Tourism service information push method and system based on artificial intelligence
Through artificial intelligence models, the user portraits are constructed and original video information is screened, which solves the problems of inappropriate and difficult to judge the authenticity of video push in the existing technology, and realizes efficient and personalized travel service information push, improving user experience and information Reference value.
Patent Information
- Application Number
- CN202411156469.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The prior art is difficult to push videos to suitable users, and since the video content is difficult to judge its authenticity after editing and splicing, the reference value of the video is reduced.
Use artificial intelligence models to build user portraits, and filter the recommended information that matches the user portrait, push the original information to the user, and use TF-IDF, Apriori algorithm, YOLOv3 model and CNN model to identify and filter original videos information.
It improves the reference value and user experience of video push, ensures that the information pushed is original and meets user preferences, reduces user annoyance, and improves the accuracy of information and user attention.
Smart Images

Figure CN118694818B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and system for pushing tourism service information based on artificial intelligence. Background Art
[0002] In today's society, tourism has become an important part of people's lives. With the development of the Internet and artificial intelligence technology, it has become more convenient to obtain tourism service information. Various tourism service agencies push tourism product information, promotional videos and other service information to users to enhance users' travel interest and experience. At the same time, the development of self-media has made video sharing an important source of tourism service information. Through the video information shared by other users, viewers can intuitively understand the situation of the corresponding tourist attractions as a reference for their own travel plans.
[0003] However, how to push videos to the right users and ensure that the pushed videos have reference value is an issue that each tourism service agency needs to consider. In terms of user matching, a common method is to match the corresponding video content by building user portraits. However, after some video content has been edited and spliced, it is difficult to judge its authenticity, which reduces the reference value of the video. Summary of the invention
[0004] The present invention uses an artificial intelligence model to build a user's character portrait, matches it with corresponding push information, further screens the push information, and pushes original information to the user, so as to solve the problems raised in the background technology.
[0005] The technical solution provided by the present invention is: a method for pushing tourism service information based on artificial intelligence, characterized in that the method comprises:
[0006] Collect user information and build user portraits based on the collected user information;
[0007] Obtain the information to be recommended, extract key features from the information to be recommended, match it with the user profile, and form an original information database with the information to be recommended that matches the user profile;
[0008] Screening the original information database, dividing the original information database into a valid information set and multiple invalid information sets, pushing the information to be recommended in the valid information set to the user terminal, and receiving feedback;
[0009] The information to be recommended includes one or more of video information, text information and audio information, the valid information set includes one or more of original video, original text and original audio, and the invalid information set includes one or more of non-original video, non-original text and non-original audio.
[0010] Preferably, the collecting of user information and constructing a user portrait based on the collected user information includes:
[0011] Collect users' personal information, browsing history data, and purchase history data through API calls or crawler technology;
[0012] Clean and pre-process the collected personal information, browsing history, and purchase history data to form an original user information database;
[0013] Extract user key features from the original user information database to form a user key feature set; specifically:
[0014] Extract the text data browsed by users from the browsing history data, and use the TF-IDF algorithm to extract keywords to obtain the key features of user browsing preferences;
[0015] Using the Apriori algorithm, we analyze the frequently appearing keywords of tourist attractions from the user's browsing history data, thereby inferring the user's tourist attraction preferences and obtaining the key features of the user's tourist attraction preferences;
[0016] The key features of user browsing preferences and the key features of user tourist attraction preferences are combined into a user key feature set;
[0017] Use machine learning algorithms to perform cluster analysis on key user features, divide users into different user groups, and build user portraits for them;
[0018] The personal information includes name, age, and occupation; the browsing history data is the user's browsing behavior on the travel website, including browsing time, travel location, and browsing duration; the purchase history is the user's purchase record on the travel website, including travel package information and reserved air ticket information.
