Travel popularity analysis method and system based on artificial intelligence

Abstract

The invention provides a tourism popularity analysis method and system based on artificial intelligence, and relates to the field of data processing, and the method comprises the steps: obtaining historical tourism popularity information of a plurality of scenic spots; based on the historical tourism popularity information of the plurality of scenic spots, determining an association relationship among the plurality of scenic spots; obtaining user evaluation information of a plurality of scenic spots from the tourism platform; obtaining user search information of a plurality of scenic spots from the tourism platform; obtaining user interaction information of a plurality of scenic spots from a social media platform; for each scenic spot, predicting an initial popularity parameter of the scenic spot in a future tourism cycle based on artificial intelligence according to historical tourism popularity information, user evaluation information, user search information and user interaction information of the scenic spot; based on the association relationship among the plurality of scenic spots, the initial popularity parameter of each scenic spot in the future tourism cycle is iteratively optimized, the popularity parameters of the plurality of scenic spots in the future tourism cycle are generated, and the tourism popularity analysis method has the advantage of improving the accuracy of tourism popularity analysis.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a tourism heat analysis method and system based on artificial intelligence. Background Art

[0002] Existing tourism heat analysis mainly relies on historical tourism data for prediction, with a single data source, resulting in low accuracy.

[0003] Therefore, it is necessary to provide a tourism heat analysis method and system based on artificial intelligence to improve the accuracy of tourism heat analysis. Summary of the invention

[0004] The present invention provides a tourism heat analysis method based on artificial intelligence, comprising: obtaining historical tourism heat information of multiple scenic spots; determining the correlation relationship between the multiple scenic spots based on the historical tourism heat information of the multiple scenic spots; obtaining user evaluation information of the multiple scenic spots from a tourism platform; obtaining user search information of the multiple scenic spots from the tourism platform; obtaining user interaction information of the multiple scenic spots from a social media platform; for each of the scenic spots, predicting the initial heat parameters of the scenic spot in the future tourism cycle based on the historical tourism heat information, user evaluation information, user search information and user interaction information of the scenic spot based on artificial intelligence; based on the correlation relationship between the multiple scenic spots, iteratively optimizing the initial heat parameters of each of the scenic spots in the future tourism cycle, and generating the heat parameters of the multiple scenic spots in the future tourism cycle.

[0005] Furthermore, based on the historical tourism popularity information of the multiple scenic spots, the association relationship between the multiple scenic spots is determined, including: for each of the scenic spots, based on the historical tourism popularity information of the multiple scenic spots, generating historical tourism popularity change curves for multiple historical periods corresponding to the scenic spot; for the historical tourism popularity change curve of each of the historical periods, based on a preset time window, segmenting the historical tourism popularity change curve to generate multiple first historical tourism popularity change curve segments; based on the multiple first historical tourism popularity change curve segments, determining a popularity time period sequence corresponding to the historical period, wherein the popularity time period sequence is composed of multiple popularity time periods; based on the popularity time period sequence corresponding to each of the historical periods, generating a popularity time period matrix corresponding to the scenic spots; for any two of the scenic spots, based on the popularity time period matrices corresponding to the two scenic spots, calculating the popularity time similarity between the two scenic spots; clustering the multiple scenic spots based on the popularity time similarity between any two of the scenic spots through the K-means clustering algorithm, and determining the association relationship between the multiple scenic spots based on the clustering results.

[0006] Furthermore, based on the multiple first historical tourism heat change curve segments, a heat time period sequence corresponding to the historical period is determined, including: for each of the first historical tourism heat change curve segments, calculating the historical tourism heat mean and the historical tourism heat discrete parameter of the first historical tourism heat change curve segment, and judging whether the first historical tourism heat change curve segment is a target first historical tourism heat change curve segment based on the historical tourism heat mean and the historical tourism heat discrete parameter of the first historical tourism heat change curve segment; for any two adjacent target first historical tourism heat change curve segments, calculating the time difference between two adjacent target first historical tourism heat change curve segments; according to the time difference between any two adjacent target first historical tourism heat change curve segments, merging multiple target first historical tourism heat change curve segments to generate multiple second historical tourism heat change curve segments, determining multiple heat time periods, and generating a heat time period sequence corresponding to the historical period.

[0007] Furthermore, the multiple scenic spots are clustered based on the heat time similarity between any two of the scenic spots through the K-means clustering algorithm, and the correlation relationship between the multiple scenic spots is determined based on the clustering result, including: clustering the multiple scenic spots based on the heat time similarity between any two of the scenic spots through the K-means clustering algorithm to determine multiple scenic spot classes; for each scenic spot class, determining the correlation type of any two scenic spots included in the scenic spot class, wherein the correlation type is one of positive correlation, irrelevant and negative correlation, for the two scenic spots with positive correlation, calculating the positive correlation parameters of the two scenic spots, and for the two scenic spots with negative correlation, calculating the negative correlation parameters of the two scenic spots.

