Intelligent device and method for generating marketing content related to theme

Through natural language processing and artificial intelligence analysis of the computing module, combining structured and unstructured data, optimized keyword sets and ontology sentences are generated, which solves the challenge of creating high-quality content in advertising marketing and achieves rapid and automated marketing Content generation improves market insight and marketing efficiency.

CN119067079APending Publication Date: 2024-12-03NATIONAL UNIVERSITY OF SINGAPORE +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310626287.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-30
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

In the field of advertising marketing, it is challenging and time-consuming to accurately understand market trends and dynamics and to create high-quality marketing content to stand out.

Method used

Provide a computing module that uses natural language input to identify structured and unstructured data, analyzes and combines data to generate optimized keyword sets, and then uses converter models to generate body sentences and utilizes artificial intelligence. The content generator model generates marketing content related to the topic based on the ontological sentences selected by the user.

Benefits of technology

It realizes the rapid and automated generation of high-quality marketing content, improves market insights and marketing efficiency, and reduces the time and cost of manual creation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119067079A_ABST
    Figure CN119067079A_ABST
Patent Text Reader

Abstract

A computing module for generating marketing content about a topic and a method of generating marketing content about a topic using a computing module are described. The calculation method comprises the following steps: receiving natural language input related to the theme, and processing the natural language input to obtain a processing result; identifying structured data and unstructured data originating from an external source related to the processing result; analyzing the unstructured data by using a first artificial intelligence model, and combining the analyzed unstructured data with the structured data to obtain an optimized keyword set; using a converter model to generate an ontology sentence set related to the theme according to the optimized keyword set and the processing result, and providing the ontology sentence set to a user module; receiving selected ontology sentences in the ontology sentence set from the user module; and using an artificial intelligence content generator model to generate marketing content about the theme according to the selected ontology sentence.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to a device and method for generating marketing content about a topic based on big data and artificial intelligence. Background Art

[0002] In the field of advertising and marketing, it is an extremely challenging, time-consuming and laborious task to accurately understand market trends, dynamics and hot spots, and to create high-quality marketing content based on marketing goals so that the brand can stand out and attract more consumers. Summary of the invention

[0003] In one aspect, the present application provides a computing module for generating marketing content about a topic, wherein the computing module includes a processing unit and a non-transitory medium readable by the processing unit. When the processing unit executes the instructions stored in the medium, the processing unit receives a natural language input related to the topic and processes the natural language input to obtain a processing result. Once completed, the processing unit will identify structured and unstructured data related to the processing result from an external source. Then, the processing unit further analyzes the unstructured data using a first artificial intelligence model and combines the analyzed unstructured data with the structured data to obtain an optimized keyword set. Then, the processing unit generates a set of ontology sentences related to the topic using a converter model, wherein the set of ontology sentences is generated based on the optimized keyword set and the processing result. Then, the processing unit provides the ontology sentences to a user module. In response to receiving a selection of an ontology sentence from a user module, the processing unit then generates marketing content for the topic based on the selected ontology sentence using an artificial intelligence content generator model.

[0004] In another aspect, the present application discloses a method for generating marketing content about a topic using a computing module. The method includes: receiving natural language input related to the topic, and processing the natural language input to obtain a processing result. Then, structured and unstructured data related to the processing result from an external source are determined. In the next step, the method involves analyzing the unstructured data using a first artificial intelligence model, and combining the analyzed unstructured data with the structured data to obtain an optimized keyword set. Then, a set of ontology sentences related to the topic based on the optimized keyword set and the processing result is generated using a converter model. In the next step, the method involves providing the ontology sentence set to the user module. In response to receiving a selection of an ontology sentence from the user module, the method then generates marketing content about the topic using an artificial intelligence content generator model based on the selected ontology sentence.