[0019] Preferably, the method further includes constructing a tourism preference feature set, which specifically includes the following steps:
[0020] Extract browsing time features;
[0021] Extract the user's historical browsing timestamp from the browsing history;
[0022] Use the pandas library to convert the browsing timestamp into a time format, extract the hour, week, and month features, and convert them into integers respectively, using the hour feature as the time feature ;
[0023] Get multiple time features The average time characteristics of ;in:
[0024] ,in is the number of views;
[0025] Extracting tourist location features;
[0026] Extracting text data of the user’s travel locations from browsing history data;
[0027] The tourist locations are converted into numerical features through one-hot encoding, specifically:
[0028] Convert text data into numerical features using the bag-of-words model or TF-IDF;
[0029] Extract browsing time features:
[0030] Extract browsing time from browsing history data ;
[0031] Convert browsing time to minutes;
[0032] Check out outliers and remove them; the outliers are browsing durations less than 0 or greater than 10,000 minutes;
[0033] The browsing time characteristics, travel location characteristics and average time characteristics form the travel preference feature set A.
[0034] Preferably, the obtaining of the information to be recommended and extracting key features from the information to be recommended includes:
[0035] Use crawler technology to obtain travel video information shared by other users from the Internet;
[0036] Extract key video features from travel videos, including: video theme, copyright statement, scenic spots, and shooting time; specifically:
[0037] The pre-trained YOLOv3 model is used to loop through the video frames of the travel video, detect objects in the video, and pre-add category name mapping to make the YOLOv3 model output object names;
[0038] Use OCR technology to extract text from the video and generate text format data from the detected text;
[0039] Extract scenic spot names, copyright notices, and time keywords from text format data through keyword algorithms;
[0040] The extracted keywords form a keyword database, and the keyword database is sent to the cloud for storage.
[0041] Preferably, extracting key features from the information to be recommended, matching them with the user portrait, and forming an original information database with the information to be recommended that matches the user portrait includes:
[0042] Use TF-IDF to vectorize keywords and user key features;
[0043] Calculate the similarity between the user's key feature vector and the keyword vector through the pandas library, output the similarity score, and sort the similarity scores in ascending order;
[0044] Sort the videos in descending order according to the similarity scores and select the top travel videos, submit similarity score labels for each travel video, and select The travel videos constitute the original information database.
[0045] Preferably, the screening of the original information database to divide the original information database into a valid information set and an invalid information set includes:
[0046] Acquire tourism video information from the original information database;
[0047] Extract multiple video frames from the travel video through a sliding window;
[0048] Input the video frames into the trained CNN model, and use the CNN model to compare the video frames with the images in the original video library to identify whether the travel video is an original video;
[0049] The travel videos identified as original videos are classified as valid information sets; and the travel videos identified as non-original videos are classified as invalid information sets.
[0050] Preferably, the step of inputting the video frame into a trained CNN model, comparing the video frame with images in an original video library through the CNN model, and identifying whether the travel video is an original video comprises the following steps:
[0051] Use TensorFlow or PyTorch to build a CNN model; the CNN model includes multiple convolutional layers, pooling layers, and fully connected layers;
[0052] Acquire multiple original image data from an original video library;
[0053] Acquire image data of a plurality of corresponding non-original videos from the Internet;
[0054] The original image data and the non-original video image data form an image sample set, and the image sample set is divided into a training set and a validation set;
[0055] Use the training set to train the CNN model, use the cross entropy loss function and optimizer to minimize the loss of the CNN model, and the output of the CNN model is the probability value of the image being the original video. ;
[0056] Set comparison threshold , when | - |≥0, it is considered to be an original video;
[0057] The obtained multiple video frames are input into the trained CNN model to determine whether it is an original video.
