Tourism product recommendation engine and method based on user portrait

Abstract

The invention relates to the technical field of information science, in particular to a tourism product recommendation engine and method based on a user portrait, and the engine comprises a data acquisition module which is used for regularly acquiring an original user portrait of a to-be-updated user and corresponding historical tourism data; the portrait auditing module is used for acquiring tourism frequency data corresponding to the to-be-updated user according to the historical tourism data, judging whether the to-be-updated user belongs to a long-term tourism type or not according to the tourism frequency data, and if the to-be-updated user belongs to the long-term tourism type, updating an original user portrait based on the historical tourism data; if the to-be-updated user does not belong to the long-term tourism type, obtaining corresponding effective tourism data according to the historical tourism data, and updating an original user portrait according to the effective tourism data; and the tourism recommendation module generates corresponding tourism recommendation information based on the updated original user portrait. The travel strategy searching time of the to-be-updated user is shortened, and the travel interest of the to-be-updated user can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of information science, and in particular to a travel product recommendation engine and method based on user portraits. Background Art

[0002] In recent years, research in the field of recommender systems has rapidly developed. Various recommendation systems have been widely applied in a wide range of fields, including e-commerce, social networking sites, e-tourism, and internet advertising, demonstrating excellent results and promising prospects. With the rise of online travel websites (such as Expedite, Travel Zoo, and Tuniu), online data that can capture user travel preferences is becoming increasingly abundant, making travel product recommendations a hot topic in the field of recommender system research.

[0003] At present, in order to reduce the user's pre-travel preparation work, a profile is generated based on the recommendation system based on the user's personal habits, historical travel data and browsing information, and the corresponding tourist attractions are recommended to the user.

[0004] However, once the user's corresponding portrait is generated, the recommendation system will make travel recommendations based on the user's corresponding portrait. The recommended travel products are relatively simple, and users will experience aesthetic fatigue, resulting in low interest in travel. Summary of the Invention

[0005] In order to enhance users' interest in travel and to adjust users' corresponding travel profiles in a timely manner, this application provides a travel product recommendation engine and method based on user profiles.

[0006] In the first aspect, this application provides a travel product recommendation engine based on user portraits, which adopts the following technical solutions: A travel product recommendation engine based on user portraits, including: A data acquisition module, which is used to regularly acquire the original user portrait of the user to be updated and the corresponding historical travel data; A portrait review module, the portrait review module is used to obtain travel frequency data corresponding to the user to be updated based on the historical travel data, and determine whether the user to be updated is a long-term travel type based on the travel frequency data; if the user to be updated is a long-term travel type, the original user portrait is updated based on the historical travel data; if the user to be updated is not a long-term travel type, the corresponding valid travel data is obtained based on the historical travel data, and the original user portrait is updated based on the valid travel data; A travel recommendation module is connected to the portrait review module via a network to receive the updated original user portrait, and generates corresponding travel recommendation information based on the updated original user portrait.

[0007] By adopting the above technical solution, the portrait review module processes historical travel data and determines whether the user to be updated belongs to the long-term travel type through the travel frequency data of the user to be updated, thereby obtaining different ways to update the original user portrait, which is more in line with the choice of the user to be updated, and can adjust the user's corresponding travel portrait in time to accurately update the original user portrait, improve the accuracy of product recommendations for the user to be updated, and increase the travel interest of the user to be updated; then, the travel recommendation module generates travel recommendation information based on the updated original user portrait, reduces the travel strategy time of the user to be updated, and can increase the travel interest of the user to be updated.

[0008] In some embodiments, the portrait review module includes a frequency division unit and a frequency analysis unit, wherein the frequency division unit processes the historical travel data to obtain travel frequency data and generates a travel frequency curve based on the travel frequency data; The frequency analysis unit obtains the travel status corresponding to the user to be updated based on the travel frequency curve, and determines whether there is regularity in the travel status; if there is regularity in the travel status, it is determined that the user to be updated belongs to the long-term travel type; if there is no regularity in the travel status, it is determined that the user to be updated does not belong to the long-term travel type.

[0009] By adopting the above technical solution, the travel frequency status is analyzed based on the frequency analysis unit, and then it is obtained whether there is any regularity in the travel status of the user to be updated. If there is regularity, it means that the user to be updated is a planned long-term tourist, and the travel of the user to be updated is purposeful. Then, the original user portrait can be updated according to the recent travel characteristics of the historical travel data. If the travel status of the user to be updated is random and there is no regularity, it means that the user to be updated is random and pays more attention to the inherent characteristics of the travel product. Therefore, the original user portrait can be updated according to the effective travel data that the user to be updated likes.

[0010] In some embodiments, the updated original user portrait includes a product recommendation tag, and the portrait review module further includes a portrait evaluation unit and a portrait update unit. The portrait evaluation unit is used to determine whether the updated original user portrait has a travel evaluation tag. If so, the portrait update unit divides the updated original user portrait according to the frequency change of the travel frequency curve to obtain a data set to be processed, the data set to be processed includes a number of the travel frequency data and corresponding product type data, and updates the product recommendation tag based on the product type data and the personal information of the user to be updated; If the travel evaluation tag does not exist, the portrait updating unit generates a user portrait based on the historical travel data and the personal information of the user to be updated.