[0005] According to one embodiment, the present application discloses a computing module for generating marketing content about a topic. The computing module includes: a processing unit; and a non-transitory medium readable by the processing unit, in which instructions are stored, wherein when executed by the processing unit, the instructions cause the processing unit to perform the following steps: receiving natural language input related to the topic, and processing the natural language input to obtain a processing result; identifying structured data and unstructured data related to the processing result and originating from an external source; using a first artificial intelligence model to analyze the unstructured data, and combining the analyzed unstructured data with the structured data to obtain an optimized keyword set; using a converter model, generating an ontology sentence set related to the topic based on the optimized keyword set and the processing result, and providing the ontology sentence set to a user module; receiving selected ontology sentences from the ontology sentence set from the user module; and using an artificial intelligence content generator model, generating marketing content about the topic based on the selected ontology sentences.

[0006] Preferably, the non-transitory medium further includes instructions for instructing the processing unit to execute the following steps: publishing the generated marketing content on a digital marketing platform.

[0007] Preferably, the non-transitory medium also includes instructions for instructing the processing unit to perform the following steps: obtaining feedback related to the published marketing content; and providing feedback for the training data set of the first artificial intelligence model and the converter model.

[0008] In one embodiment, the computing module, wherein the instructions for identifying structured data and unstructured data related to processing results from an external source, also includes instructing the processing unit to perform the following steps: using a web crawler module to crawl the external source to identify the structured data and the unstructured data.

[0009] Preferably, the external source comprises a database comprising information from social media platforms, information from public websites and information from search engines.

[0010] Preferably, the instructions for combining the analyzed unstructured data with the structured data to obtain an optimized keyword set also include instructing the processing unit to perform the following steps: using the first artificial intelligence model to assign a weight score to each keyword in the structured and unstructured data, wherein the weight score is generated based on the spatiotemporal occurrence rate of the keyword.

[0011] Preferably, the first artificial intelligence model includes at least one of a natural language processing model, an image processing model and a video analysis model.

[0012] Preferably, the non-transitory medium further comprises instructions for instructing the processing unit to perform the following steps:

[0013] Identifying a digital marketing platform associated with the generated marketing content; and publishing the generated marketing content to the identified digital marketing platform.

[0014] Preferably, the ontology sentences include statements for defining the nature of the subject, the concept of the subject, and the relationship between the optimized keyword set.

[0015] The embodiment also includes a method for generating marketing content about a topic using a computing module, the method comprising: receiving natural language input related to the topic, and processing the natural language input to obtain a processing result; identifying structured data and unstructured data from external sources related to the processing result; using a first artificial intelligence model to analyze the unstructured data, and combining the analyzed unstructured data with the structured data to obtain an optimized keyword set; using a converter model to generate an ontology sentence set related to the topic based on the optimized keyword set and the processing result, and providing the ontology sentence set to a user module; receiving selected ontology sentences from the user module; and using an artificial intelligence content generator model to generate marketing content about the topic based on the selected ontology sentences. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Various embodiments of the present invention are described below with reference to the accompanying drawings, in which:

[0017] Figure 1 is a block diagram of modules in a computing device for generating marketing content on a topic according to one embodiment of the present invention;

[0018] Figure 2 A block diagram of a processing system for executing an embodiment of the present invention is shown;

[0019] Figure 3 is a module block diagram according to an embodiment of the present invention, illustrating the architecture and steps of importing data, cleaning, feature engineering, conversion and alignment to generate ontology sentences;

[0020] Figure 4 is a submodule block diagram according to an embodiment of the present invention, illustrating the architecture and steps of a market insight analysis module;

[0021] Figure 5 is a submodule block diagram according to an embodiment of the present invention, illustrating the architecture and steps of an artificial intelligence marketing content generation module; and

[0022] Figure 6is a flow chart illustrating a process or method for generating marketing content about a topic according to one embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following detailed description refers to the accompanying drawings, which show in an illustrated manner specific details and embodiments in which the present invention may be practiced. Features described in the context of one embodiment may be correspondingly applicable to the same or similar features in other embodiments. Features described in the context of one embodiment may be correspondingly applicable to other embodiments, even if not explicitly described in these other embodiments. In addition, additions and / or combinations and / or substitutions described for a feature in the context of one embodiment may be correspondingly applicable to the same or similar features in other embodiments.