[0058] Preferably, the step of pushing the information to be recommended in the valid information set to the user terminal and receiving feedback includes:
[0059] Arrange the information to be recommended in the valid information set in descending order according to the similarity scores to form a recommended information sequence;
[0060] Setting recommendation thresholds , Push time period ( , ), push time interval , , Indicates the push start time point. Indicates the push end time point;
[0061] choose Similarity score greater than The recommended information is Time is greater than When the user sends the recommended information, the user receives the feedback information, and sends the received feedback information to the cloud storage; ,in, , is the time parameter, ;if , , then the The video information is returned to the valid information set and inserted at the first place of the recommended information sequence to ensure that it will be given priority in the next recommendation.
[0062] The present invention also provides a tourism service information push system based on artificial intelligence, and the system is used to execute the tourism service information push method based on artificial intelligence.
[0063] The present invention also provides a computer-readable storage medium, which stores a computer program. The computer program is executed by a processor to implement the artificial intelligence-based tourism service information push method.
[0064] Beneficial effects of the present invention:
[0065] 1. In the present invention, the collected personal information, browsing history and purchase history data are cleaned and preprocessed to form an original user information database; the travel video information is extracted from the original information database; a plurality of video frames are extracted from the travel video using a sliding window; the video frames are imported into a trained CNN model, and the video frames are compared with images in an original video library using the CNN model to identify whether the travel video is an original video; the travel videos identified as original are divided into a valid information set; the non-original travel videos are divided into an invalid information set, and when information is pushed, only the video data in the valid information set is screened to ensure the reference value of the pushed video.
[0066] 2. The present invention arranges the information to be recommended in the valid information set in descending order according to the similarity score to form a recommended information sequence; determines which information to push first; and sets a recommendation threshold , Push time period ( , ), push time interval , , set a reasonable push interval to avoid disturbing users too frequently. Too frequent pushes may cause users to feel annoyed, thereby reducing their use of the platform; analyze users' past browsing data to determine their active time periods, and push information during these time periods to increase users' attention and acceptance. BRIEF DESCRIPTION OF THE DRAWINGS
[0067] Figure 1 The present invention is a flow chart of a method for pushing tourism service information based on artificial intelligence. DETAILED DESCRIPTION
[0068] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles of the present invention defined in the following description can be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not deviate from the spirit and scope of the present invention.
[0069] It is to be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element may be one, while in another embodiment, the number of the element may be multiple, and the term "one" should not be understood as a limitation on the quantity.
[0070] Please combine Figure 1 The present invention provides a tourism service information push method based on artificial intelligence, the method comprising: collecting user information, and constructing a user portrait based on the collected user information, specifically comprising the following steps:
[0071] 1. Collect users’ personal information, browsing history data, and purchase history data through API calls or crawler technology;
[0072] The personal information includes name, age, and occupation; the browsing history data is the browsing behavior of the user on the travel website, including browsing time, travel location, and browsing duration; the purchase history is the purchase record of the user on the travel website, including travel package information and reserved air ticket information;
[0073] Clean and pre-process the collected personal information, browsing history, and purchase history data to form an original user information database;
[0074] Extract user key features from the original user information database to form a user key feature set; specifically:
[0075] Extract the text data browsed by users from the browsing history data, and use the TF-IDF algorithm to extract keywords to obtain the key features of user browsing preferences;
[0076] Using the Apriori algorithm, we analyze the frequently appearing keywords of tourist attractions from the user's browsing history data, thereby inferring the user's tourist attraction preferences and obtaining the key features of the user's tourist attraction preferences;
[0077] The key features of user browsing preferences and the key features of user tourist attraction preferences are combined into a user key feature set.
[0078] Use machine learning algorithms to perform cluster analysis on key user features, divide users into different user groups, and build user portraits for them.