[0008] Further, determining the related types of the two scenic spots included in the scenic spot class includes: For each of the scenic spots, perform empirical mode decomposition on each second historical tourism heat change curve segment corresponding to the scenic spot, generate the historical tourism heat change intrinsic mode function and the historical tourism heat change residual of each second historical tourism heat change curve segment, extract the time domain characteristics and frequency domain characteristics of the historical tourism heat change intrinsic mode function, and extract the time domain characteristics and frequency domain characteristics of the historical tourism heat change residual; determine the relevant types of the two scenic spots through a type determination model based on the time domain characteristics and frequency domain characteristics of the historical tourism heat change intrinsic mode function of each second historical tourism heat change curve segment corresponding to the two scenic spots and the time domain characteristics and frequency domain characteristics of the historical tourism heat change residual; Calculating the positive correlation parameters of the two scenic spots, including: for each of the scenic spots, calculating the historical tourism heat mean and the historical tourism heat discrete parameter of the second historical tourism heat change curve segment corresponding to each of the historical periods of the scenic spot, generating a historical tourism heat mean sequence and a historical tourism heat discrete parameter sequence for the scenic spot corresponding to each of the historical periods; for each of the historical periods, calculating the periodic positive correlation parameters of the two scenic spots corresponding to the historical periods based on the historical tourism heat mean sequence and the historical tourism heat discrete parameter sequence of the two scenic spots corresponding to the historical periods; calculating the positive correlation parameters of the two scenic spots based on the periodic positive correlation parameters of the two scenic spots corresponding to each of the historical periods.

[0009] Furthermore, based on the historical tourism popularity information, user evaluation information, user search information and user interaction information of the scenic spot, the initial heat parameters of the scenic spot in the future tourism cycle are predicted, including: predicting the first heat parameter of the scenic spot in the future tourism cycle based on the historical tourism popularity information of the scenic spot through a heat prediction model; generating a first heat adjustment parameter based on the user evaluation information; generating a second heat adjustment parameter based on the user search information; generating a third heat adjustment parameter based on the user interaction information; adjusting the first heat parameter of the scenic spot in the future tourism cycle based on the first heat adjustment parameter, the second heat adjustment parameter and the third heat adjustment parameter to generate the initial heat parameter of the scenic spot in the future tourism cycle.

[0010] Furthermore, based on the user evaluation information, a first heat adjustment parameter is generated, including: for each user evaluation included in the user evaluation information, identifying the emotion type of the user evaluation; based on the emotion type of each user evaluation, generating the first heat adjustment parameter.

[0011] Furthermore, based on artificial intelligence, according to the historical tourism heat information, user evaluation information, user search information and user interaction information of the scenic spot, the initial heat parameters of the scenic spot in the future tourism cycle are predicted, including: S1, establishing a heat adjustment model; S2, using the heat adjustment model to generate the heat parameters of each scenic spot in the future tourism cycle corresponding to the current round based on the initial heat parameters of each scenic spot in the future tourism cycle; S3, based on the initial heat parameters of each scenic spot in the future tourism cycle and the heat parameters of each scenic spot in the current round in the future tourism cycle, the global heat fluctuation parameters are calculated; S4, according to the The global heat fluctuation parameters are used to determine whether to perform the next round of optimization. If so, execute S5. If not, the heat parameters of each scenic spot in the current round in the future tourism cycle are used as the heat parameters of each scenic spot in the future tourism cycle; S5. Based on the heat parameters of each scenic spot in the previous round in the future tourism cycle, the heat parameters of each scenic spot in the current round in the future tourism cycle are generated through the heat adjustment model; S6. Based on the heat parameters of each scenic spot in the previous round in the future tourism cycle and the heat parameters of each scenic spot in the current round in the future tourism cycle, the global heat fluctuation parameters are calculated and S4 is executed.

[0012] Furthermore, the global fluctuation parameter is calculated based on the following formula: ,in, is the global volatility parameter of the tth round, is the heat parameter of the kth scenic spot in the tth round in the future tourism cycle, is the heat parameter of the kth scenic spot in the t-1th round in the future tourism cycle, is the total number of scenic spots.

[0013] The present invention provides a tourism heat analysis system based on artificial intelligence, comprising: an information acquisition module, used to acquire historical tourism heat information of multiple scenic spots; an association analysis module, used to determine the association relationship between the multiple scenic spots based on the historical tourism heat information of the multiple scenic spots; the information acquisition module is also used to acquire user evaluation information of the multiple scenic spots from a tourism platform; the information acquisition module is also used to acquire user search information of the multiple scenic spots from the tourism platform; the information acquisition module is also used to acquire user interaction information of the multiple scenic spots from a social media platform; a heat analysis module is used to predict the initial heat parameters of each scenic spot in the future tourism cycle based on artificial intelligence according to the historical tourism heat information, user evaluation information, user search information and user interaction information of the scenic spot; the heat analysis module is also used to iteratively optimize the initial heat parameters of each scenic spot in the future tourism cycle based on the association relationship between the multiple scenic spots, and generate the heat parameters of the multiple scenic spots in the future tourism cycle.

