Method and system for creating advertisements based on user input

Abstract

The invention relates to a method and a system for creating an advertisement based on user input. The method for creating the advertisement based on the user input comprises the following steps: displaying a link for inputting a user input text comment about a predetermined commodity on a public display device; in response to receiving a user input text comment, the server performs a filtering operation to obtain a filtered input character string; feeding the filtered input string to a generative neural network model, thereby obtaining an advertisement text file; generating a user-related advertisement for the predetermined commodity at least based on the advertisement text file; and sending the user-related advertisement to a public display device for display. In this manner, the resulting user-related advertisement may be more readily accessible to the public because the content is obtained directly from the user's user experience.

Description

Technical Field

[0001] This invention relates to a method and system for creating advertisements based on user input. Background Technology

[0002] In the marketing field, advertising in public places is crucial because people typically spend a significant amount of time in various public spaces (e.g., public transportation, stadiums, restaurants, shopping malls, airports, hotel lobbies, etc.). Therefore, advertising providers are constantly striving to find innovative ways to reach the public more effectively. Generally, each public space can install one or more screens to display various advertisements in addition to other multimedia content. Summary of the Invention

[0003] Therefore, the object of this invention is to provide a method for creating advertisements based on user input.

[0004] Therefore, the present invention provides a user input-based advertising creation method implemented using a server placed in a public place and a public display device. The method includes: the server sending a link to the public display device, the link being associated with an input website for inputting user-inputted text reviews about a pre-ordered product; the public display device displaying the link thereon; in response to receiving user-inputted text reviews from a user device via the input website, the server performing a filtering operation on the user-inputted text reviews to obtain a filtered input string; the server feeding the filtered input string as input to a generative neural network model to obtain an advertising text file as the output of the generative neural network model; the server generating user-related advertisements for the pre-ordered product based at least on the advertising text file, and sending the user-related advertisements to the public display device; and the public display device displaying the user-related advertisements.

[0005] In some embodiments, the method further includes: in response to receiving the link, the public display device encodes the link into a QR code that can be read by the user device, and further displays the QR code, the QR code being a Quick Response (QR) code.

[0006] In some embodiments, the filtering operation includes determining whether the user-input text comment is associated with positive, negative, or mixed emotions, and discarding the user-input text comment associated with the negative emotions.

[0007] In some embodiments, the filtering operation further includes determining whether the user-input text comment is relevant to the user experience of the pre-ordered product, and discarding user-input text comments that are irrelevant to the user experience of the pre-ordered product.

[0008] In some embodiments, the filtering operation further includes segmenting the user-input text comment into multiple segments, and determining for each segment whether the segment is associated with the positive emotion, the negative emotion, or the mixed emotion.

[0009] In some embodiments, the method further includes the following steps: the server receiving a user multimedia file from the user device via the input website, and the creation of the user-related advertisement is also based on the user multimedia file.

[0010] In some embodiments, the server includes a database storing multimedia files of the pre-ordered goods, and the creation of the user-related advertisements is also based on the multimedia files.

[0011] In some embodiments, the method further includes providing a reward token to the user device after creating the user-related advertisement based on the advertisement text file.

[0012] Another objective of this invention is to provide an advertising creation system based on user input.

[0013] Therefore, the present invention provides an advertising creation system based on user input, comprising a server and a public display device placed in a public place and communicating with the server. The system further comprises: the server sending a link to the public display device, the link being associated with an input website for inputting user-inputted text reviews about a pre-ordered product; the public display device displaying the link thereon; in response to receiving user-inputted text reviews from a user device via the input website, the server performing a filtering operation on the user-inputted text reviews to obtain a filtered input string; the server feeding the filtered input string as input to a generative neural network model to obtain an advertising text file as the output of the generative neural network model; the server generating user-related advertisements for the pre-ordered product based at least on the advertising text file, and sending the user-related advertisements to the public display device; and the public display device displaying the user-related advertisements.

[0014] In some embodiments, in response to receiving the link, the public display device encodes the link into a QR code that can be read by the user device, and further displays the QR code, which is a Quick Response (QR) code.