[0024] In the context of various embodiments, the articles “a,” “an,” and “the” used with respect to features or elements include reference to one or more of the features or elements.

[0025] In the context of the various embodiments, the terms "about" or "approximately" as applied to numerical values ​​include the exact value and reasonable variations, for example within 10% of the specified value.

[0026] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0027] “Including” means including, but not limited to, whatever follows the word “including.” Thus, the use of the term “including” indicates that the listed elements are required or mandatory, but other elements are optional and may or may not be present.

[0028] In addition, those skilled in the art will appreciate that some functional units herein are labeled as modules. Those skilled in the art will also appreciate that modules can be implemented as circuits, logic chips, or discrete components of any kind. In addition, those skilled in the art will also appreciate that modules can also be implemented in software, which can then be executed by different types of processor architectures. In an embodiment of the present invention, a module may also include computer instructions or executable code, and may instruct a computer processor to perform a series of events based on the instructions received. Those skilled in the art may choose to implement the module according to their design, without limiting the scope of protection of the claims.

[0029] Figure 1 1 is a block diagram of a module for generating marketing content about a topic according to an embodiment of the present invention. In one embodiment, a computing device such as computing module 100 may include Figure 1All or part of the modules described. Those skilled in the art will appreciate that other types of computing devices may also include all or part of the modules without departing from the scope of the present invention. As shown in the figure, the computing module 100 includes a cross-platform content publishing module 101, an artificial intelligence content generator module 111, a market insight analysis module 121, and a data storage module 131.

[0030] The data storage module 131 may include an SQL database 132 and a NoSQL database 134. The SQL database 132 may include a database such as Oracle, MySQL, etc., for storing structured table data. The NoSQL database 134 may include MangoDB, Couchbase, etc., and its database is configured to store data, such as text data, image data, video data, and other types of unstructured data.

[0031] The market insight analysis module 121 includes three sub-modules, for example, a data collection and preprocessing sub-module 122 , a big data analysis sub-module 124 , and a market trend and insight generation sub-module 126 .

[0032] The data collection and preprocessing submodule 122 is configured to identify structured data and unstructured data related to a specific topic or a data set from an external source. The data collection and preprocessing submodule 122 can obtain these data using a stream processing platform such as Apache Kafka or Amazon Kinesis. In one embodiment of the present invention, the stream processing platform is configured to obtain multiple images from at least one external source, extract at least one visual semantic concept from the multiple images, detect an event from the multiple images by distinguishing an event semantic concept signal from a background semantic concept signal, and obtain one or more images related to the event semantic concept signal for visual description presentation of the detected event. Then, the preprocessing of the obtained data includes management with a specific workflow script, and the script is configured to perform cleaning and exploratory data analysis, as well as quality checks for each data type. Then, the submodule 122 stores the preprocessing results of the data in a database through SQL / NoSQL stream processing. In addition, for each stream processing platform, the data collection and preprocessing submodule 122 can be configured to use independent preprocessing scripts and databases to store each data type (text, image, audio and video content, etc.).

[0033] The big data analysis submodule 124 is configured to use natural language processing techniques, such as topic models, named entity recognition, tokenization, stemming and word grouping, word bag, sentiment analysis, etc., to analyze text data placed in unstructured data and convert the analyzed text data into structured data. These data are then integrated with other structured data. After the above steps are completed, the big data analysis submodule 124 will further process the integrated data and classify and store important market-related data according to pre-defined categories, such as by author, subject, entity, object, event, date, etc.