[0011] By adopting the above technical solution, the portrait evaluation unit searches the updated original user portrait to determine whether there is a travel evaluation tag. If so, it means that the product recommendation tag of the original user portrait is separate and can be updated separately, reducing the cost of regenerating the portrait and improving the update efficiency of the updated user portrait. If it does not exist, it is necessary to regenerate the user portrait based on historical travel data and personal information. The product recommendation tag of the user portrait can be updated separately to facilitate subsequent portrait updates.

[0012] In some embodiments, the travel recommendation module includes a data analysis unit and a product recommendation unit. The data analysis unit is used to process the data set to be processed in sequence to determine whether there is similarity in the product type data in the data set to be processed. If there is similarity, the product recommendation unit updates the product recommendation tag of the user to be updated based on the travel frequency data and the similar features of the product type data in the data set to be processed. If there is no similarity, the product recommendation unit updates the product recommendation tag of the user to be updated based on the personal information of the user to be updated and the product type data.

[0013] By adopting the above technical solution, the data analysis unit processes the data set to be processed in turn to determine whether the product type data in the data set to be processed are similar. If there is similarity, the product recommendation unit updates the product recommendation label of the user to be updated based on the travel frequency data and the similar features of the product type data in the data set to be processed. If there is no similarity, the product recommendation unit updates the product recommendation label of the user to be updated based on the personal information of the user to be updated and the product type data. The product recommendation module of the user to be updated can be adjusted separately and timely, thereby improving the efficiency of updating the original user portrait. By analyzing the product type data in the data set to be processed, the product recommendation label is updated based on the travel products selected by the user to be updated and the personal information, thereby improving the accuracy of travel product recommendations and thereby increasing the travel interest of the user to be updated.

[0014] In some embodiments, the travel recommendation module includes a recommendation evaluation unit, which is used to obtain corresponding prediction data based on the product recommendation tag, and obtain recommendation information based on the original user portrait, and generate corresponding travel recommendation information based on the prediction data and the recommendation information.

[0015] By adopting the above technical solution, travel recommendation information is generated based on recommendation information and prediction data, thereby being able to provide more accurate travel products to users to be updated, thereby increasing the travel interest of users to be updated.

[0016] In some embodiments, the prediction data includes prediction success data and prediction failure data, and the recommendation evaluation unit includes a comparison subunit and an output subunit; The comparison subunit is used to obtain the prediction success rate based on the prediction success data and the prediction failure data, and to determine whether the prediction success rate is greater than a preset value; if so, the output subunit obtains the predicted product information based on the prediction data, and uses the predicted product information as the travel recommendation information; if not, the output subunit generates the travel recommendation information based on the recommendation information.

[0017] By adopting the above technical solution, the comparison subunit determines whether the prediction success rate is greater than the preset value; if so, the output subunit obtains the predicted product information based on the prediction data, and uses the predicted product information as the travel recommendation information; if not, the output subunit uses the recommendation information as the travel recommendation information, and then can make real-time adjustments based on the predicted product information and the recommendation information to generate accurate travel recommendation information, which is convenient for use by users to be updated and increases users' interest in travel.

[0018] In some embodiments, the travel recommendation module further includes a travel prediction unit, a result comparison unit, and a result analysis unit. The travel prediction unit is configured to obtain personal information of a user to be updated, obtain several groups of estimated time periods based on the inference of the user's personal information, and generate predicted product recommendations based on the estimated time periods and the historical travel data. The result comparison unit is used to obtain the actual travel data of the user to be updated during the estimated time period, and compare the actual travel data with the predicted product recommendation to determine whether the actual travel data is similar to the predicted product recommendation; If they are similar, the result analysis unit generates a prediction success signal and marks the prediction data according to the prediction success signal; if they are not similar, the result analysis unit marks the prediction data according to prediction failure.

[0019] By adopting the above technical solution, during the use process of the user to be updated, the result comparison unit compares the actual travel data with the predicted product recommendation to determine whether the actual travel data is similar to the predicted product recommendation, and gradually improves the accuracy of the predicted data. It can then make real-time adjustments based on the predicted product information and recommendation information to generate accurate travel recommendation information, which is convenient for the use of the user to be updated and increases the user's interest in travel.

[0020] In a second aspect, the present application provides a method for recommending travel products based on user portraits, which adopts the following technical solutions: A travel product recommendation method based on user portraits includes the following steps: Regularly obtain the original user profile of the user to be updated and the corresponding historical travel data; Acquire travel frequency data corresponding to the user to be updated based on the historical travel data, and determine whether the user to be updated belongs to the long-term travel type based on the travel frequency data; If the user to be updated is a long-term traveler, then the original user profile is updated based on the historical travel data; If the user to be updated does not belong to the long-term travel type, then obtain corresponding valid travel data based on the historical travel data; The original user portrait is updated according to the valid travel data, and corresponding travel recommendation information is generated based on the updated original user portrait.