[0015] In some embodiments, the filtering operation includes determining whether the user-input text comment is associated with positive, negative, or mixed emotions, and discarding the user-input text comment associated with the negative emotions.

[0016] In some embodiments, the filtering operation further includes determining whether the user-input text comment is relevant to the user experience of the pre-ordered product, and discarding user-input text comments that are irrelevant to the user experience of the pre-ordered product.

[0017] In some embodiments, the filtering operation further includes segmenting the user-input text comment into multiple segments, and determining for each segment whether the segment is associated with the positive emotion, the negative emotion, or the mixed emotion.

[0018] In some embodiments, the server further receives user multimedia files from the user device via the input website, and the creation of the user-related advertisements is also based on the user multimedia files.

[0019] In some embodiments, the server includes a database storing multimedia files of the pre-ordered goods, and the creation of the user-related advertisements is also based on the multimedia files.

[0020] The beneficial effects of this invention are as follows: Public display devices installed in different public places can provide links for passersby to input information about specific products. Upon receiving a user's input text comment, the server performs a filtering operation to obtain a filtered input string, and feeds this filtered input string as input to a generative neural network model to obtain an advertising text file as the output of the model. The server then generates a user-relevant advertisement for the specific product based at least on the advertising text file. The user-relevant advertisement is then displayed by the public display device. In this way, the resulting user-relevant advertisement is more easily accessible to the public because its content is directly derived from the user's user experience. In some embodiments, by creating more personalized user-relevant advertisements that include user multimedia files, ordinary users can identify the specific product in the user-relevant advertisement. Attached Figure Description

[0021] Other features and effects of the present invention will be clearly presented in the embodiments with reference to the accompanying drawings, wherein:

[0022] Figure 1 A block diagram illustrating an embodiment of the present invention for creating an advertisement based on user input is provided.

[0023] Figure 2 A flowchart illustrating the steps of a method for creating an advertisement based on user input according to an embodiment of the present invention is provided.

[0024] Figure 3 An illustrative description of the input website for the described embodiment;

[0025] Figure 4 The filtering neural network model for filtering operations in the embodiments described are illustrated by way of example;

[0026] Figure 5 The generative neural network model of the embodiment is illustrated by way of example; and

[0027] Figures 6 to 8 This example illustrates the display of user-related advertisements on a public display device. Detailed Implementation

[0028] Before describing the invention in more detail, it should be noted that, where deemed appropriate, reference numerals are repeatedly used in the drawings to indicate corresponding or similar components, which may have similar characteristics.

[0029] It should be noted that throughout this disclosure, spatial relative terms such as “up,” “down,” “left,” and “right” are used herein to distinguish the relative spatial positions of the different elements shown in the figures of this disclosure, and are intended to cover different orientations of elements other than those depicted in the figures.

[0030] Before the invention is described in detail, it should be noted that similar components are represented by the same numbers in the following description.

[0031] Figure 1 This is a block diagram illustrating a user-input-based advertising creation system 100 according to an embodiment of the present invention. The user-input-based advertising creation system 100 includes a server 110 and a common display device 160 communicating with the server 110.

[0032] In this embodiment, the server 110 includes a processor 112, a data storage unit 114, and a communication unit 116.

[0033] The processor 112 may use a central processing unit (CPU), microprocessor, microcontroller, single-core processor, multi-core processor, dual-core mobile processor, digital signal processor (DSP), field-programmable gate array (FPGA), application-specific integrated circuit (ASIC), radio frequency integrated circuit (RFIC), etc.

[0034] The data storage unit 114 is connected to the processor 112 and can use, for example, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), firmware, flash memory, etc. In this embodiment, the data storage unit 114 stores software applications and multiple neural network models. The software applications include instructions that, when executed by the processor 112, cause the processor 112 to perform the operations described below. Each neural network model can use a convolutional neural network (CNN), a recurrent neural network (RNN), or other forms of neural network and can be pre-trained to perform the specific operations described below. In some embodiments, the data storage unit 114 may include a database storing multiple preset multimedia files. Each preset multimedia file may be a pre-existing advertisement for a predetermined product.