[0034] Market Trend and Insight Generation The marketing trend and insight generation submodule 126 is configured to input the data generated by the big data analysis submodule 124 into a natural language model, such as inputting important market-related data such as authors, topics, entities, objects, events, dates, etc. into a natural language model, such as a transformer model (e.g., Keytotext, etc.). The natural language model is configured to use a self-attention mechanism to capture long-range dependencies and relationships between words and / or tags. In other words, the transformer model is configured to focus on each position in the input sequence to calculate the representation of each word or tag based on its relationship with other words or tags in the sequence. In this embodiment, the transformer model can be pre-trained using data from the Internet, data from various texts (e.g., books, articles, and any other digital files), and / or specific marketing data sets (e.g., marketing texts or materials). Then, training is performed on a smaller labeled data set according to a specific task to further improve the pre-trained model.

[0035] Based on these important market-related data, the converter model automatically generates relevant ontology sentences to describe market trends and insights. As the market-related data increases and the scale of the architecture in the database scale increases, the semantic relationships in the ontology sentences will be improved with each training period. Ontology sentences are defined as statements that can describe the properties of entities, concepts, and relationships within a specific market or knowledge domain. In other words, ontology sentences can be defined as statements that define the properties of a subject, the concepts of a subject, and the relationship between an optimized set of keywords. In an embodiment of the present invention, ontology sentences can be expressed in a formal language, such as a Web Ontology Language or a Resource Description Framework, allowing computers to understand and reason about a specific market or knowledge domain. The generated ontology sentences can then be used in artificial intelligence models to define objects of a specific category, describe the relationships between objects, or provide constraints on the properties of objects in artificial intelligence models.

[0036] The artificial intelligence content generator module 111 includes a text content generation submodule 112, an image content generation submodule 114, a video content generation submodule 116, an audio content generation submodule 117, and a comprehensive content generation submodule 118. The artificial intelligence content generator module 111 is configured to automatically generate relevant marketing content based on the ontology sentences generated by the market insight analysis module 121.

[0037] In the process of generating relevant marketing content, module 111 is configured to utilize some or all of the submodules, such as text content generation submodule 112, which is configured to call artificial intelligence text generation tools, such as ChatGPT, etc. Image content generation submodule 114 is configured to call artificial intelligence image generation tools, such as Dall-E2, Fotor, etc. Video content generation submodule 116 is configured to call artificial intelligence video generation tools, such as Pictory, Synthesys, etc. Audio content generation submodule 117 is configured to call audio content generation tools to generate relevant text, image, video and audio marketing content based on market trends and insight content. These text, image, audio and video content can also be further integrated through artificial intelligence tools to generate comprehensive marketing content through comprehensive content generation submodule 118. In some embodiments, content files can be stored in data lake-based storage, such as Amazon S3, and can be reviewed internally or externally before other modules in computing device 100 perform further processing.

[0038] The cross-platform content publishing module 101 is configured to publish the marketing content generated by the artificial intelligence content generator module 111 on digital marketing platforms including but not limited to social media platforms, public websites or other digital marketing media channels, using the application program interface (API) submodule 102, the network service submodule 104 and other types of submodules 106 configured to publish digital information online. In an embodiment of the present invention, the module 101 is configured to automatically recommend digital marketing channels based on the marketing content generated by the artificial intelligence content generator module 111 and the historical data of each digital marketing platform. These historical data include but are not limited to the cost of each digital marketing platform, the effectiveness of each marketing platform, the return on investment (ROI) of each marketing content or activity, etc. Therefore, the marketing activities can be optimized according to the marketing goals.

[0039] According to an embodiment of the present invention, Figure 2 1 is a block diagram illustrating components of a processing system 200 that may be provided within a computing module 100, any module and / or sub-module of a module 100. Those skilled in the art will appreciate that the exact configuration of each processing system provided within these modules may be different, and the exact configuration of the processing system 200 may be different. Figure 2 The arrangements shown are provided by way of non-limiting examples only.