[0021] In some embodiments, the travel recommendation information includes a travel route, and after generating corresponding travel product information based on the updated user profile, the following steps are further included: Obtaining the starting location and destination location of the user to be updated, and determining whether the user to be updated accepts the recommendation based on the destination location; If so, predicting the travel status based on the travel route, and determining whether the destination needs to be changed based on the travel status; If it is determined that the travel status requires a change of destination, the travel recommendation information is changed according to the travel route.

[0022] By adopting the above technical solution, the destination location is analyzed in real time during the travel process of the user to be updated, and it is determined whether the destination needs to be changed based on the travel status prediction. Then, the travel recommendation information of the user to be updated can be adjusted in time to enhance the travel experience of the user to be updated, so that the user to be updated can achieve a relaxed state through travel.

[0023] In some embodiments, determining whether the destination needs to be changed based on the travel status includes the following steps: Obtaining a waiting time based on the travel status, and obtaining an estimated number of people based on the destination; generating a travel replacement value according to the waiting time and the estimated number of people, and determining whether the travel replacement value is greater than a preset replacement value; If the travel change value is greater than the preset change value, it is determined that the destination needs to be changed; If the travel change value is not greater than the preset change value, it is determined that there is no need to change the destination.

[0024] By adopting the above technical solution, when there are a large number of people at the destination and the road is congested, the travel recommendation information for the user to be updated can be adjusted in time to enhance the travel experience of the user to be updated, so that the user to be updated can achieve a relaxed state through travel.

[0025] In summary, this application includes at least one of the following beneficial technical effects: The portrait review module processes historical travel data and determines whether the user to be updated is a long-term traveler based on the user's travel frequency data. This allows the user to update the original user portrait in different ways to better suit the user's choices. The user's corresponding travel portrait can be adjusted in a timely manner to accurately update the original user portrait, improve the accuracy of product recommendations for the user to be updated, and increase the user's travel interest. The travel recommendation module then generates travel recommendation information based on the updated original user portrait, reducing the time it takes for the user to be updated to search for travel strategies and increasing the user's travel interest. The portrait evaluation unit searches the updated original user portrait to determine whether there is a travel evaluation tag. If so, it means that the product recommendation tag of the original user portrait is independent and can be updated separately, which reduces the cost of regenerating the portrait and improves the update efficiency of the updated user portrait. If not, it is necessary to regenerate the user portrait based on historical travel data and personal information. The product recommendation tag of the user portrait can be updated separately to facilitate subsequent portrait updates. During the travel of the user to be updated, the destination location is analyzed in real time, and it is determined whether the destination needs to be changed based on the travel status prediction. The travel recommendation information for the user to be updated can be adjusted in time to enhance the travel experience of the user to be updated, so that the user to be updated can achieve a relaxed state through travel. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a schematic diagram of the structure of a travel product recommendation engine based on user portraits provided in an embodiment of the present application; Figure 2 This is a schematic diagram of the specific structure of the travel recommendation module provided in the embodiment of the present application; Figure 3 This is a block diagram of a travel product recommendation method based on user portraits provided in an embodiment of the present application; Figure 4 This is another method block diagram provided in an embodiment of the present application.

[0027] Explanation of the accompanying symbols: 10. Data acquisition module; 20. Portrait review module; 21. Frequency division unit; 22. Frequency analysis unit; 23. Portrait evaluation unit; 24. Portrait update unit; 30. Travel recommendation module; 31. Data analysis unit; 32. Product recommendation unit; 33. Recommendation evaluation unit; 34. Travel prediction unit; 35. Result comparison unit; 36. Result analysis unit. DETAILED DESCRIPTION

[0028] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. However, it should be understood by those skilled in the art that the present application can be implemented without these details. In some cases, in order to avoid unnecessary descriptions that make various aspects of the present application obscure, the well-known methods, processes, systems, components and / or circuits that have been described at a higher level will not be described in detail. It is obvious to those skilled in the art that various changes can be made to the embodiments disclosed in the present application, and the general principles defined in the present application can be applied to other embodiments and application scenarios without departing from the principles and scope of the present application. Therefore, the present application is not limited to the embodiments shown, but conforms to the broadest scope consistent with the scope claimed for protection in the present application.

[0029] The embodiment of the present application discloses a travel product recommendation engine based on user portraits.

[0030] like Figure 1 As shown, the travel product recommendation engine based on user portraits includes a data acquisition module 10, a portrait review module 20 and a travel recommendation module 30. The data acquisition module 10 is used to regularly obtain the original user portrait of the user to be updated and the corresponding historical travel data. The portrait review module 20 is used to obtain the travel frequency data corresponding to the user to be updated based on the historical travel data, and to determine whether the user to be updated belongs to the long-term travel type based on the travel frequency data. If the user to be updated belongs to the long-term travel type, the original user portrait is updated based on the historical travel data. If the user to be updated does not belong to the long-term travel type, the corresponding valid travel data is obtained based on the historical travel data, and the original user portrait is updated based on the valid travel data. The travel recommendation module 30 is connected to the portrait review module 20 through a network to receive the updated original user portrait, and generates corresponding travel recommendation information based on the updated original user portrait.

[0031] Among them, the original user portrait represents the travel portrait that has been generated for the user to be updated. The original user portrait is generated based on the historical travel data and personal information data of the user to be updated. The historical travel data includes travel product data and travel data, and the personal information data includes the user's own financial status and interest data.