[0079] 2. Obtain the information to be recommended, extract key features from the information to be recommended, match it with the user portrait, and form the original information database with the information to be recommended that matches the user portrait. The specific steps include:
[0080] The step of obtaining the information to be recommended and extracting key features from the information to be recommended includes:
[0081] Use crawler technology to obtain travel video information shared by other users from the Internet; extract key video features from travel videos, the key video features include: video theme, copyright statement, scenic spots, shooting time; specifically:
[0082] The pre-trained YOLOv3 model is used to loop through the video frames of the travel video, detect objects in the video, and pre-add category name mapping to make the YOLOv3 model output object names;
[0083] Use OCR technology to extract text from the video, and generate text format data from the detected text; extract attraction names, copyright statements, and time keywords from the text format data through keyword algorithms; construct a keyword database with the extracted keywords, and send the keyword database to cloud storage.
[0084] Use TF-IDF to vectorize keywords and user key features; use the pandas library to calculate the similarity between the user key feature vector and the keyword vector, output the similarity score, and sort the similarity scores in ascending order; sort the videos in descending order according to the similarity score sorting, and select the top videos. travel videos, submit similarity score labels for each travel video, and select The travel videos constitute the original information database.
[0085] 3. Screening the original information database, dividing the original information database into a valid information set and multiple invalid information sets, pushing the recommended information in the valid information set to the user terminal, and receiving feedback; specifically including the following steps:
[0086] Acquire tourism video information from the original information library; extract multiple video frames from the tourism video through a sliding window; input the video frames into the trained CNN model, compare the video frames with the images in the original video library through the CNN model, and identify whether the tourism video is an original video; classify the tourism videos identified as original into a valid information set; and classify the non-original tourism videos into an invalid information set.
[0087] In some embodiments, distinguishing whether the travel video information is an original video by using a CNN model can be achieved by the following steps:
[0088] The video frame is input into the trained CNN model, and the video frame is compared with the image in the original video library by the CNN model to identify whether the travel video is an original video, including the following steps:
[0089] Use TensorFlow or PyTorch to build a CNN model; the CNN model includes multiple convolutional layers, pooling layers, and fully connected layers;
[0090] Acquire multiple original image data from an original video library;
[0091] Acquire image data of a plurality of corresponding non-original videos from the Internet;
[0092] The original image data and the non-original video image data form an image sample set, and the image sample set is divided into a training set and a validation set;
[0093] Use the training set to train the CNN model, use the cross entropy loss function and optimizer to minimize the loss of the CNN model, and the output of the CNN model is the probability value of the image being the original video. ;
[0094] Set comparison threshold , when | - |≥0, it is considered to be an original video;
[0095] The obtained multiple video frames are input into the trained CNN model to determine whether it is an original video.
[0096] In some preferred embodiments, the method further includes constructing a travel preference feature set, which specifically includes the following steps:
[0097] Extract browsing time features;
[0098] Extract the user's historical browsing timestamp from the browsing history;
[0099] Use the pandas library to convert the browsing timestamp into a time format, extract the hour, week, and month features, and convert them into integers respectively, using the hour feature as the time feature ;
[0100] Get multiple time features The average time characteristics of ;in:
[0101] ,in is the number of views;
[0102] Extracting tourist location features;
[0103] Extracting text data of the user’s travel locations from browsing history data;
[0104] The tourist locations are converted into numerical features through one-hot encoding, specifically:
[0105] Convert text data into numerical features using the bag-of-words model or TF-IDF;
[0106] Extract browsing time features:
[0107] Extract browsing time from browsing history data ;
[0108] Convert browsing time to minutes;
[0109] Check out outliers and remove them; the outliers are browsing durations less than 0 or greater than 10,000 minutes;
[0110] The browsing time characteristics, travel location characteristics and average time characteristics form the travel preference feature set A.
[0111] After two steps of judging whether the travel video information is consistent with the user portrait (may be consistent with the user's travel preferences) and the credibility of the travel video (whether it is an original video), it is necessary to establish a push rule to push the video information to the user.
[0112] In this embodiment, the following steps are implemented:
[0113] First, the information to be recommended in the valid information set is arranged in descending order according to the similarity score to form a recommended information sequence to determine which information to push first;
[0114] Setting recommendation thresholds , Push time period ( , ), push time interval , According to the user's browsing preferences, we choose to push information when the user is browsing travel information, so as to achieve accurate push in time. No information will be pushed at other times, which will reduce the disturbance to the user and reduce the probability of user resistance.
