Volume 2023, Issue 1 4586047

Research Article

Open Access

A Real-Time Tourism Route Recommendation System Based on Multitime Scale Constraints

Yongmei Song,

Corresponding Author

Yongmei Song

[email protected]

orcid.org/0000-0002-7204-2918

School of Management, Xi’an University of Science and Technology, Xi’an, Shaanxi 710054, China xust.edu.cn

Research Center for Western Tourism Development, Xi’an University of Science and Technology, Xi’an, Shaanxi 710054, China xust.edu.cn

Search for more papers by this author

Xuelian Jiao,

Xuelian Jiao

School of Management, Xi’an University of Science and Technology, Xi’an, Shaanxi 710054, China xust.edu.cn

Search for more papers by this author

Yongmei Song,

Corresponding Author

Yongmei Song

[email protected]

orcid.org/0000-0002-7204-2918

School of Management, Xi’an University of Science and Technology, Xi’an, Shaanxi 710054, China xust.edu.cn

Research Center for Western Tourism Development, Xi’an University of Science and Technology, Xi’an, Shaanxi 710054, China xust.edu.cn

Search for more papers by this author

Xuelian Jiao,

Xuelian Jiao

School of Management, Xi’an University of Science and Technology, Xi’an, Shaanxi 710054, China xust.edu.cn

Search for more papers by this author

First published: 26 April 2023

https://doi.org/10.1155/2023/4586047

Academic Editor: Muhammad Zakarya

Share a link

Email
Wechat
Bluesky

Abstract

In order to increase the capability of real-time intelligent recommendation of tourists’ information on cross-regional city-level tourist routes with epidemic normalization, a real-time intelligent recommendation algorithm for cross-regional city-level tourist routes with epidemic normalization based on multi-time scale constraints is proposed. Under the training of limited samples, the tourist correlation model of the epidemic normalization of cross-regional city-level tourist routes is created. In addition, two kernel functions i.e. the mixed and the global are assembled to excerpt the correspondence features of the epidemic normalization cross-regional city-level tourist route recommendation information. As a result, the well-known particle swarm optimization (PSO) procedure and algorithm with multitime scale constraints are adopted to carry out the adaptive learning of the epidemic normalization cross-regional city-level tourist route recommendation, and the convergence control of the recommended method is comprehended through mining the geographic information data sets of cities. This paper analyzes the universality and ergodicity of tourists′ personal interest preferences and social characteristics in urban tourism and combines a gradient algorithm to carry out particle swarm evolution and self-adaptive optimization for the recommendation of cross-regional city-level tourist routes with a normalized epidemic situation, so as to realize the group real-time intelligent recommendation of tourists’ information on cross-regional city-level tourist routes with the normalized epidemic situation. The model outcomes indicate that the exactitude and precision of cross-regional city-level tourism route information recommendation with this algorithm are decent, and the convergence of the swarm intelligence optimization (SIO) problem is robust, which can circumvent dipping into the local optimal solution in the process of real-time intelligent recommendation of tourism routes and improve the intelligence and global stability of cross-regional city-level tourism route recommendation with epidemic normalization.

1. Introduction

With the speedy improvement of the economy, the public′s corporeal standard has enriched obviously, and to a greater extent, people begin to pay attention to rich spiritual life, which promotes the development of tourism. However, under the background of epidemic normalization prevention and control, it brings great challenges to the tourism industry. At present, people mainly search for travel information on the Internet to customize their travel plans, but the information overload problem on the Internet is becoming more and more serious, so it is necessary to establish a model for recommending cross-regional city-level travel routes with epidemic normalization. Combining the distribution characteristics of cross-regional city-level travel routes with epidemic normalization, the reliability of recommending cross-regional city-level travel routes with network modeling big data cost analysis can be improved [1]. Therefore, to a greater extent, academics begin to recompense consideration to the exploration of travel route planning. Most tourists will be restricted by many factors such as traffic, expenses, time, scenic spots, and hotels. When making travel plans. However, the existing research does not fully consider tourists’ preferences and constraints, and it is difficult to really meet the individual needs of tourists [2].