[0032] It should be noted that the original user profile is generated by collecting historical travel data and personal information, analyzing this data, for example, completing data cleaning and integration, then extracting features and labeling the data to form tag features. Based on these tag features, a virtual user model is constructed, and the corresponding original user profile is generated. The original user profile can be presented using visualization tools such as XMind or PPT.

[0033] Travel product data can be obtained from the product descriptions or from the data in various social media accounts, thereby obtaining the data corresponding to each travel product. Travel data and personal information data can be obtained from the social media accounts and personal information of the user to be updated, or from data directly entered by the user to be updated.

[0034] It should be noted here that Google Analytic, Sensors Data, and tracking SDKs (such as Sensors Analytic) can be used to obtain various data, while SQL (Hive / Spark SQL) and Python (Pandas) are used for data processing, and rule engines (Drools), NLP (BERT, Pacy), and image recognition (TensorFlow) are used to label data.

[0035] Historical travel data also includes the duration of each trip, and the frequency data of each trip is obtained through the travel time. The long-term travel type indicates that the user to be updated is a long-term and planned tourist. For the users to be updated who belong to the long-term travel type, there is a high tendency to choose the type of travel products. Therefore, for this type of users to be updated, it is only necessary to analyze the similarities of historical travel data and update the original user portrait through the product types of more recent historical travel data.

[0036] It should be noted here that how to analyze the similarities of historical tourism data is mainly to compare various characteristics of multiple historical tourism data, such as tourism product characteristics, tourism experience characteristics, and tourism journey characteristics, so as to regard the similar characteristics of historical tourism data as the similarities of historical tourism data.

[0037] For users who are not long-term travelers, they do not have higher requirements for travel products. Therefore, users of this type who are updated focus more on travel status. Therefore, valid travel data with good travel experiences can be screened out from their historical travel data, and the original user portrait can be updated based on the valid travel data.

[0038] It should be noted that valid travel data can be obtained through the social media platforms of the user to be updated, or through the user's travel experience with previous travel products. This primarily involves obtaining the user's travel reviews. Evaluation feature words are extracted from these reviews and then filtered against a pre-set database to obtain corresponding comparison feature words. A check is then made to determine whether a match exists in the pre-set database. If a match exists, the trip is considered valid and the travel data is then considered valid.

[0039] Updating the original user profile based on valid travel data involves obtaining the updated label features required for the user profile based on the valid travel data, replacing the label features of the original user profile with the updated label features, and then constructing a virtual user model based on the updated label features to generate the corresponding original user profile. The original user profile can be presented using visualization tools such as XMind or PPT.

[0040] Travel recommendations are generated based on the updated user profile of the user being updated. The updated profile includes travel products, and travel recommendations include route reminders, travel guides, and travel tips. Route reminders include departure time, travel duration, and itinerary. Real-time traffic analysis is used to promptly remind the user to depart to avoid congestion. Users are also promptly notified of potential issues that may arise during their travels, and are reminded to bring their own belongings.

[0041] Reference Figure 1 In one embodiment, the profile review module 20 includes a frequency division unit 21 and a frequency analysis unit 22. The frequency division unit 21 processes historical travel data to obtain travel frequency data and generates a travel frequency curve based on the travel frequency data. The frequency analysis unit 22 obtains the travel status corresponding to the user to be updated based on the travel frequency curve and determines whether there is any regularity in the travel status. If there is any regularity in the travel status, the user to be updated is determined to be a long-term traveler. If there is no regularity in the travel status, the user to be updated is determined not to be a long-term traveler.

[0042] The travel frequency data is the time interval between two consecutive trips of the user to be updated. The travel frequency curve is obtained by first displaying the travel frequency data as data points in chronological order and then smoothing the data points using a smooth curve to obtain the corresponding travel frequency curve. The frequency analysis unit 22 determines the travel status of the user to be updated based on the trend of the travel frequency curve and determines whether there is any regularity in the trend of the travel frequency curve. If there is any regularity, the user to be updated is determined to be a long-term traveler. If there is no regularity, the user to be updated is determined not to be a long-term traveler.

[0043] What needs to be explained here is how to judge whether there is a pattern based on the trend of the tourism frequency curve. First, we can observe it through visualization, that is, observe the tourism frequency curve, and then use statistical tests. Specifically, the ADF test can be used to test whether the curve maintains its statistical characteristics over time. Finally, through mathematical modeling, multiple models are used for curve fitting to determine whether there is a regularity in the tourism frequency curve.

[0044] Reference Figure 1In one embodiment, the updated original user profile includes a product recommendation tag. The profile review module 20 also includes a profile evaluation unit 23 and a profile update unit 24. The profile evaluation unit 23 is used to determine whether the updated original user profile has a travel evaluation tag. If so, the profile update unit 24 divides the updated original user profile according to the frequency change of the travel frequency curve to obtain a data set to be processed. The data set to be processed includes a number of travel frequency data and corresponding product type data, and updates the product recommendation tag based on the product type data and the personal information of the user to be updated. If the travel evaluation tag does not exist, the profile update unit 24 generates a user profile based on the historical travel data and the personal information of the user to be updated.