[0035] It is worth noting that throughout this disclosure, the term "goods" can refer to the brand of products or services offered by a company or store. Typically, advertising is for the purpose of promoting one or more specific goods.

[0036] The communication unit 116 is connected to the processor 112 and may include one or more of the following: a radio frequency integrated circuit (RFIC), a near-field wireless communication module supporting near-field wireless communication networks using wireless technologies such as Bluetooth and / or Wi-Fi, and a mobile communication module supporting telecommunications using Long Term Evolution (LTE), 3G, 4G, or 5G wireless mobile telecommunications technologies. The communication unit 116 enables the server 110 to communicate with the public display device 160 via a wireless network (e.g., the Internet).

[0037] The public display device 160 may use a display screen, projector, or other device capable of displaying information thereon. In some embodiments, the public display device 160 is installed in a public place (which may be a place where a large number of people may pass through or gather, such as public transportation, stadiums, restaurants, shopping malls, airports, hotel lobbies, etc.). It is worth noting that in other embodiments, the server 110 may communicate with multiple public display devices 160 simultaneously, each of which uses one of the aforementioned devices and is installed in a specific public place.

[0038] The public display device 160 includes a processor 162, a data storage unit 164, a communication unit 166, and a display screen 168.

[0039] The processor 162 can be implemented using the same components as the processor 112. The data storage unit 164 is connected to the processor 162, can be implemented using the same components as the data storage unit 114, and stores one or more output multimedia files. The communication unit 166 is connected to the processor 162 and can use the same components as the communication unit 116. The display screen 168 is connected to the processor 162 and is controlled by the processor 162 to display the multimedia files. In some embodiments, the public display device 160 further includes an audio output component to output the audio portion of the multimedia files. In some examples, the multimedia files may be advertising videos, and the data storage unit 164 may store multiple multimedia files, and the processor 162 may control the display screen 168 to cycle through the multimedia files. Each advertising video may be associated with a product.

[0040] Figure 2 This is a flowchart illustrating the steps of a method for creating an advertisement based on user input, according to an embodiment of the present invention. In this embodiment, the method employs... Figure 1 The user-input-based advertising creation system 100 shown is implemented to achieve this.

[0041] When a group associated with server 110 (e.g., an advertising company) intends to create user-related advertisements for a predetermined merchant or product, the group can operate server 110 to control processor 112 to generate a link associated with an input website in step 202. In step 204, the link is sent to the public display device 160 via communication unit 116. The link (e.g., a URL) is associated with an input website for entering user-input text reviews about the predetermined product.

[0042] In step 206, in response to receiving the link, the processor 162 of the public display device 160 controls the display screen 168 to display the link. It is worth noting that in some embodiments, the processor 112 may further send one of the preset multimedia files stored in the data storage unit 114 to the public display device 160 via the communication unit 116, and the link may be displayed together with one of the preset multimedia files serving as an advertisement for the predetermined product.

[0043] In some embodiments, in addition to the link, the processor 162 may further encode the link as a QR code (e.g., a Quick Response (QR) code, a barcode, etc.) and control the display screen 168 to further display the QR code. The QR code can be read by an electronic device using a camera (and optionally, a code reading software application), while the link can be seen by a passerby without an electronic device equipped with a camera, or can be manually entered by the passerby.

[0044] In some embodiments, the processor 162 of the public display device 160 controls the display screen 168 to further display a text message encouraging passersby to interact with the link or the QR code. In some embodiments, the text message may include, for example, “Tell us about your experience using this product! You may win a prize if your review is selected.” In some embodiments, the processor 162 may control the audio output unit to output an audio file that can read the text message aloud. That is, the public display device 160 installed in a public place can be used to attract passersby and encourage them to use the link or the QR code to access the input website.

[0045] In step 208, when the passerby intends to access the input website, he / she can operate the user device 180 (e.g., smartphone, tablet, laptop, etc.) to use the camera 184 to read the QR code, thereby launching a browser to access the input website, or manually enter the link in the browser. After the passerby successfully accesses the input website, the processor 182 downloads the input website and controls the display screen 186 to display the input website.