[0014] Compared with the prior art, the tourism heat analysis method and system based on artificial intelligence provided by the present invention have at least the following beneficial effects: Based on the historical tourism popularity information of multiple scenic spots, the correlation between multiple scenic spots can be determined, which can clarify the mutual influence and correlation between two scenic spots, and provide a new perspective for predicting the popularity of scenic spots. User evaluation information reflects tourists' true feelings about scenic spot services, facilities, attractions, etc. User search information reveals tourists' intentions and needs, as well as their interest in a certain scenic spot. By monitoring and analyzing search data, the popularity of scenic spots in the future can be predicted. User interaction information, such as discussions and sharing on social media, reflects tourists' attention and participation in scenic spots. This information is of great significance for predicting the popularity of scenic spots. By determining the correlation between two scenic spots in terms of tourism popularity, and combining the analysis of the scenic spot's user evaluation information, user search information and user interaction information, combined with multi-dimensional information, and comprehensively considering the impact of multiple factors on the popularity of scenic spots, a comprehensive, accurate and timely prediction result is provided. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein: Figure 1 It is a module schematic diagram of a tourism heat analysis system based on artificial intelligence according to some embodiments of this specification; Figure 2 It is a flow chart of a tourism heat analysis method based on artificial intelligence according to some embodiments of this specification; Figure 3 It is a flowchart of iterative optimization according to some embodiments of this specification. DETAILED DESCRIPTION

[0016] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0017] Figure 1 is a module diagram of a tourism heat analysis system based on artificial intelligence according to some embodiments of this specification, such as Figure 1As shown, a tourism heat analysis system based on artificial intelligence may include a data acquisition module, a correlation analysis module and a heat analysis module.

[0018] The information acquisition module can be used to obtain historical tourism popularity information of multiple scenic spots.

[0019] The association analysis module can be used to determine the association relationship between multiple scenic spots based on the historical tourism popularity information of multiple scenic spots.

[0020] The information acquisition module can also be used to obtain user evaluation information of multiple scenic spots from the tourism platform.

[0021] The information acquisition module can also be used to obtain user search information of multiple scenic spots from the travel platform.

[0022] The information acquisition module can also be used to obtain user interaction information of multiple scenic spots from social media platforms.

[0023] The heat analysis module can be used for each scenic spot. Based on artificial intelligence, it can predict the initial heat parameters of the scenic spot in the future tourism cycle according to the historical tourism heat information, user evaluation information, user search information and user interaction information of the scenic spot.

[0024] The heat analysis module can also be used to iteratively optimize the initial heat parameters of each scenic spot in the future tourism cycle based on the correlation between multiple scenic spots, and generate heat parameters of multiple scenic spots in the future tourism cycle.

[0025] An artificial intelligence-based tourism heat analysis system can be used to execute an artificial intelligence-based tourism heat analysis method. For more descriptions of an artificial intelligence-based tourism heat analysis system, please refer to the relevant descriptions of an artificial intelligence-based tourism heat analysis method, which will not be repeated here.

[0026] Figure 2 is a flow chart of a tourism heat analysis method based on artificial intelligence according to some embodiments of this specification, such as Figure 2 As shown, a tourism heat analysis method based on artificial intelligence may include the following steps.

[0027] Step 210, obtaining historical tourism popularity information of multiple scenic spots.

[0028] Specifically, the number of tourists in multiple scenic spots at different time points in recent years can be obtained. For example, the number of tourists in a scenic spot every day in the past ten years can be obtained. The historical tourism popularity information of multiple scenic spots can be crawled from the network platform, or the historical tourism popularity information of the scenic spot can be obtained through relevant equipment set up in the scenic spot (for example, gates, cameras, etc.).

[0029] Step 220, based on the historical tourism popularity information of the multiple scenic spots, determine the association relationship between the multiple scenic spots.

[0030] Specifically include: For each scenic spot, based on the historical tourism heat information of multiple scenic spots, generate historical tourism heat change curves corresponding to multiple historical periods (for example, one year) of the scenic spot; for the historical tourism heat change curve of each historical period, segment the historical tourism heat change curve based on a preset time window (for example, 3 days) to generate multiple first historical tourism heat change curve segments; based on the multiple first historical tourism heat change curve segments, determine the heat time period sequence corresponding to the historical period, wherein the heat time period sequence is composed of multiple heat time periods; based on the heat time period sequence corresponding to each historical period, generate a heat time period matrix corresponding to the scenic spot, wherein one row of the heat time period matrix is ​​the heat time period sequence corresponding to one historical period; For any two scenic spots, the heat time similarity between the two scenic spots is calculated based on the heat time period matrix corresponding to the two scenic spots; The K-means clustering algorithm is used to cluster multiple scenic spots based on the heat time similarity between any two scenic spots, and the correlation between multiple scenic spots is determined based on the clustering results.