[0045] Among them, the travel evaluation tag is a sign that the product recommendation tag of the user portrait can be updated separately. The travel evaluation tag can specifically be a tag generated by the portrait generation module when establishing the user portrait, and the product recommendation tag is the travel product recommendation information generated for the user to be updated, specifically a certain travel product.

[0046] If there is no travel evaluation tag, the portrait updating unit 24 generates a user portrait based on the historical travel data and the personal information of the user to be updated. The user portrait generated here is a user portrait with a travel evaluation tag. The specific generated user portrait is generated according to the generation method of the original user portrait, which will not be described in detail here.

[0047] The data set to be processed includes a number of travel frequency data and corresponding product type data. The profile updating unit 24 divides the data set to be processed based on the frequency changes of the travel frequency curve. The data set to be processed is generated by performing a binomial fit on the travel frequency curve to obtain a corresponding binomial formula. The binomial formula is then differentiated and nodes with a derivative of 0 are divided to obtain multiple data sets to be processed. The binomial fitting is performed using software such as MATLAB, Excel, or Python.

[0048] It should be noted here that if the obtained binomial is approximately a straight line, the travel frequency of the user to be updated is very stable. Therefore, the profile update can be directly performed based on the historical product type data of the user to be updated.

[0049] Reference Figure 1In one embodiment, the travel recommendation module 30 includes a data analysis unit 31 and a product recommendation unit 32. The data analysis unit 31 is used to process the data set to be processed in sequence, and determine whether there is similarity in the product type data in the data set to be processed. If there is similarity, the product recommendation unit 32 updates the product recommendation label of the user to be updated based on the travel frequency data and the similar features of the product type data in the data set to be processed. If there is no similarity, the product recommendation unit 32 updates the product recommendation label of the user to be updated based on the personal information of the user to be updated and the product type data.

[0050] Among them, the product type data corresponding to several groups of data in the data set to be processed are compared to determine whether there are similarities in the product type data. If there are similarities, the product recommendation unit 32 updates the user portrait of the user to be updated based on the travel frequency data and the similar features of the product type data in the data set to be processed, so that when the user to be updated needs to travel, the travel product recommendation engine can provide corresponding travel recommendations in a timely manner, and the recommended travel products have the same features as previous historical data.

[0051] If there is no similarity, the product recommendation unit 32 updates the user profile of the user to be updated based on the personal information of the user to be updated and the product type data. Specifically, the user to be updated can obtain the most recent travel time and the current time, obtain the travel interval based on the most recent travel time and the current time, compare the travel interval with the preset interval, and determine whether the travel interval is greater than the preset interval. When the travel interval is greater than the preset interval, the user to be updated is judged to be in the life stage type according to the current time, and the travel profile corresponding to the life stage type is obtained, and the user profile of the user to be updated is replaced with the travel profile. When the travel interval is not greater than the preset interval, it is determined that the user to be updated is in a travel state of the corresponding frequency, so the proportion of the current travel type recommendation of the user to be updated can be increased, thereby obtaining the product type data at the preset time, and generating the corresponding user profile based on the product type data.

[0052] It should be noted here that the life stage type is specifically obtained based on the current age stage of the user to be updated. For example, the life stage types include student period, working period, family status, etc. Different life stages have different demands for the type of travel products. Therefore, corresponding travel products can be recommended based on the current life stage type of the user to be updated.

[0053] Reference Figure 1 In one embodiment, the travel recommendation module 30 includes a recommendation evaluation unit 33, which is used to obtain corresponding prediction data based on product recommendation tags, and obtain recommendation information based on the original user portrait, and generate corresponding travel recommendation information based on the prediction data and recommendation information.

[0054] Among them, the prediction data represents the travel product recommendation data obtained based on the updated original user portrait, and the recommendation information is the travel product recommendation information obtained based on the original user portrait.

[0055] It should be noted here that the predicted data and the recommended information are subjected to feature extraction respectively to obtain the corresponding recommended feature sets, and the two sets of recommended feature sets are compared, and the same features in the two sets of recommended feature sets are used as first-level response recommendations, and the different features in the two sets of recommended feature sets are used as second-level response recommendations. Preferably, a search is performed based on the first-level response recommendation to obtain tourism products with the same features as a temporary recommendation set, and then the tourism products that the user to be updated has visited are eliminated from the temporary recommendation set. Finally, the recommended products in the temporary recommendation set are compared based on the second-level response features to obtain feature similarity, and are arranged in reverse order based on the feature similarity, and the top three tourism products are recommended as tourism products. Corresponding travel guides are generated based on the tourism products, and the tourism product recommendations and the corresponding travel guides are used as tourism product recommendation information.

[0056] It should be noted here that the travel guide can be obtained by following the detailed travel guide corresponding to the tourism product to obtain other people's travel process, which can be obtained through public procedures, such as Weibo, Douyin, etc.

