video data respectively; Through multimodal data fusion, the features of text, image and video are integrated; Use machine learning models to classify the fused features; Use clustering algorithms to aggregate similar data into news leads; Extracting topics from text data; Generate multiple news clues based on classification, clustering and topic modeling results.

4. The news production system based on big data and artificial intelligence according to claim 3 is characterized by: The evaluation and sorting module includes an evaluation unit for evaluating the news value of news clues. The evaluation unit evaluates the news value based on six factors: newsworthiness factor, importance factor, audience attention factor, verifiability factor, reporting feasibility factor, and potential impact factor.

5. The news production system based on big data and artificial intelligence according to claim 4 is characterized in that: The news value is expressed by the following formula: Where V S represents the news value of the S-th news clue, N represents the newsworthiness score, I represents the importance score, A represents the audience attention score, C represents the verifiability score, F represents the report feasibility score, and P represents the potential impact score; K N , K I , K A , K C , K F , K P They respectively represent the power exponents of newsworthiness score, importance score, audience attention score, verifiability score, reporting feasibility score and potential impact score.

6. The news production system based on big data and artificial intelligence according to claim 5 is characterized in that: The calculation formula of newsworthiness score N is: In the formula, N1 represents the timeliness score, N2 represents the novelty score, and N3 represents the exclusivity score; W N2 Represents the weight coefficient The calculation formula of importance score I is: In the formula, I1 represents the score of geographical scope, I2 represents the score of the population involved, I3 represents the score of industry impact, I4 represents the score of public interest relevance, and I5 represents the score of historical significance; W I2 , W I3 represents the weight coefficient; The calculation formula of audience attention score A is: In the formula, A1 represents the target audience range score, A2 represents the broad interest score, and A3 represents the social media popularity score; represents the weight coefficient; The calculation formula of the verifiability score C is: In the formula, C1 represents the reliability score of the information source, C2 represents the sufficiency score of evidence support, and C3 represents the difficulty score of verification; and represents the weight coefficient; The calculation formula for the report feasibility score F is: In the formula, F1 represents the rationality score of resource demand, F2 represents the legal compliance score, and F3 represents the operational difficulty score; represents the weight coefficient; The calculation formula of potential impact score P is: In the formula, P1 represents the positive impact significance score, P2 represents the negative impact controllability score, and P3 represents the social response expectation score; Represents the weight coefficient.

7. The news production system based on big data and artificial intelligence according to claim 6 is characterized by: The evaluation and sorting module also includes a sorting unit for evaluating the news value of news clues based on the evaluation unit and prioritizing them. For news clues with the highest priority, multiple reporters can be dispatched simultaneously to form a temporary interview team for in-depth reporting.

8. A news production method based on big data and artificial intelligence, characterized in that: The following steps are involved: Capture multi-source data information in real time through big data technology; Through classification, clustering and topic modeling techniques, the captured multi-source data information is generated into multiple news clues; Evaluate the news value of news leads and prioritize them; Based on the priority order of news value of news clues and the background factors of reporters, reporters are matched to go to the news scene with the highest priority news clues for interviews; Obtain journalists' interview manuscripts, edit and process them using artificial intelligence technology to obtain news articles; Review the edited news releases; Distribute the approved news releases to various platforms; Follow up on the distributed news releases to obtain tracking data.

Images