[0031] Specifically, the tourist number information of the scenic spot at multiple time points in a year can be fitted to generate a historical tourism heat change curve, which can be used to characterize the change of the tourist number of the scenic spot over time within a year.

[0032] In some embodiments, based on a plurality of first historical tourism heat change curve segments, determining a heat time period sequence corresponding to a historical period includes: For each first historical tourism heat change curve segment, calculate the historical tourism heat mean value and the historical tourism heat discrete parameter of the first historical tourism heat change curve segment, and determine whether the first historical tourism heat change curve segment is a target first historical tourism heat change curve segment based on the historical tourism heat mean value and the historical tourism heat discrete parameter of the first historical tourism heat change curve segment; For any two adjacent target first historical tourism heat change curve segments, calculate the time difference between the two adjacent target first historical tourism heat change curve segments; According to the time difference between any two adjacent target first historical tourism heat change curve segments, multiple target first historical tourism heat change curve segments are merged to generate multiple second historical tourism heat change curve segments, determine multiple heat time periods, and generate a heat time period sequence corresponding to the historical period, wherein an element of the heat time period sequence is a heat time period corresponding to a second historical tourism heat change curve segment, and the length of the heat time period sequence is fixed. When an element of a heat time period sequence is missing, it can be supplemented by the time period (0, 0).

[0033] For example, the historical tourism heat mean and the historical tourism heat discrete parameter of the first historical tourism heat change curve segment can be calculated based on the following formula:

[0034] in, is the mean value of the historical tourism heat of the i-th historical tourism heat change curve segment of the k-th scenic spot, is the historical tourism heat corresponding to the t-th sampling time point of the i-th first historical tourism heat change curve segment of the k-th scenic spot, is the total number of sampling time points corresponding to the i-th first historical tourism heat change curve segment of the k-th scenic spot, is the discrete parameter of the historical tourism heat of the i-th historical tourism heat change curve segment of the k-th scenic spot.

[0035] When the historical tourism heat mean of the first historical tourism heat change curve segment is greater than the preset mean threshold, and the historical tourism heat discrete parameter is less than the preset discrete parameter threshold, the first historical tourism heat change curve segment is determined to be the target first historical tourism heat change curve segment.

[0036] The time difference between the end time point of the previous target first historical tourism heat change curve segment and the start time point of the next target first historical tourism heat change curve segment in two adjacent target first historical tourism heat change curve segments can be calculated as the time difference between the two adjacent target first historical tourism heat change curve segments. When the time difference between the two adjacent target first historical tourism heat change curve segments is less than the preset time difference threshold, the two adjacent target first historical tourism heat change curve segments can be merged. The second historical tourism heat change curve segment can be formed by merging multiple consecutive target first historical tourism heat change curve segments, and the heat time period is the time period corresponding to the second historical tourism heat change curve segment.

[0037] The heat time similarity between two scenic spots can be calculated based on the following formula:

[0038] in, is the heat time similarity between the e-th scenic spot and the f-th scenic spot, is the preset parameter, greater than 0, is the time difference between the starting time point of the hot time period in the i-th row and j-th column in the hot time period matrix corresponding to the e-th scenic spot and the starting time point of the hot time period in the i-th row and j-th column in the hot time period matrix corresponding to the f-th scenic spot, is the time difference between the end time point of the hot time period in the i-th row and j-th column in the hot time period matrix corresponding to the e-th scenic spot and the end time point of the hot time period in the i-th row and j-th column in the hot time period matrix corresponding to the f-th scenic spot, The total number of rows in the heat time period matrix, The total number of columns in the heat time period matrix.

[0039] In some embodiments, a plurality of scenic spots are clustered based on the heat time similarity between any two scenic spots by using a K-means clustering algorithm, and a correlation relationship between the plurality of scenic spots is determined based on the clustering result, including: Based on the heat time similarity between any two scenic spots, the K-means clustering algorithm is used to cluster multiple scenic spots and determine multiple scenic spot classes; For each scenic spot class, determine the correlation type of any two scenic spots included in the scenic spot class, where the correlation type is one of positive correlation, irrelevant and negative correlation. For two scenic spots with positive correlation, calculate the positive correlation parameters of the two scenic spots, and for two scenic spots with negative correlation, calculate the negative correlation parameters of the two scenic spots.