[0046] Figure 3 An input website is illustrated by way of example according to an embodiment of the present invention. In this embodiment, the input website may include multiple fields, each field allowing the passerby to input specific information. Figure 3 In one example, the input website includes a field for users to input text comments, and one or more fields for inputting the passerby's contact information (e.g., email address, phone number, etc.). In other embodiments, the input website includes additional fields for inputting other information.

[0047] The user-input text comment can indicate how the passerby views the pre-ordered product and may include one or more sentences. After the user completes the field, he / she can press a button (e.g., a confirmation button) to submit the user-input text comment and the contact information to the server 110.

[0048] In step 210, in response to receiving the user-input text comment from the user device 180 via the input website, the server 110 performs a filtering operation on the user-input text comment to obtain a filtered input string that will be used as a subsequent advertisement creation.

[0049] More specifically, in some embodiments, the filtering operation may include the processor 112 applying the user-input text comments as input to a pre-trained filtering neural network model to filter out some of the user-input text comments that are deemed unsuitable for generating user-related advertisements, and obtaining the filtered input string.

[0050] Figure 4 A filtering neural network model for the filtering operation is exemplarily described according to an embodiment of the present invention. In this embodiment, the filtering neural network model 400 may use the functionality provided by a basic model, such as the GPT-n model developed by OpenAI, Inc., and includes some mechanisms trained to achieve the following functionality. In some embodiments, the filtering neural network model 400 may use one or more of recurrent neural networks (RNNs), long short-term memory neural networks, convolutional neural networks (CNNs), etc. It is worth noting that the operation of training the filtering neural network model 400 implemented using the above-described neural networks is readily available in related technologies, and therefore, for the sake of brevity, its details are omitted herein.

[0051] See Figure 4 In this embodiment, in response to the received user-input text comment, the filtering neural network model 400 includes an emotion mechanism 410 configured to determine whether the user-input text comment is associated with positive emotion, negative emotion, neutral or ambiguous emotion (i.e., emotion that is difficult to definitively categorize as positive or negative), or mixed emotion (i.e., the same user-input text comment containing both positive and negative emotions). If a user-input text comment is associated with the negative emotion, the user-input text comment can be discarded. If a user-input text comment is associated with the positive emotion, neutral or ambiguous emotion, or mixed emotion, the user-input text comment can be retained for further processing.

[0052] Table 1 below lists some exemplary user-input text comments from different passersby (hereinafter referred to as "users"), and the sentiment determination of each of the user-input text comments.

[0053] Table 1

[0054]

[0055]

[0056] Subsequently, in the case of a user-input text comment containing the aforementioned mixed emotions, the filtering neural network model 400 can apply the user-input text comment with the mixed emotions to the partitioning mechanism 420, thereby segmenting the user-input text comment into multiple fragments. For example, the user-input text comment "The food was good, but the service was slow" can be segmented into separate fragments, "The food was good" and "The service was slow". Then, the fragments are brought back to the emotion mechanism 410 for individual processing, wherein the fragments associated with negative emotions (i.e., slow service) are discarded.

[0057] In some embodiments, in addition to the user-input text comment having the mixed emotion, if the user-input text comment comprises multiple sentences or a long paragraph, the processor 112 may first apply the partitioning mechanism 420 to the user-input text comment to segment the user-input text comment into multiple distinct segments. Then, the processor 112 feeds each segment as input to the filtering neural network model 400, and the emotion mechanism 410 can process each segment in the manner described above. In this configuration, the chance of obtaining user-input text comments with the mixed emotion may be reduced.

[0058] Then, the user-input text comments or the fragments that were not discarded are fed to the relevance detection mechanism 430 to determine whether the user-input text comments or the fragments are related to the user experience of the pre-ordered product.

[0059] For example, in one embodiment, the pre-ordered product may be a product sold in a physical store, and the user-input text comments collected in Table 1 may be processed to determine whether the user-input text comments are meaningful in the context of the pre-ordered product. If it is determined that the user-input text comments or the fragments are indeed relevant to the user experience of the pre-ordered product, the user-input text comments or the fragments are then output as a filter input string, which can be used to generate user-related advertisements.