[0057] In one embodiment, the prediction data includes prediction success data and prediction failure data, and the recommendation evaluation unit 33 includes a comparison subunit and an output subunit. The comparison subunit is configured to obtain a prediction success rate based on the prediction success data and the prediction failure data, and to determine whether the prediction success rate is greater than a preset value. If so, the output subunit obtains predicted product information based on the prediction data and uses the predicted product information as travel recommendation information. If not, the output subunit generates travel recommendation information based on the recommendation information.

[0058] The prediction success data indicates whether the predicted recommended travel products meet the needs of the user to be updated, while the prediction failure data indicates whether the predicted recommended travel products do not meet the needs of the user to be updated. The prediction success rate indicates the ratio of the prediction success data to the total prediction data, where the total prediction data is the sum of the prediction success data and the prediction failure data. The preset value indicates the minimum standard for successful prediction for the user to be updated. When the prediction success rate exceeds the preset value, the output subunit obtains predicted product information based on the prediction data and uses the predicted product information as travel recommendation information. If the prediction success rate is not greater than the preset value, the output subunit uses the recommended information as travel recommendation information.

[0059] Reference Figure 2In one embodiment, the travel recommendation module 30 further includes a travel prediction unit 34, a result comparison unit 35, and a result analysis unit 36. The travel prediction unit 34 is used to obtain the personal information of the user to be updated, and based on the calculation of the user's personal information, to obtain several groups of estimated time periods, and generate predicted product recommendations based on the estimated time periods and historical travel data. The result comparison unit 35 is used to obtain the actual travel data of the user to be updated in the estimated time period, and compare the actual travel data with the predicted product recommendations to determine whether the actual travel data and the predicted product recommendations are similar. If similar, the result analysis unit 36 ​​generates a prediction success signal and marks the predicted data according to the prediction success signal. If not similar, the result analysis unit 36 ​​marks the predicted data according to the prediction failure.

[0060] The estimated time period representation is calculated based on the personal information of the user to be updated, so as to obtain the life stage of the user to be updated. For example, the estimated time period includes student period, working period, family status, etc.

[0061] It should be noted that the travel prediction unit 34 primarily receives the personal information acquired by the data acquisition module 10. The similarity between the actual travel data and the predicted product recommendations is determined by determining whether there is a 70% or greater similarity between the actual travel data and the predicted product recommendations. If so, the two are considered similar; otherwise, they are considered dissimilar.

[0062] It should be noted that, assuming the user to be updated accepts the predicted product recommended by the travel product recommendation engine and visits the corresponding product, the actual travel data is 100% similar to the predicted product recommendation. However, after the user to be updated visits the product, the user's experience is poor. In this case, the predicted data can be marked as a prediction failure. Specifically, the user's experience can be obtained by collecting data based on the user's comments, whether they recommend the product, and the information they post on various social platforms. The result analysis unit 36 ​​obtains the predicted data and sends it to the recommendation evaluation unit 33.

[0063] To determine whether the user accepts the recommended product, the system can determine the user's travel destination. If the user's destination matches the recommended product, the user is deemed to have chosen the recommended product. To determine the travel destination, the user can access the travel product recommendation engine and retrieve the destination location based on the user's current location. If the user does not click on the travel product recommendation engine, the user's mobile phone location can be used to determine the user's destination location.

[0064] For travel products that are not recommended by the travel product recommendation engine to the user to be updated, the travel prediction unit 34 can perform corresponding analysis on the personal information obtained in real time by the data acquisition module 10, obtain the travel area that the user to be updated is about to visit, search for the travel experience of previous tourists, and obtain corresponding travel recommendation information so that the user to be updated can travel better.

[0065] The embodiment of the present application also discloses a method for recommending travel products based on user portraits.

[0066] like Figure 3 As shown in FIG, the method for recommending tourism products based on user portraits includes the following steps: S100, regularly obtain the original user profile of the user to be updated and the corresponding historical travel data.

[0067] S200: Acquire travel frequency data corresponding to the user to be updated based on historical travel data, and determine whether the user to be updated belongs to the long-term travel type based on the travel frequency data.

[0068] S300: If the user to be updated is a long-term traveler, the original user profile is updated based on historical travel data.

[0069] S400: If the user to be updated does not belong to the long-term travel type, then corresponding valid travel data is obtained based on historical travel data.

[0070] S500, based on the updated original user portrait after valid travel data, corresponding travel recommendation information is generated based on the updated original user portrait.

[0071] The other functions performed in the above steps S100 to S500 and the technical details of each function are the same as or similar to the corresponding features in the user portrait-based travel product recommendation engine described above, so they will not be repeated here.

[0072] Reference Figure 4 In one embodiment, to meet the needs of the user to be updated, the travel recommendation information can be timely and adaptively adjusted based on the actual travel situation of the user to be updated during the travel process, thereby increasing the travel interest of the user to be updated. The travel recommendation information includes travel routes. After generating corresponding travel product information based on the updated user profile, the following steps are also included: S510: Obtain the starting location and destination location of the user to be updated, and determine whether the user to be updated accepts the recommendation based on the destination location.

[0073] S520: If yes, predict the travel status based on the travel route, and determine whether the destination needs to be changed based on the travel status.

[0074] S530: If it is determined that the travel status requires a change of destination, the travel recommendation information is changed according to the travel route.