[0040] In some embodiments, determining the related types of two scenic spots included in the scenic spot class includes: For each scenic spot, perform empirical mode decomposition on each second historical tourism heat change curve segment corresponding to the scenic spot, generate the historical tourism heat change intrinsic mode function and historical tourism heat change residual of each second historical tourism heat change curve segment, extract the time domain characteristics (for example, local extreme points, mean, mean square value, standard deviation, curve curvature, etc.) and frequency domain characteristics (for example, spectrum energy distribution, etc.) of the historical tourism heat change intrinsic mode function, and extract the time domain characteristics and frequency domain characteristics of the historical tourism heat change residual; Through the type determination model, the relevant types of the two scenic spots are determined based on the time domain characteristics and frequency domain characteristics of the historical tourism heat change intrinsic mode function of each second historical tourism heat change curve segment corresponding to the two scenic spots and the time domain characteristics and frequency domain characteristics of the historical tourism heat change residual, wherein the type determination model can be a Convolutional Neural Network (CNN) model.

[0041] In some embodiments, calculating the positive correlation parameters of two scenic spots includes: For each scenic spot, the historical tourism heat mean and historical tourism heat discrete parameters of the second historical tourism heat change curve segment corresponding to each historical period of the scenic spot are calculated to generate the historical tourism heat mean value sequence and historical tourism heat discrete parameter sequence of the scenic spot corresponding to each historical period; For each historical period, the periodic positive correlation parameters of the historical periods corresponding to the two scenic spots are calculated based on the historical tourism heat mean sequence and the historical tourism heat discrete parameter sequence of the two scenic spots in the corresponding historical periods; Based on the periodic positive correlation parameters of the two scenic spots corresponding to each historical period, the positive correlation parameters of the two scenic spots are calculated.

[0042] For example, the periodic positive correlation parameter can be calculated based on the following formula:

[0043] in, is the periodic positive correlation parameter between the e-th scenic spot and the f-th scenic spot, is the preset parameter, greater than 0, is the mean value of the historical tourism heat of the e-th scenic spot in the j-th second historical tourism heat change curve segment of the historical period, is the mean value of the historical tourism heat of the f-th scenic spot in the j-th second historical tourism heat change curve segment of the historical period, is the discrete parameter of the historical tourism heat of the e-th scenic spot in the j-th second historical tourism heat change curve segment of the historical period, is the discrete parameter of the historical tourism heat of the f-th scenic spot in the j-th second historical tourism heat change curve segment of the historical period.

[0044] The positive correlation parameters of the two scenic spots corresponding to each historical period can be averaged to obtain the positive correlation parameters of the two scenic spots.

[0045] In some embodiments, calculating the negative correlation parameters of two scenic spots includes: For each scenic spot, the historical tourism heat mean and historical tourism heat discrete parameters of the second historical tourism heat change curve segment corresponding to each historical period of the scenic spot are calculated to generate the historical tourism heat mean value sequence and historical tourism heat discrete parameter sequence of the scenic spot corresponding to each historical period; For each historical period, the negative correlation parameters of the historical periods corresponding to the two scenic spots are calculated based on the mean value sequence of the historical tourism heat and the discrete parameter sequence of the historical tourism heat of the two scenic spots. Based on the periodic negative correlation parameters of the two scenic spots corresponding to each historical period, the negative correlation parameters of the two scenic spots are calculated.

[0046] For example, the periodic negative correlation parameter can be calculated based on the following formula:

[0047] in, is the periodic negative correlation parameter between the e-th scenic spot and the f-th scenic spot, is the preset parameter, greater than 0, is the mean value of the historical tourism heat of the x-th scenic spot in the j-th second historical tourism heat change curve segment of the historical period, is the mean value of the historical tourism heat of the yth scenic spot in the jth second historical tourism heat change curve segment of the historical period, is the discrete parameter of the historical tourism heat of the x-th scenic spot in the j-th second historical tourism heat change curve segment of the historical period, It is the discrete parameter of the historical tourism heat of the jth second historical tourism heat change curve segment of the yth scenic spot in the historical period.

[0048] The negative correlation parameters of the two scenic spots corresponding to each historical period can be averaged to obtain the negative correlation parameters of the two scenic spots.

[0049] Step 230, obtaining user evaluation information of multiple scenic spots from the travel platform.

[0050] Specifically, the tourism platform may be a software platform that provides tourism-related services to users, and user evaluation information of multiple scenic spots may be obtained from the tourism platform through crawler and other technologies.

[0051] Step 240, obtaining user search information of multiple scenic spots from the travel platform.

[0052] Specifically, user search information of users using the travel platform may be obtained from the server of the travel platform.

[0053] Step 250: Obtain user interaction information of multiple scenic spots from the social media platform.