[0060] Table 2 below illustrates some user-input text comments with the aforementioned positive sentiment processed by the relevance detection mechanism 430. In Table 2, a yes / no question is asked to determine whether the user-input text comment or the fragment can be logically linked to 'shopping at Acme supermarket'.

[0061] Table 2

[0062]

[0063]

[0064] As can be seen from Table 2, the user-input text comments "The food is good" and "Your joke is funny" cannot be logically combined with the shopping experience and will therefore be discarded. Other user-input text comments will be output as the filtered input string.

[0065] Table 3 below illustrates some user-input text comments with neutral or ambiguous emotions processed by the relevance detection mechanism 430. In Table 3, a yes-or-no question is asked to determine whether the user-input text comment or the fragment can logically be combined with 'shopping at Acme supermarket'. If a combination is possible, a resulting combined sentence is fed back to the emotion mechanism 410, where the combined sentence associated with negative emotions is discarded.

[0066] Table 3

[0067]

[0068] Table 4 below illustrates the combined sentences processed by the emotion mechanism 410. If the emotion is determined to be positive, the combined sentence is then output as a filtered input string.

[0069] Table 4

[0070]

[0071] After completing the filtering operations described above, some filtered input strings can be used to generate the user-related advertisements. Therefore, the process proceeds to step 212, where the processor 112 of the server 110 uses the filtered input strings as input to a generative neural network model to obtain an advertisement text file as the output of the generative neural network model.

[0072] Figure 5 A generative neural network model 500 is illustrated exemplaryly according to an embodiment of the present invention. In this embodiment, the generative neural network model 500 may use a commercially available generative neural network model (e.g., ChatGPT), but is not limited thereto.

[0073] In some embodiments, step 212 can be implemented by applying the filtered input string as input to the generating neural network model, accompanied by the prompt "Please prepare a poem or prose about your shopping experience at Acme supermarket. The following filtered input string must be included."

[0074] In one example, the product is related to a specific restaurant (e.g., The food is processed using the above process to obtain the filtered input strings, "I like this food", "I like this taste", and "The food is good".

[0075] In this example, the advertising text file output by the generative neural network model 500 might be: "I like this food, and I like the taste. I can't deny that McDonald's is a place that makes me happy. The food is good, and McDonald's is great with many choices. There's no reason to hesitate, why not stop by and enjoy it? I guarantee you'll love it, and it will be a delicious feast." It's worth noting that the resulting advertising text file output by the generative neural network model 500 may differ depending on different prompts, different filtered input strings, and different products.

[0076] In another example, the goods are related to a specific airline (e.g., The aviation services provided are obtained using the above process, and the filtered input strings are "I feel very high", "I am very happy" and "You make me happy".

[0077] In this example, the advertising text file output by the generative neural network model 500 might read: "I fly with joy, happy and carefree. With you, my heart is full. You make me happy and I'm never bored. Darling, let's take an Emirates flight together and fly to the boundless horizon. We will see the world together. Our love is forever open."

[0078] In another example, the product is related to a particular store (e.g., Shopping, and using the above process to obtain the filtered input strings, "This is durable", "This is comfortable", "This is affordable" and "Shopping at IKEA".

[0079] In this example, the advertising text file output by the generative neural network model 500 might be "At IKEA, you will find unique, durable, comfortable, authentic and affordable furniture that suits you and me."

[0080] When a large number of user-input text comments are received, and a large number of filter input strings are generated accordingly, the filter input strings can be divided into several groups, each group containing a smaller number of filter input strings. Therefore, the operation in step 212 can then be performed on each group to generate multiple advertising text files.

[0081] In step 214, after the advertising text file is generated, the processor 112 of the server 110 continues to generate user-related advertisements for the predetermined product based at least on the advertising text file, and transmits the user-related advertisements to the public display device 160.

[0082] Specifically, in some embodiments, the user-related advertisement may include the advertisement text file and the logo of the related product. Alternatively, the user-related advertisement may include the advertisement text file and existing advertisements for the related product.