[0115] choose Similarity score greater than The recommended information is Time is greater than When the user receives the recommended information, the recommended information is sent to the user and feedback information is received; the feedback information includes the user's click operation, playback frequency, etc., which can be obtained from the operation log;
[0116] Send the received feedback information to the cloud storage for system backup and later analysis and summary;
[0117] in , , , browsing time hours; if , , then the The video information is returned to the valid information set and inserted at the top of the recommended information sequence to ensure that this information is recommended first in the next information push. In this way, after the push time period has passed, the information will no longer be pushed, even if there is information that has not been pushed yet; the unpushed information is put back into the valid information set and ranked at the top of all video information, so that the video can be selected first the next time. If the user is active and displays browsing behavior at non-scheduled times, the system can flexibly adjust and temporarily push relevant information based on the current behavior pattern, increasing the immediacy and personalization of the user experience.
[0118] In this embodiment, only video information is described. In addition to video information, the information to be recommended may also include text information and audio information; in addition to original videos, the valid information set may also include original text and original audio; similarly, in addition to non-original videos, the invalid information set may also include non-original text and non-original audio.
[0119] The embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. The embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part, and / or installed from a removable medium. When the computer program is executed by the central processing unit (CPU), the above-mentioned functions defined in the method of the present application are executed. It should be noted that the above-mentioned computer-readable medium of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of an electrical, magnetic, optical, electromagnetic, infrared segment, or semiconductor, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more wire segments, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, electrical wire, optical cable, RF, etc., or any suitable combination of the foregoing.
[0120] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0121] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments. Without departing from the principles, the implementation methods of the present invention may be deformed or modified in any way.
Claims
1. A method for pushing tourism service information based on artificial intelligence, characterized in that: The method comprises: Collect user information and build user portraits based on the collected user information; Obtain the information to be recommended, extract key features from the information to be recommended, match it with the user profile, and form an original information database with the information to be recommended that matches the user profile; The original information database is screened, the original information database is divided into a valid information set and multiple invalid information sets, the information to be recommended in the valid information set is pushed to the user terminal, and feedback is received; specifically: the travel video information is obtained from the original information database; multiple video frames are extracted from the travel video through a sliding window; Input the video frame into the trained CNN model, compare the video frame with the image in the original video library through the CNN model, and identify whether the travel video is an original video, specifically: use TensorFlow or PyTorch to build a CNN model; the CNN model includes multiple convolutional layers, pooling layers and fully connected layers; Acquire multiple original image data from an original video library; acquire multiple corresponding non-original video image data from the Internet; The original image data and the non-original video image data form an image sample set, and the image sample set is divided into a training set and a validation set; Use the training set to train the CNN model, use the cross entropy loss function and optimizer to minimize the loss of the CNN model, and the output of the CNN model is the probability value of the image being the original video. ; Set comparison threshold , when | - |≥0, it is considered to be an original video; Input the acquired multiple video frames into the trained CNN model to determine whether it is an original video; The travel videos identified as original videos are classified as valid information sets; the travel videos identified as non-original videos are classified as invalid information sets; The information to be recommended includes one or more of video information, text information and audio information, the valid information set includes one or more of original video, original text and original audio, and the invalid information set includes one or more of non-original video, non-original text and non-original audio.