[0054] Specifically, a social media platform may be a software platform that provides social services to users, and user interaction information of multiple scenic spots may be obtained from the social media platform through crawler and other technologies, wherein user interaction may include but is not limited to posting dynamic information related to the scenic spot, commenting on dynamic information related to the scenic spot, forwarding dynamic information related to the scenic spot, etc.

[0055] Step 260, for each scenic spot, based on artificial intelligence, according to the historical tourism popularity information, user evaluation information, user search information and user interaction information of the scenic spot, predict the initial popularity parameters of the scenic spot in the future tourism cycle.

[0056] Specifically include: Predicting the first heat parameter of the scenic spot in the future tourism cycle based on the historical tourism heat information of the scenic spot through the heat prediction model, wherein the heat prediction model may be a long short-term memory network (LSTM, Long Short-Term Memory) model; Based on the user evaluation information, generating a first heat adjustment parameter; Based on the user search information, generating a second heat adjustment parameter; Based on the user interaction information, generating a third heat adjustment parameter; Based on the first heat adjustment parameter, the second heat adjustment parameter and the third heat adjustment parameter, the first heat parameter of the scenic spot in the future tourism cycle is adjusted to generate the initial heat parameter of the scenic spot in the future tourism cycle.

[0057] In some embodiments, generating a first heat adjustment parameter based on user evaluation information includes: For each user evaluation included in the user evaluation information, identifying the sentiment type of the user evaluation; Based on the sentiment type of each user evaluation, a first heat adjustment parameter is generated.

[0058] Specifically, the emotion type of the user's evaluation may be identified through an emotion recognition model, wherein the emotion recognition model may be a convolutional neural network (CNN) model, and the emotion type may be one of positive, neutral, negative, and the like.

[0059] Based on the sentiment type of each user evaluation, the proportion of user evaluations corresponding to each sentiment type is determined, and the first heat adjustment parameter is calculated based on the proportion of user evaluations corresponding to each sentiment type according to the following formula:

[0060] in, is the first heat adjustment parameter of the kth scenic spot, is the proportion of positive user reviews for the kth scenic spot, is the proportion of neutral user reviews for the kth scenic spot, is the proportion of negative user reviews for the kth scenic spot, and is the preset parameter, and greater than 0, Adjust parameters for the preset first heat level.

[0061] The second heat adjustment parameter can be calculated based on the following formula:

[0062] in, is the second heat adjustment parameter of the kth scenic spot, is the search ratio of the kth scenic spot, is the preset parameter, greater than 0, To preset the second heat adjustment parameters, is the number of searches for the kth scenic spot, The number of searches for all scenic spots.

[0063] Based on the user interaction information of the scenic spot obtained from the social media platform, the scores of the scenic spot in multiple interaction indicators (for example, playback indicators, comment indicators, forwarding indicators, etc.) can be determined, and the third heat adjustment parameter can be calculated based on the scores of the scenic spot in multiple interaction indicators.

[0064] The third heat adjustment parameter can be calculated based on the following formula:

[0065] in, is the third heat adjustment parameter of the kth scenic spot, is the preset parameter, greater than 0, is the score of the k-th scenic spot in the m-th interaction index, Adjust parameters for the third heat setting.

[0066] Step 270, based on the association relationship between the multiple scenic spots, iteratively optimize the initial heat parameters of each scenic spot in the future travel cycle, and generate the heat parameters of the multiple scenic spots in the future travel cycle.

[0067] Figure 3 is a schematic diagram of an iterative optimization process according to some embodiments of this specification, such as Figure 3 As shown in the figure, based on artificial intelligence, the initial heat parameters of the scenic spot in the future tourism cycle are predicted according to the historical tourism heat information, user evaluation information, user search information and user interaction information of the scenic spot, including: S1. Establishing a heat adjustment model, wherein the heat adjustment model may be a convolutional neural network (CNN) model; S2, generating heat parameters of each scenic spot in the future tourism cycle corresponding to the current round based on the initial heat parameters of each scenic spot in the future tourism cycle through the heat adjustment model; S3, calculating a global heat fluctuation parameter based on the initial heat parameter of each scenic spot in the future tourism cycle and the heat parameter of each scenic spot in the current round of the future tourism cycle; S4, according to the global heat fluctuation parameter, determine whether to perform the next round of optimization, if yes, execute S5, if no, take the heat parameter of each scenic spot in the current round in the future tourism cycle as the heat parameter of each scenic spot in the future tourism cycle, specifically, when the global heat fluctuation parameter is greater than the preset global heat fluctuation parameter threshold, determine to perform the next round of optimization; S5, generating heat parameters of each scenic spot in the future tourism cycle corresponding to the current round based on the heat parameters of each scenic spot in the previous round in the future tourism cycle through the heat adjustment model; S6. Based on the heat parameters of each scenic spot in the previous round in the future tourism cycle and the heat parameters of each scenic spot in the current round in the future tourism cycle, calculate the global heat fluctuation parameter and execute S4.