[0083] In some embodiments, for each product, the database of the data storage unit 114 can store multiple product multimedia files, such as video files, audio files, and images related to the product. Each product multimedia file is associated with one or more related words. Therefore, when the filtering input string includes related words of one of the product multimedia files, that product multimedia file can be used to generate the user-related advertisement. In the case of multiple advertisement text files being generated, multiple user-related advertisements can also be created for the same related product.

[0084] In step 216, after the user-related advertisement is generated, the processor 112 of the server 110 transmits the user-related advertisement to the public display device 160, and in response to receiving the user-related advertisement, the public display device 160 may display the user-related advertisement thereon.

[0085] In one example Figure 6 User-related advertisements are exemplarily displayed on the public display device 160. Figure 6 In the example, the user-related advertisement includes the logo of the company offering the goods (e.g., ( ), Images depicting specific product promotions and the accompanying advertising text files.

[0086] In another example, Figure 7 User-related advertisements are exemplarily displayed on the public display device 160. Figure 7 In the example, the user-related advertisement includes the logo of the company offering the goods (e.g., (This includes) multiple images promoting the company's services and the aforementioned advertising text file.

[0087] In some embodiments, the input website may further utilize the user device 180 to enable the user to upload user multimedia files (e.g., user photos, user-recorded videos or audio describing user experience, etc.), and the server 110 may store the user multimedia files in the database of the data storage unit 114. Therefore, after generating the advertising text file, the user multimedia files can be further used to create user-related advertisements. In some embodiments, the input website may request the user to upload the user multimedia files.

[0088] In response to receiving the user multimedia file, the public display device 160 may first process the user multimedia file. In some embodiments, the input website may embed instructions to process the user multimedia file. In some examples where the user multimedia file is an image, the image may be cropped so that only the user's face is displayed. In the case where multiple faces are detected, the image may be cropped so that only the faces are displayed. The image may be applied to a filter to determine whether inappropriate content exists. In the case of multiple images uploaded, the input website may select to use one of the images.

[0089] Figure 8 User-related advertisements are exemplarily displayed on the public display device 160. Figure 8 In the example, the user-related advertisement includes the logo of the company offering the product (e.g., The advertising text file and the image of the user who provides the user-input text comment used to generate the advertising text file.

[0090] In some embodiments, in step 218, after successfully creating the user-related advertisement, the processor 112 of the server 110 may continue to provide reward tokens to the user device 180, whose owner provides the user-input text comment used to generate the advertisement text file. The reward tokens can then be redeemed to purchase the product related to the user-related advertisement. Therefore, the product can be further promoted to users who already have a positive feeling about the product, in a manner similar to a loyalty program.

[0091] In some embodiments, upon detecting a user-entered text comment with negative sentiment, the processor 112 of the server 110 may send a message to the associated user (if the user provides his / her email address or phone number) requesting further feedback regarding his / her user experience. Thus, the company providing the product can become aware of potential problems with the product.

[0092] In summary, embodiments of the present invention provide a method and system for creating advertisements based on user input. In the method, public display devices 160 installed in different public places can provide links for passersby to input information about specific products. Upon receiving a user-input text comment, server 110 performs a filtering operation to obtain a filtered input string, and feeds the filtered input string as input to a generative neural network model 500 to obtain an advertisement text file as the output of the generative neural network model 500. Then, server 110 generates user-related advertisements for the specific product based at least on the advertisement text file. The user-related advertisements are then displayed by the public display devices 160. In this way, the resulting user-related advertisements are more easily accessible to the public because the content is directly derived from the user's user experience. In some embodiments, the present invention achieves its objectives by creating more personalized user-related advertisements that include user multimedia files, enabling ordinary users to identify the specific product in the user-related advertisements.