[0040] In an embodiment of the present invention, the processing system 200 may include a controller 201 and a user interface 202. The user interface 202 is configured to allow manual interaction between the user and the computing modules as needed, and includes user input instructions to provide input / output components required for updating each of these modules. The components of the user interface 202 may vary depending on the implementation, but generally include one or more of a display 240, a keyboard 235, and a touchpad 236.

[0041] The controller 201 communicates data with the user interface 202 via the bus 215, and may also include a memory 220, a processor 205 mounted on a circuit board for processing instructions and data for executing the method of the present embodiment, an operating system 206, an input / output (Input / Output) interface 230 and a communication interface for communicating with the user interface 202, which is presented in the form of a network card 250 in this embodiment. In one example, the network card 250 can be used to send data of the module to other processing devices through a wired or wireless network, or to receive data through a wired or wireless network. The wireless network can be implemented through the network card 250, which includes but is not limited to: Wireless Fidelity, Bluetooth, Near Field Communication, cellular networks, satellite networks, telecommunication networks, wide area networks, etc.

[0042] The memory 220 and the operating system 206 communicate data with the CPU 205 via the bus 210. The memory components include volatile and non-volatile memory, and various types of memory, including one or more of random access memory (RAM) 223, read-only memory (ROM) 225, and mass storage devices 245, wherein the mass storage device 245 includes one or more solid-state drives (SSDs). It will be understood by those skilled in the art that the above-mentioned memory components include non-transitory computer-readable media and should be considered to include all computer-readable media, except temporary, propagated signals. In general, instructions are stored in the memory components in the form of program code, but may also be hard-wired. The memory 220 may include a kernel and / or programming modules, such as software applications that may be stored in volatile or non-volatile memory.

[0043] In this article, the term "processor" is used to refer to any device or component capable of processing such instructions, and may include: a microprocessor, a microcontroller, a programmable logic device or other computing device. That is, the processor 205 may be provided by any suitable logic circuit for receiving input, processing according to instructions stored in a memory and generating output (e.g., a memory component or a display 240). In this embodiment, the processor 205 may be a single-core or multi-core processor with a memory addressing space. In one example, the processor 205 may be multi-core, including, for example, an 8-core central processing unit (Central Processing Unit). In another example, it may be a cluster of CPU cores running in parallel to accelerate calculations. In another example, it may also include a graphics processing unit core (e.g., NVIDIA V100 TensorCore Graphic Processing Unit), which is configured to run and continue deep learning models and artifacts.

[0044] In operation, the computing module 100 is configured to receive natural language input related to a topic. The natural language input may include a description of the topic, a marketing plan for the topic, and a business goal related to the topic. In one example, the description may include "selling mobile phone A to high-spending groups."

[0045] Then, the computing module 100 may process the natural language input using the market insight analysis module 121 to obtain a processing result. In an embodiment of the present invention, the module 121 may utilize a natural language processing model such as, but not limited to, a named entity recognition (Named Entity Recognition) model, a tokenization (Tokenization) model, etc., to extract keywords related to the marketing goal from the description, such as "sales", "mobile phone", "high consumer group", etc. The natural language processing model may also be configured to recommend other keywords related to the marketing goal (using a built-in recommendation engine) based on these directly extracted keywords, such as "consumer behavior", "mobile phone features", etc. In an embodiment of the present invention, the recommendation engine may include content-based, collaborative filtering, and a hybrid model combining these two methods. The content-based recommendation engine is configured to analyze previous marketing behaviors and preferences, and then recommend content similar to previously utilized content. On the other hand, the collaborative filtering recommendation engine may analyze the behaviors and preferences of marketing methods of other topics and recommend content accordingly. The hybrid model combines these two methods to provide more accurate and relevant suggestions. These keywords may then be further compiled by the computing module 100 or may be directly utilized by the data collection and pre-processing sub-module 122 .