The normalized cross-regional city-level tourist routes are realized by a personalized network of independent education and statistics facilities rendering to tourists′ personal predilections. Tourists’ statistics and knowledge resource information on the normalized cross-regional city-level tourist routes want to be boosted, categorized, and kept adaptively rendering to the tourists’ predilections [3]. The resource scheduling and personalized recommendation of the normalized cross-regional city-level tourist routes are combined with the multisource distributed design method to improve the resource scheduling ability of the normalized cross-regional city-level tourist routes and the ability to extract tourists’ urban tourism interests and preferences [1]. In the cross-regional city-level tourist routes with complex epidemic normalization, personalized recommendation and resource optimization scheduling are carried out for complex and diverse online learning resources rendering to the tourists’ past visit archives and predilections, so as to encourage the information practice and progress the energetic communication aptitude of the cross-regional city-level tourist routes with epidemic normalization [4]. It is of countless implications to revise the personalized real-time intellectual recommendation procedure of cross-regional city-level tourist routes with epidemic normalization in the optimization design of cross-regional city-level tourist routes with epidemic normalization. In the traditional methods, the personalized endorsement and communal detection approach for cross-regional city-level tourist routes with normalized epidemic situations mainly comprise intelligent PSO procedure, data clustering recommendation system, association rule-based mining process, and fuzzy PID recommendation procedure, etc., and the topological structure of cross-regional city-level tourist routes is optimized according to multilevel epidemic situation normalization [5].

Combining the appropriate features of the recommendation tasks and information sharing, a data clustering center for a personalized recommendation of epidemic normalization cross-regional city-level tourist routes is established, and an algorithm for a recommendation of epidemic normalization cross-regional city-level tourist routes is designed by using association rule mining method, so as to develop and progress the personalized appearance aptitude of epidemic normalization cross-regional city-level tourist routes. Related literature has designed the recommendation algorithm for epidemic normalization cross-regional city-level tourist routes, and achieved certain research results. Among them, in [6] it is proposed based on a community discovery algorithm of epidemic normalization cross-regional city-level tourist routes based on parallel recommendation, compelling the tourist manners value, tourist ingesting value, and faithfulness of the epidemic normalization cross-regional city-level tourist routes as independent variables. The fuzzy decision-making model for public detection of the epidemic normalization cross-regional city-level tourist routes is established, and the optimization design of the epidemic normalization cross-regional city-level tourist route recommendation model is carried out by combining the association mining method. Nevertheless, the calculation cost of this technique is enormous, and the real-time performance of the epidemic normalization cross-regional city-level tourist route recommendation is not good.

In [7], a community discovery and recommendation algorithm of epidemic normalization cross-regional city-level tourist routes founded on the transformation factor assessment and intellectual and adaptive PSO is proposed. The synchronization tag of epidemic normalization cross-regional city-level tourist routes communities is established, the community association attribute features of epidemic normalization cross-regional city-level tourist routes are extracted, and the association amongst the interaction degree within groups and recommendation consequence is investigated. To realize personalized recommendations of cross-regional city-level tourist routes for epidemic normalization, this method has poor intelligence in the process of recommending cross-regional city-level tourist routes for epidemic normalization, and its global optimization ability is not strong [8].

To elucidate the aforementioned difficulties, in this paper we suggest a real-time intelligent recommendation algorithm for cross-regional city-level tourist routes based on multitime scale constraints. Firstly, under the training of limited samples, the tourist correlation model of the epidemic normalization cross-regional city-level tourist routes is created, and the two kernel functions, i.e., the mixed and the global are assembled to excerpt the association features of the epidemic normalization cross-regional city-level tourist route recommendation information, and the well-known PSO procedure with multitime scale constraints is adopted to carry out the adaptive learning of the epidemic normalization cross-regional city-level tourist route recommendation. At that moment, the convergence control of the recommended procedure is comprehended through mining the geographic information data sets of cities, and the universality and ergodicity of individual tourists′ urban tourism interest preferences and social characteristics are analyzed. Combined with the gradient algorithm, the particle swarm evolution and self-adaptive optimization of the epidemic normalization cross-regional city-level tourist route recommendation are carried out, and the real-time intelligent recommendation of the epidemic normalization cross-regional city-level tourist route information is recognized. As a final point, the simulation exploration illustrates the loftier performance of the anticipated technique in enlightening the personalized real-time intelligent recommendation ability of cross-regional city-level tourist routes with the normalized epidemic situation. The following are some of the fundamental contributions of our work:

(1)
We propose a real-time intelligent recommendation algorithm for cross-regional city-level tourist routes based on multitime scale constraints
(2)
The tourist correlation model of the epidemic normalization cross-regional city-level tourist routes is assembled, and two kernel functions, i.e., the mixed and the global are created
(3)
The particle swarm optimization algorithm with multitime scale constraints is adopted to carry out the adaptive learning of the epidemic normalization cross-regional city-level tourist route recommendation

The remaining manuscript is systemized in the subsequent fashion. In section 2, we discuss the tourism recommendation information model and correlation feature extraction. The recommended algorithm optimization process is deliberated in section 3. The simulation experiment and result analysis are validated in section 4. As a final point, section 5 recaps the manuscript and talks about promising future research.

2. Model for Tourism Recommendation and Extraction of Correlation Features

2.1. Recommended Information Transfer Model of Urban Tourist Routes

In order to comprehend the personalized recommendation and characteristics identification of epidemic normalization cross-regional city-level tourist routes, the parallel recommendation procedure is practiced to excerpt the tourists′ interest preferences and mine statistics, and a hybrid system of the mixed and global kernel functions is assembled to excerpt related structures of epidemic normalization cross-regional city-level tourist route recommendation information. The fuzzy decision function of epidemic normalization cross-regional city-level tourist route recommendation evidence is acquired through the restricted sample training. This should be noted that the self-organizing nonlinear mapping, which is denoted by Φ : x ∈ Rⁿ⟶F, is created to characterize the information transmission space of epidemic normalization cross-regional city-level tourist routes, and the recommended information of epidemic normalization cross-regional city-level tourist routes is mapped to high-dimensional feature space by combining fuzzy decision-making and intelligent swarm optimization [9–11]. In fact, this is expected that the recommended sample set for training of cross-regional city-level tourist routes with the normalized epidemic situation is x_i ∈ Rⁿ, in which the personalized characteristic quantity {(x₁, y₁), (x₂, y₂), ⋯, (x_n, y_n)} of cross-regional city-level tourist routes exemplifies the input vector of the anticipated recommendation system, y_i ∈ Rⁿ is the target quantity worth for the personalized recommendation of cross-regional city-level tourist routes, and x_i ∈ Rⁿ is the sample number of cross-regional city-level tourist routes. Collective with the association rule mining procedure, the objective function of the anticipated real-time intelligent and adaptive recommendation system of cross-regional city-level tourist routes with the normalized epidemic situation is as follows:

(1)

Wherein ξ_i and

exemplify the semantic ontology attributes, characteristics, and features and the corresponding association rule variables of cross-regional city-level tourist routes distribution. The generalized learning procedure is implemented to carry out and perform adaptive learning in the development of cross-regional city-level tourist routes recommendation, which is the cost factor recommended for cross-regional city-level tourist routes, and is implemented to understand the punishment control of wrong and incorrect samples. Note that, by adopting the PSO control, the dissimilarity function of the personalized recommendation of cross-regional city-level tourist routes is as follows:

(2)

Wherein α_i and are the recommended attribute values and template category numbers of cross-regional city-level tourist routes, respectively. In addition, the variable K(x_i, x_j) is a symmetric kernel function, in fact, sustaining the Mercer circumstance, and b represents the recommended threshold of cross-regional city-level tourist routes.