[0068] In some embodiments, the global fluctuation parameter is calculated based on the following formula:

[0069] in, is the global volatility parameter of the tth round, is the heat parameter of the k-th scenic spot in the t-th round in the future tourism cycle, is the heat parameter of the kth scenic spot in the t-1th round in the future tourism cycle, is the total number of scenic spots.

[0070] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A tourism heat analysis method based on artificial intelligence, characterized in that: include: Obtain historical tourism popularity information for multiple scenic spots; Determining the association relationship between the plurality of scenic spots based on the historical tourism popularity information of the plurality of scenic spots; Obtaining user evaluation information of the plurality of scenic spots from a travel platform; Acquire user search information of the plurality of scenic spots from the travel platform; Obtaining user interaction information of the plurality of scenic spots from a social media platform; For each of the scenic spots, predict the initial popularity parameters of the scenic spot in the future tourism cycle based on the historical tourism popularity information, user evaluation information, user search information and user interaction information of the scenic spot based on artificial intelligence; Based on the association relationship between the multiple scenic spots, the initial heat parameters of each scenic spot in the future travel cycle are iteratively optimized to generate the heat parameters of the multiple scenic spots in the future travel cycle.

2. According to the artificial intelligence-based tourism heat analysis method of claim 1, it is characterized in that: Determining the association relationship between the plurality of scenic spots based on the historical tourism popularity information of the plurality of scenic spots includes: For each of the scenic spots, based on the historical tourism heat information of the multiple scenic spots, generate the historical tourism heat change curves of the multiple historical periods corresponding to the scenic spot, for each of the historical tourism heat change curves of the historical period, based on the preset time window, segment the historical tourism heat change curve to generate multiple first historical tourism heat change curve segments, based on the multiple first historical tourism heat change curve segments, determine the heat time period sequence corresponding to the historical period, wherein the heat time period sequence is composed of multiple heat time periods, based on the heat time period sequence corresponding to each of the historical periods, generate the heat time period matrix corresponding to the scenic spot; For any two of the scenic spots, based on the heat time period matrices corresponding to the two scenic spots, calculate the heat time similarity between the two scenic spots; The plurality of scenic spots are clustered based on the heat time similarity between any two of the scenic spots by using a K-means clustering algorithm, and the association relationship between the plurality of scenic spots is determined based on the clustering result.

3. The tourism heat analysis method based on artificial intelligence according to claim 2 is characterized in that: Determining a heat time period sequence corresponding to the historical period based on the plurality of first historical tourism heat change curve segments includes: For each of the first historical tourism heat change curve segments, calculating the historical tourism heat mean value and the historical tourism heat discrete parameter of the first historical tourism heat change curve segment, and judging whether the first historical tourism heat change curve segment is a target first historical tourism heat change curve segment based on the historical tourism heat mean value and the historical tourism heat discrete parameter of the first historical tourism heat change curve segment; For any two adjacent target first historical tourism heat change curve segments, calculate the time difference between the two adjacent target first historical tourism heat change curve segments; According to the time difference between any two adjacent target first historical tourism heat change curve segments, multiple target first historical tourism heat change curve segments are merged to generate multiple second historical tourism heat change curve segments, determine multiple heat time periods, and generate a heat time period sequence corresponding to the historical period.

4. The tourism heat analysis method based on artificial intelligence according to claim 3 is characterized in that: Clustering the plurality of scenic spots based on the heat time similarity between any two of the scenic spots by using a K-means clustering algorithm, and determining the association relationship between the plurality of scenic spots based on the clustering result, including: Clustering the plurality of scenic spots based on the heat time similarity between any two of the scenic spots by using a K-means clustering algorithm to determine a plurality of scenic spot classes; For each scenic spot class, determine the correlation type of any two scenic spots included in the scenic spot class, wherein the correlation type is one of positive correlation, irrelevant and negative correlation. For the two scenic spots with positive correlation, calculate the positive correlation parameters of the two scenic spots, and for the two scenic spots with negative correlation, calculate the negative correlation parameters of the two scenic spots.