[0046] The data collection and preprocessing submodule 122 is configured to collect relevant data from different external data sources based on these keywords. This can be achieved by using a web crawler, an application program interface, or a direct database connector. The external data sources accessed by the submodule 122 include, but are not limited to, social media platforms, public websites, and search engines. The data obtained from the external data source may include structured data, such as tabular data, or unstructured data, such as text, pictures, videos, and other types of unstructured data.

[0047] The big data analysis submodule 124 is configured to analyze the obtained unstructured data using an artificial intelligence model, and combine the analyzed unstructured data with the structured data to obtain an optimized keyword set. In an embodiment of the present invention, the artificial intelligence model may include a natural language processing model, an image processing machine learning model, or a video analysis machine learning model. For the sake of brevity, the detailed working principles of some or all models known to those skilled in the art are omitted herein. In the process of analyzing unstructured and structured data, the artificial intelligence model may optimize the data according to the subject, time, platform, industry, etc., so as to more accurately and completely describe market trends and insight content in the optimized keyword set. In addition, the optimization process may be performed by assigning a weight score to each keyword in the structured and unstructured data, wherein the weight score is generated according to the spatiotemporal occurrence rate of the keyword. In one example, the artificial intelligence model may generate optimized keywords such as "user experience", "high-definition camera", "unique", "material", etc. based on the analysis results.

[0048] The market trend and insight generation submodule 126 is configured to utilize a large-scale language model, such as a converter model, based on an optimized keyword set (such as obtained from submodule 124) and the processing results obtained from submodule 122, to generate a set of ontological sentences related to the topic. In an embodiment of the present invention, the converter model may include a large-scale language model, such as Keytotext, or a text-to-text converter, to automatically generate relevant sentences with ontological meaning, i.e., ontological sentences, to describe market trends and insight content using keywords. The Keytotext model is configured to take a set of keywords or phrases as input and generate a complete sentence or paragraph to capture the meaning of the keyword. It is noteworthy that Keytotext uses a converter-based architecture to generate its responses, and is typically trained on large text data sets including books, articles, and web pages, and can generate responses in several different languages. For the sake of brevity, this description omits some or all of the detailed working principles of Keytotext.

[0049] Using the previous example, the ontology sentence can be generated by the converter model as "How to emphasize user experience, camera functions, mobile phone materials and other unique advantages to generate advertising content for mobile phone A?". Then, the ontology sentence generated by the submodule 126 is provided to the user module (not shown). The user module may include a user terminal that is communicatively connected to the computing module 100, or may include a communication terminal that is remotely connected to the computing module 100 by wireless or wired means, which is a setting choice of those skilled in the art. Then, the user can use the user terminal to edit and / or select the ontology sentence, and then convey the selected ontology sentence to the artificial intelligence content generator module 111.

[0050] The artificial intelligence content generator module 111 is configured to generate marketing content for a topic using an artificial content generator model, which is done based on the ontology sentence selected and provided by the user module. In other words, the artificial intelligence content generator module 111 automatically generates relevant marketing content based on the content and format specified in the selected ontology sentence using the artificial content generator model. In one example, based on the previous example, the marketing content may be generated as "Mobile phone A has a high-end titanium alloy shell, powerful camera function, and super user experience. It is the best choice for the high-income class." In an embodiment of the present invention, the artificial content generator model may include an artificial intelligence text generation tool using artificial intelligence technologies such as natural language processing and machine learning to automatically generate text based on input data or prompts; or an artificial intelligence image generation tool using artificial intelligence technologies such as machine learning and deep learning to automatically generate images. Examples of usable artificial intelligence text generation tools include, but are not limited to, GPT-3 (Generative Pre-trained Transformer 3) developed by OpenAI, Grover, Textio, Copysmith and Hugging Face developed by OpenAI. Examples of AI image generation tools that can be used include, but are not limited to, Generative Adversarial Networks, StyleGAN, Deep Dream developed by Google, DALL-E developed by OpenAI, and Pix2Pix. For the sake of brevity, the present invention omits some or all of the detailed working principles of these generation tools.