[0093] However, the above description is merely an embodiment of the present invention and should not be construed as limiting the scope of the present invention. Any simple equivalent changes and modifications made in accordance with the claims and description of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for creating advertisements based on user input, wherein the method is implemented using a server placed in a public place and a public display device; characterized in that, The method includes: The server sends a link to the public display device, the link being associated with an input website for entering user-inputted text reviews about pre-ordered goods; The links thereon are displayed on the public display device; In response to receiving a user-input text comment from a user device via the input website, the server performs a filtering operation on the user-input text comment to obtain a filtered input string; The server feeds the filtered input string as input to the generative neural network model, thereby obtaining an advertising text file as the output of the generative neural network model; The server generates user-related advertisements for the pre-ordered product based at least on the advertisement text file, and sends the user-related advertisements to the public display device; as well as The user-related advertisements are displayed on the public display device.

2. The method for creating advertisements based on user input according to claim 1, characterized in that, The method further includes: In response to receiving the link, the public display device encodes the link into a QR code that can be read by the user device, and further displays the QR code, which is a quick-response code.

3. The method for creating advertisements based on user input according to claim 1, characterized in that, The filtering operation includes determining whether the user-input text comment is associated with positive, negative, or mixed emotions, and discarding user-input text comments associated with negative emotions.

4. The method for creating advertisements based on user input according to claim 3, characterized in that, The filtering operation further includes determining whether the user's input text review is relevant to the user experience of the pre-ordered product, and discarding reviews related to the pre-ordered product. (BCPI-240228, page 2 / 3) User-input text comments are irrelevant to the user experience.

5. The method for creating advertisements based on user input according to claim 3, characterized in that, The filtering operation further includes segmenting the user-input text comment into multiple segments, and determining for each segment whether the segment is associated with the positive emotion, the negative emotion, or the mixed emotion.

6. The method for creating advertisements based on user input according to claim 1, characterized in that, The method further includes the following steps: the server receives user multimedia files from the user device through the input website, and the creation of user-related advertisements is also based on the user multimedia files.

7. The method for creating advertisements based on user input according to claim 1, characterized in that, The server includes a database storing multimedia files of the pre-ordered goods, and the creation of user-related advertisements is also based on the multimedia files of the goods.

8. The method for creating advertisements based on user input according to claim 1, characterized in that, The method further includes providing a reward token to the user device after creating the user-related advertisement based on the advertisement text file.

9. A user-input-based advertising creation system, the user-input-based advertising creation system comprising a server and a public display device placed in a public place and communicating with the server, characterized in that: The server sends a link to the public display device, the link being associated with an input website for entering user-inputted text reviews about pre-ordered goods; The public display device displays the links thereon; In response to receiving a user-input text comment from a user device via the input website, the server performs a filtering operation on the user-input text comment to obtain a filtered input string; The server feeds the filtered input string as input to the generative neural network model, thereby obtaining an advertising text file as the output of the generative neural network model; The server, at least based on the advertising text file, produces BCPI-240228 (page 3 / 3). Generate user-related advertisements and send the user-related advertisements to the public display device; and The public display device displays advertisements related to the user.

10. The advertising creation system based on user input according to claim 9, characterized in that: In response to receiving the link, the public display device encodes the link into a QR code that can be read by the user device, and further displays the QR code, which is a quick-response code.

11. The advertising creation system based on user input according to claim 9, characterized in that, The filtering operation includes determining whether the user-input text comment is associated with positive, negative, or mixed emotions, and discarding user-input text comments associated with negative emotions.

12. The advertising creation system based on user input according to claim 11, characterized in that, The filtering operation further includes determining whether the user-input text comment is relevant to the user experience of the pre-ordered product, and discarding user-input text comments that are irrelevant to the user experience of the pre-ordered product.

13. The advertising creation system based on user input according to claim 11, characterized in that, The filtering operation further includes segmenting the user-input text comment into multiple segments, and determining for each segment whether the segment is associated with the positive emotion, the negative emotion, or the mixed emotion.

14. The advertising creation system based on user input according to claim 9, characterized in that, The server further receives user multimedia files from the user device through the input website, and the creation of user-related advertisements is also based on the user multimedia files.

15. The advertising creation system based on user input according to claim 9, characterized in that, The server includes a database storing multimedia files of the pre-ordered goods, and the creation of user-related advertisements is also based on the multimedia files of the goods.

Images