5. According to the artificial intelligence-based tourism heat analysis method of claim 4, it is characterized in that: Determining the related types of the two scenic spots included in the scenic spot class includes: For each of the scenic spots, perform empirical mode decomposition on each second historical tourism heat change curve segment corresponding to the scenic spot, generate the historical tourism heat change intrinsic mode function and the historical tourism heat change residual of each second historical tourism heat change curve segment, extract the time domain features and frequency domain features of the historical tourism heat change intrinsic mode function, and extract the time domain features and frequency domain features of the historical tourism heat change residual; Determine the relevant types of the two scenic spots through a type determination model based on the time domain characteristics and frequency domain characteristics of the historical tourism heat change intrinsic mode function of each second historical tourism heat change curve segment corresponding to the two scenic spots and the time domain characteristics and frequency domain characteristics of the historical tourism heat change residual; Calculating the positive correlation parameters of the two scenic spots, including: For each of the scenic spots, the historical tourism heat mean and the historical tourism heat discrete parameter of the second historical tourism heat change curve segment corresponding to each of the historical periods of the scenic spot are calculated to generate a historical tourism heat mean value sequence and a historical tourism heat discrete parameter sequence corresponding to each of the historical periods of the scenic spot; For each of the historical periods, based on the historical tourism heat mean value sequence and the historical tourism heat discrete parameter sequence of the two scenic spots corresponding to the historical period, the period positive correlation parameters of the historical periods corresponding to the two scenic spots are calculated; Based on the periodic positive correlation parameters of the two scenic spots corresponding to each of the historical periods, the positive correlation parameters of the two scenic spots are calculated.

6. The tourism heat analysis method based on artificial intelligence according to claim 1 is characterized in that: Based on the historical tourism popularity information, user evaluation information, user search information and user interaction information of the scenic spot, the initial popularity parameters of the scenic spot in the future tourism cycle are predicted, including: Predicting a first heat parameter of the scenic spot in a future tourism cycle based on the historical tourism heat information of the scenic spot through a heat prediction model; Based on the user evaluation information, generating a first heat adjustment parameter; Based on the user search information, generating a second heat adjustment parameter; Based on the user interaction information, generating a third heat adjustment parameter; Based on the first heat adjustment parameter, the second heat adjustment parameter and the third heat adjustment parameter, the first heat parameter of the scenic spot in the future travel period is adjusted to generate the initial heat parameter of the scenic spot in the future travel period.

7. The tourism heat analysis method based on artificial intelligence according to claim 6 is characterized in that: Based on the user evaluation information, generating a first heat adjustment parameter includes: For each user evaluation included in the user evaluation information, identifying the sentiment type of the user evaluation; The first heat adjustment parameter is generated based on the emotion type of each of the user evaluations.

8. A tourism heat analysis method based on artificial intelligence according to any one of claims 1 to 7, characterized in that: Based on artificial intelligence, the initial heat parameters of the scenic spot in the future tourism cycle are predicted according to the historical tourism heat information, user evaluation information, user search information and user interaction information of the scenic spot, including: S1. Establish a heat adjustment model; S2, generating heat parameters of each scenic spot in the future tourism cycle corresponding to the current round based on the initial heat parameters of each scenic spot in the future tourism cycle through a heat adjustment model; S3, calculating a global heat fluctuation parameter based on the initial heat parameter of each scenic spot in the future tourism cycle and the heat parameter of each scenic spot in the current round of the future tourism cycle; S4, judging whether to perform the next round of optimization according to the global heat fluctuation parameter, if yes, executing S5, if no, taking the heat parameter of each scenic spot in the current round in the future tourism cycle as the heat parameter of each scenic spot in the future tourism cycle; S5, generating a heat parameter of each scenic spot in the future tourism cycle corresponding to the current round based on the heat parameter of each scenic spot in the previous round in the future tourism cycle through a heat adjustment model; S6. Based on the heat parameters of each of the scenic spots in the previous round in the future travel cycle and the heat parameters of each of the scenic spots in the current round in the future travel cycle, calculate the global heat fluctuation parameter and execute S4.

9. The tourism heat analysis method based on artificial intelligence according to claim 8 is characterized in that: The global volatility parameter is calculated based on the following formula: in, is the global volatility parameter of the tth round, is the heat parameter of the kth scenic spot in the tth round in the future tourism cycle, is the heat parameter of the kth scenic spot in the t-1th round in the future tourism cycle, is the total number of scenic spots.

10. A tourism heat analysis system based on artificial intelligence, characterized in that: include: The information acquisition module is used to obtain the historical tourism popularity information of multiple scenic spots; A correlation analysis module, used to determine the correlation relationship between the plurality of scenic spots based on the historical tourism popularity information of the plurality of scenic spots; The information acquisition module is also used to obtain user evaluation information of the plurality of scenic spots from the travel platform; The information acquisition module is also used to acquire user search information of the plurality of scenic spots from the travel platform; The information acquisition module is also used to obtain user interaction information of the plurality of scenic spots from the social media platform; A heat analysis module is used to predict the initial heat parameters of each scenic spot in the future tourism cycle based on artificial intelligence according to the historical tourism heat information, user evaluation information, user search information and user interaction information of the scenic spot; The heat analysis module is also used to iteratively optimize the initial heat parameters of each scenic spot in the future travel cycle based on the association relationship between the multiple scenic spots, and generate the heat parameters of the multiple scenic spots in the future travel cycle.

Images