[0051] According to other embodiments of the present invention, the marketing content generated by the artificial intelligence content generator module 111 can be directly provided to the cross-platform content publishing module 101. Then, the module 101 publishes the generated marketing content on the digital marketing platform. In further embodiments, the module 101 can obtain feedback related to the published marketing content, and then the feedback can be used to improve the marketing content generated by the artificial intelligence content generator module 111. This can be done by using the feedback to improve the training data set, which is used to train the converter model in the market trend and insight generation submodule 126 and the artificial intelligence model in the big data analysis submodule 124.

[0052] Figure 3 1 shows an overall overview of the processes performed by the submodules in the market insight analysis module 121. Specifically, according to an embodiment of the present invention, the flowchart 300 shows the processes or methods performed by the various submodules provided in the market insight analysis module 121 to generate marketing content about a topic. Figure 3 As shown, content from social media 302, public websites 304, search engines 306, and any other external data sources 308 are collected and aggregated by the data collection and preprocessing submodule 122. This information is then provided to the submodule 124, which performs big data analysis 312 on the data to obtain an optimized keyword set 314. The market trend and insight generation submodule 126 then converts the keyword set 314 into ontology sentences 318.

[0053] Figure 4 400 shows an overall overview of the process performed by the submodules in the artificial intelligence content generator module 111. Specifically, the flowchart 400 shows the process or method performed by the submodules provided in the artificial intelligence content generator module 111. Figure 4 As shown, the ontology sentence 402 (as obtained from the marketing insight analysis module 121) is first provided to the artificial intelligence content generator engine 410. Then, the artificial intelligence content generator engine 410 uses the text generation model 412, the image generation model 414, the video generation model 416, the audio generation model 418 and / or other content generation models 419 to process the ontology sentence 402 to form the corresponding artificial intelligence content generator text marketing content 422, artificial intelligence content generator image marketing content 424, artificial intelligence content generator video marketing content 426, artificial intelligence content generator audio marketing content 428 and / or other artificial intelligence content generator related marketing content 429. The detailed working principles of these models are described in the above sections.

[0054] Figure 5An overall overview of the processes performed by the submodules in the cross-platform content publishing module 101 is shown. Specifically, the flowchart 500 shows the process or method performed by the submodules provided in the cross-platform content publishing module 101. The content 422, 424, 426, 428, and 429 are cross-published across multiple digital platforms, such as a social media platform 522, a website 524, a search engine 526, and any other external digital platform 528, at step 510.

[0055] Figure 6 A method 600 of generating marketing content about a topic is shown, wherein the method 600 may be implemented in a computing module, or by modules and / or components in a digital system.

[0056] The method 600 begins at step 602 by receiving a natural language input related to the topic. The received natural language input is then processed to obtain a processing result. The method 600 identifies structured and unstructured data from external sources related to the processing result at step 604. In an embodiment of the present invention, the external source may include a database containing information obtained from a social media platform, information from a public website, and information from a search engine. In other embodiments, the method 600 may crawl the external source using a web crawler module to identify the structured and unstructured data.

[0057] After the structured and unstructured data are identified, the method 600 proceeds to step 606. In step 606, the method 600 uses the first artificial intelligence model to analyze the unstructured data and combines the analyzed unstructured data with the structured data to obtain an optimized keyword set.

[0058] Method 600 utilizes the optimized keyword set and generates a set of ontology sentences related to the topic using the transformer model according to the optimized keyword set and the processing results at step 608. Then, method 600 provides the ontology sentence set to the user module.

[0059] In response to receiving the ontology sentence from the user module in step 610, the method 600 proceeds to step 612. At step 612, the method 600 will continue to generate marketing content about the topic using the artificial intelligence content generator model based on the received ontology sentence.