[0075] Among them, the travel status prediction represents the state of the user to be updated after arriving at the destination from the current location, and determines whether there are a lot of tourists. If there are a lot of tourists, it will reduce the travel experience of the user to be updated. If the travel status prediction of the user to be updated is a low experience, it is determined that the destination needs to be changed. If the travel status prediction is a good experience, it is determined that the destination does not need to be changed. Of course, whether the destination needs to be changed depends on the choice of the user to be updated. The travel product recommendation engine will display the specific data to the user to be updated in a visual mode so that the user to be updated can make a choice. The alternative plan is specifically selected on the tourist route and has a tourist area similar to the tourist product. It can be quickly recommended to the user to be updated to meet the current needs of the user to be updated.

[0076] It should be noted here that the travel recommendation module is also used to execute the above steps S510-S530, adaptively adjust the travel recommendation information while the user to be updated is traveling, save the user to be updated from searching for relevant travel strategies, and improve the user to be updated's travel interest.

[0077] In one embodiment, determining whether a destination needs to be changed based on the travel status includes the following steps: S511, obtaining the waiting time based on the travel status, and obtaining the estimated number of people based on the destination.

[0078] S512, generating a travel replacement value based on the waiting time and the estimated number of people, and determining whether the travel replacement value is greater than a preset replacement value.

[0079] S513: If the travel change value is greater than the preset change value, it is determined that the destination needs to be changed.

[0080] S514: If the travel change value is not greater than the preset change value, it is determined that the destination does not need to be changed.

[0081] Among them, the travel replacement value is specifically a value obtained by adding the waiting time and the estimated number of people according to the corresponding weights. The time weight of the waiting time and the number weight of the estimated number of people can be obtained by comparing the waiting time with the average time, and the estimated number of people with the average number of people. When the waiting time is less than the average time and the estimated number of people is less than the average number of people, the sum of the time weight and the number weight is 0.5. The specific values ​​of the time weight and the number weight can be obtained according to the ratio of the waiting time to the average time and the ratio of the estimated number of people to the average number of people. When the ratio of the waiting time to the average time is greater than the ratio of the estimated number of people to the average number of people, the time weight is set to 0.3. When the ratio of the waiting time to the average time is equal to the ratio of the estimated number of people to the average number of people, the time weight and the number weight are both set to 0.25. When the ratio of the waiting time to the average time is less than the ratio of the estimated number of people to the average number of people, the number weight is both set to 0.3.

[0082] Specifically, when the waiting time is not less than the average time and / or the estimated number of people is not less than the average number of people, the sum of the time weight and the number weight is 1. The specific values ​​of the time weight and the number weight can be obtained based on the ratio of the waiting time to the average time and the ratio of the estimated number of people to the average number of people. When the ratio of the waiting time to the average time is greater than the ratio of the estimated number of people to the average number of people, the time weight is set to 0.8. When the ratio of the waiting time to the average time is equal to the ratio of the estimated number of people to the average number of people, the time weight and the number weight are both set to 0.5. When the ratio of the waiting time to the average time is less than the ratio of the estimated number of people to the average number of people, the number weight is set to 0.8.

[0083] It should be noted here that the travel recommendation module specifically determines whether the current travel destination of the user to be updated needs to be changed based on the travel status, and updates the travel recommendation information at the same time. If the destination does not need to be changed, the originally generated travel recommendation information continues to be displayed to the user to be updated.

[0084] The implementation principle is: The frequency division unit 21 processes the historical travel data to obtain travel frequency data and generates a travel frequency curve based on the travel frequency data. The frequency analysis unit 22 obtains the travel status corresponding to the user to be updated based on the travel frequency curve and determines whether there is regularity in the travel status. If there is regularity in the travel status, it is determined that the user to be updated belongs to the long-term travel type. If there is no regularity in the travel status, it is determined that the user to be updated does not belong to the long-term travel type. If the user to be updated belongs to the long-term travel type, the original user portrait is updated based on the historical travel data. If the user to be updated does not belong to the long-term travel type, the corresponding valid travel data is obtained based on the historical travel data, and the original user portrait is updated based on the valid travel data. The travel recommendation module 30 is connected to the portrait review module 20 via a network to receive the updated original user portrait and generate corresponding travel recommendation information based on the updated original user portrait.

[0085] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps and they may be performed in other orders.

[0086] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A travel product recommendation engine based on user portraits, characterized by: include: A data acquisition module (10), the data acquisition module (10) is used to regularly acquire the original user portrait of the user to be updated and the corresponding historical travel data; A portrait review module (20), the portrait review module (20) is used to obtain travel frequency data corresponding to the user to be updated based on the historical travel data, and to determine whether the user to be updated belongs to the long-term travel type based on the travel frequency data; if the user to be updated belongs to the long-term travel type, the original user portrait is updated based on the historical travel data; If the user to be updated does not belong to the long-term travel type, then corresponding valid travel data is obtained based on the historical travel data, and the original user profile is updated based on the valid travel data; A travel recommendation module (30) is connected to the portrait review module (20) via a network to receive the updated original user portrait and generate corresponding travel recommendation information based on the updated original user portrait.