2. The method for pushing tourism service information based on artificial intelligence according to claim 1, characterized in that: The collecting of user information and building of user portraits based on the collected user information includes: Collect users' personal information, browsing history data, and purchase history data through API calls or crawler technology; Clean and pre-process the collected personal information, browsing history, and purchase history data to form an original user information database; Extract user key features from the original user information database to form a user key feature set; specifically: Extract the text data browsed by users from the browsing history data, and use the TF-IDF algorithm to extract keywords to obtain the key features of user browsing preferences; Using the Apriori algorithm, we analyze the frequently appearing keywords of tourist attractions from the user's browsing history data, thereby inferring the user's tourist attraction preferences and obtaining the key features of the user's tourist attraction preferences; The key features of user browsing preferences and the key features of user tourist attraction preferences are combined into a user key feature set; Use machine learning algorithms to perform cluster analysis on key user features, divide users into different user groups, and build user portraits for them; The personal information includes name, age, and occupation; the browsing history data is the user's browsing behavior on the travel website, including browsing time, travel location, and browsing duration; the purchase history is the user's purchase record on the travel website, including travel package information and reserved air ticket information.
3. The method for pushing tourism service information based on artificial intelligence according to claim 1, characterized in that: It also includes constructing a travel preference feature set, which specifically includes the following steps: Extract browsing time features; Extract the user's historical browsing timestamp from the browsing history; Use the pandas library to convert the browsing timestamp into a time format, extract the hour, week, and month features, and convert them into integers respectively, using the hour feature as the time feature ; Get multiple time features The average time characteristics of ;in: ,in is the number of views; Extracting tourist location features; Extracting text data of the user’s travel locations from browsing history data; The tourist locations are converted into numerical features through unique hot encoding, specifically: Convert text data into numerical features using the bag-of-words model or TF-IDF; Extract browsing time features: Extract browsing time from browsing history data ; Convert browsing time to minutes; Check out outliers and remove them; the outliers are browsing durations less than 0 or greater than 10,000 minutes; The browsing time characteristics, travel location characteristics and average time characteristics form the travel preference feature set A.
4. The method for pushing tourism service information based on artificial intelligence according to claim 2, characterized in that: The step of obtaining the information to be recommended and extracting key features from the information to be recommended includes: Use crawler technology to obtain travel video information shared by other users from the Internet; Extract key video features from travel videos, including: video theme, copyright statement, scenic spots, and shooting time; specifically: The pre-trained YOLOv3 model is used to loop through the video frames of the travel video, detect objects in the video, and pre-add category name mapping to make the YOLOv3 model output object names; Use OCR technology to extract text from the video and generate text format data from the detected text; Extract scenic spot names, copyright notices, and time keywords from text format data through keyword algorithms; The extracted keywords form a keyword database, and the keyword database is sent to the cloud for storage.
5. The method for pushing tourism service information based on artificial intelligence according to claim 4 is characterized in that: The extracting key features from the information to be recommended, matching them with the user profile, and forming an original information database with the information to be recommended that matches the user profile include: Use TF-IDF to vectorize keywords and user key features; Calculate the similarity between the user's key feature vector and the keyword vector through the pandas library, output the similarity score, and sort the similarity scores in ascending order; Sort the videos in descending order according to the similarity scores and select the top travel videos, submit similarity score labels for each travel video, and select The travel videos constitute the original information database.
6. The method for pushing tourism service information based on artificial intelligence according to claim 5, characterized in that: The step of pushing the information to be recommended in the valid information set to the user terminal and receiving feedback includes: Arrange the information to be recommended in the valid information set in descending order according to the similarity scores to form a recommended information sequence; Setting recommendation thresholds , Push time period ( , ), push time interval , ; Indicates the push start time point. Indicates the push end time point; choose Similarity score greater than The recommended information is Time is greater than When the user sends the recommended information, the user receives the feedback information, and sends the received feedback information to the cloud storage; ,in, , is the time parameter, ;if , , then the The video information is returned to the valid information set and inserted at the first place of the recommended information sequence to ensure that it will be given priority in the next recommendation.
7. The tourism service information push system based on artificial intelligence is characterized by: The system is used to execute the artificial intelligence-based tourism service information push method described in any one of claims 1-6.
8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the artificial intelligence-based tourism service information push method described in any one of claims 1 to 6.