[0060] In another embodiment, if the generated marketing content is unqualified, the method will be iteratively performed until a marketing content that meets the requirements is generated.

[0061] The present invention is presented above for the purpose of illustration and description, but is not intended to be exhaustive or limiting. Many modifications and variations will be apparent to those of ordinary skill in the art. The exemplary embodiments are selected and described in order to explain the principles and practical applications, and to enable those of ordinary skill in the art to understand the various embodiments of the present invention, with various modifications suitable for the specific use contemplated.

[0062] Therefore, although illustrative example embodiments are described herein with reference to the accompanying drawings, it should be understood that the description is not restrictive and that those skilled in the art may implement various other changes and modifications therein without departing from the scope, innovative concepts and technical solutions of the present invention.

Claims

1. A computing module for generating marketing content about a topic, It is characterized in that The calculation module comprises: processing unit; and a non-transitory medium readable by a processing unit, wherein the medium stores instructions, When executed by the processing unit, the instructions cause the processing unit to perform the following steps: receiving a natural language input related to the topic, and processing the natural language input to obtain a processing result; identifying structured data and unstructured data from external sources that are relevant to the processing results; Analyzing the unstructured data using a first artificial intelligence model, and combining the analyzed unstructured data with the structured data to obtain an optimized keyword set; Using a converter model, generating an ontology sentence set related to the topic according to the optimized keyword set and the processing result, and providing the ontology sentence set to a user module; receiving a selected ontology sentence from the set of ontology sentences from the user module; and Using an artificial intelligence content generator model, marketing content about the topic is generated based on the selected ontology sentences.

2. The computing module according to claim 1, wherein the non-transitory medium further comprises instructions for instructing the processing unit to perform the following steps: Publish the generated marketing content on a digital marketing platform.

3. The computing module of claim 2, wherein the non-transitory medium further comprises instructions for instructing the processing unit to perform the following steps: Obtain feedback related to said marketing content published; and Provide feedback to the training data set of the first artificial intelligence model and the converter model.

4. The computing module according to claim 1, in, The instructions for identifying structured data and unstructured data associated with a processing result from an external source also include instructing the processing unit to perform the following steps: The external source is crawled using a web crawler module to identify the structured data and the unstructured data.

5. The computing module of claim 1, wherein the external source comprises a database comprising information from a social media platform, information from a public website, and information from a search engine.

6. The computing module according to claim 1, wherein the instruction for combining the analyzed unstructured data with the structured data to obtain an optimized keyword set further comprises instructing the processing unit to perform the following steps: Using the first artificial intelligence model, assigning a weight score to each keyword in the structured and unstructured data, in, The weight score is generated according to the spatiotemporal occurrence rate of the keyword.

7. The computing module according to claim 1, wherein the first artificial intelligence model comprises at least one of a natural language processing model, an image processing model and a video analysis model.

8. The computing module of claim 1, wherein the non-transitory medium further comprises instructions for instructing the processing unit to perform the following steps: identifying the digital marketing platform associated with the generation of said marketing content; and The generated marketing content is published to the identified digital marketing platform.

9. The computing module according to claim 1, wherein the ontology sentences include statements for defining the nature of the subject, the concepts of the subject, and the relationship between the optimized keyword sets.

10. A method for generating marketing content about a topic using a computing module, the method include: receiving a natural language input related to the topic, and processing the natural language input to obtain a processing result; identifying structured data and unstructured data from external sources related to the processing results; Analyzing unstructured data using the first artificial intelligence model, and combining the analyzed unstructured data with the structured data to obtain an optimized keyword set; Using a converter model, generating an ontology sentence set related to the topic according to the optimized keyword set and the processing result, and providing the ontology sentence set to a user module; receiving a selected ontology sentence from the user module; as well as Using an artificial intelligence content generator model, marketing content about the topic is generated based on the selected ontology sentences.