2. The user portrait-based travel product recommendation engine according to claim 1, characterized in that: The portrait review module (20) includes a frequency division unit (21) and a frequency analysis unit (22), wherein the frequency division unit (21) processes the historical tourism data to obtain tourism frequency data, and generates a tourism frequency curve based on the tourism frequency data; The frequency analysis unit (22) obtains the travel status corresponding to the user to be updated according to the travel frequency curve, and determines whether there is regularity in the travel status; If the travel status has regularity, it is determined that the user to be updated belongs to the long-term travel type; If there is no regularity in the travel status, it is determined that the user to be updated does not belong to the long-term travel type.

3. The user portrait-based travel product recommendation engine according to claim 2, characterized in that: The updated original user portrait includes a product recommendation tag, and the portrait review module (20) also includes a portrait evaluation unit (23) and a portrait update unit (24). The portrait evaluation unit (23) is used to determine whether the updated original user portrait has a travel evaluation tag. If so, the portrait update unit (24) divides the updated original user portrait according to the frequency change of the travel frequency curve to obtain a data set to be processed, the data set to be processed includes a number of the travel frequency data and the corresponding product type data, and updates the product recommendation tag according to the product type data and the personal information of the user to be updated; If the travel evaluation tag does not exist, the portrait updating unit (24) generates a user portrait based on the historical travel data and the personal information of the user to be updated.

4. The user portrait-based travel product recommendation engine according to claim 3, characterized in that: The travel recommendation module (30) includes a data analysis unit (31) and a product recommendation unit (32). The data analysis unit (31) is used to process the data set to be processed in sequence and determine whether the product type data in the data set to be processed are similar. If similar, the product recommendation unit (32) updates the product recommendation tag of the user to be updated based on the travel frequency data and the similar features of the product type data in the data set to be processed. If similar, the product recommendation unit (32) updates the product recommendation tag of the user to be updated based on the personal information of the user to be updated and the product type data.

5. The user portrait-based travel product recommendation engine according to claim 3, characterized in that: The travel recommendation module (30) includes a recommendation evaluation unit (33), which is used to obtain corresponding prediction data based on the product recommendation tag and obtain recommendation information based on the original user portrait, and generate corresponding travel recommendation information based on the prediction data and the recommendation information.

6. The user portrait-based travel product recommendation engine according to claim 5, characterized in that: The prediction data includes prediction success data and prediction failure data, and the recommendation evaluation unit (33) includes a comparison subunit and an output subunit; The comparison subunit is used to obtain a prediction success rate based on the prediction success data and the prediction failure data, and to determine whether the prediction success rate is greater than a preset value; if so, the output subunit obtains predicted product information based on the prediction data, and uses the predicted product information as the travel recommendation information; If not, the output subunit generates the travel recommendation information according to the recommendation information.

7. The user portrait-based travel product recommendation engine according to claim 5, characterized in that: The travel recommendation module (30) further includes a travel prediction unit (34), a result comparison unit (35), and a result analysis unit (36), wherein the travel prediction unit (34) is used to obtain the personal information of the user to be updated, obtain several groups of estimated time periods based on the calculation of the personal information of the user, and generate predicted product recommendations based on the estimated time periods and the historical travel data; The result comparison unit (35) is used to obtain the actual travel data of the user to be updated during the estimated time period, and compare the actual travel data with the predicted product recommendation to determine whether the actual travel data is similar to the predicted product recommendation; If they are similar, the result analysis unit (36) generates a prediction success signal and marks the prediction data according to the prediction success signal; if they are not similar, the result analysis unit (36) marks the prediction data according to prediction failure.

8. A tourism product recommendation method based on user portrait, characterized in that: The following steps are involved: Regularly obtain the original user profile of the user to be updated and the corresponding historical travel data; Acquire travel frequency data corresponding to the user to be updated based on the historical travel data, and determine whether the user to be updated belongs to the long-term travel type based on the travel frequency data; If the user to be updated is a long-term traveler, then the original user profile is updated based on the historical travel data; If the user to be updated does not belong to the long-term travel type, then obtain corresponding valid travel data based on the historical travel data; The original user portrait is updated according to the valid travel data, and corresponding travel recommendation information is generated based on the updated original user portrait.

9. The method for recommending tourism products based on user portraits according to claim 8, characterized in that: The travel recommendation information includes travel routes. After generating corresponding travel product information based on the updated user profile, the following steps are also included: Obtaining the starting location and destination location of the user to be updated, and determining whether the user to be updated accepts the recommendation based on the destination location; If so, predicting the travel status based on the travel route, and determining whether the destination needs to be changed based on the travel status; If it is determined that the travel status requires a change of destination, the travel recommendation information is changed according to the travel route.

10. The method for recommending tourism products based on user portraits according to claim 9, characterized in that: The step of determining whether the destination needs to be changed based on the travel status includes the following steps: Obtaining a waiting time based on the travel status, and obtaining an estimated number of people based on the destination; generating a travel replacement value according to the waiting time and the estimated number of people, and determining whether the travel replacement value is greater than a preset replacement value; If the travel change value is greater than the preset change value, it is determined that the destination needs to be changed; If the travel change value is not greater than the preset change value, it is determined that there is no need to change the destination.

Images