This should be noted that each data tuple′s feature vector set is built, and a hybrid kernel function is created by using the local kernel function (RBF kernel function) and the global kernel function (polynomial kernel function) as control decision functions for the recommendation of cross-regional city-level tourist routes with a normalized epidemic situation, and its expression is as follows:

(3)

Wherein represents the fuzzy personalized feature distribution kernel function recommended by cross-regional city-level tourist routes, and represents the RBF kernel function of confidence trustworthiness of cross-regional city-level tourist routes with the normalized epidemic situation. In fact, its purpose is to adjust the impact of both kernel functions over the whole mixed kernel function, which is also known as, the weight coefficient. Therefore, an information transmission model for the recommendation of cross-regional city-level tourist routes with a normalized epidemic situation is constructed, and the tourist characteristics of cross-regional city-level tourist routes with a normalized epidemic situation are extracted and personalized recommendation design is carried out by combining the hybrid optimization method of particle swarm optimization [12].

2.2. Extraction of Correlation Features of Recommended Information of Tourist Routes

Over the foundation of assembling both the hybrid and the universal kernel functions, this paper extracts the correlation features of the recommended information of cross-regional city-level tourist routes with a normalized epidemic situation and adopts the data set mining of urban geographic information to realize the convergence control of the recommendation process, so as to obtain the potential tourist variables of cross-regional city-level tourist routes with a normalized complex epidemic situation, which are expressed as: {S₁, S₂, ⋯, S_L} by four tuples. Taking the tourism interest preference and the distribution of urban geographic and cultural information resources as control constraints, the characteristic abstraction conveyance regulator prototype of the personalized recommendation system for the cross-regional city-level tourist routes is as follows:

(4)

where in

(5)

The aforementioned two formulas, in fact, symbolize the sparse coefficient from relay node U to the monitoring node of cross-regional city-level tourist routes with a normalized epidemic situation, where M_i[t_l > M_m∨M_n, M_m[t_m > M_j, M_n[t_n > M_j, represents the scoring explanatory variable of recommended information of cross-regional city-level tourist routes, and N₂(u) = {v|d_G(u, v) = 2} exemplifies the average mutual information of tourist characteristics U of cross-regional city-level tourist routes with a normalized epidemic situation, where (i ≠ m ≠ n ≠ j, a ≠ b ≠ c). Under the constriction controller of connotation instructions, the prospective association features of cross-regional city-level tourist routes with a normalized complex epidemic situation are obtained using the following equation:

(6)

wherein |Rev(u)| represents the characteristic distribution point set of cross-regional city-level tourist routes, Rev(u) represents the number of tourist nodes, a_ij represents the tourist scoring measurement information of cross-regional city-level tourist routes with a normalized epidemic situation, b_j represents the joint mutual information recommended by cross-regional city-level tourist routes, and Or de r(Rev(u)) represents the matching degree of nodes in the distribution space of cross-regional city-level tourist routes according to their personalized behavior characteristics. Uniting the connotation rule constraints and the chaotic mapping, the self-adaptive clustering of cluster intelligence information of cross-regional city-level tourist routes is carried out, and the recommended cluster output of cross-regional city-level tourist routes is gotten as mathematically illustrated in the following equation:

(7)

Note that in the above equation, the notation f_lg(z) exemplifies the score value of tourist items recommended by cross-regional city-level tourist routes with a normalized epidemic situation, thus obtaining the quadruple expression of correlation feature extraction of cross-regional city-level tourist routes with the normalized epidemic situation as follows:

(8)

In the above formula, Ch(u) represents the association rule coefficient of cross-regional city-level tourism route recommendation, and node v is added to R2(u) in order to comprehend and understand the dynamic characteristic abstraction of personalized tourist features and characteristics [13].

3. Optimization of the Recommendation Algorithm

3.1. The Gradient PSO Procedure

Over the foundation of assembling a hybrid system which is in fact from the mixed and the global kernel function to excerpt the correlation features of the recommendation statistics of epidemic normalization cross-regional city-level tourist routes, an improved design of the real-time intelligent recommendation algorithm for epidemic normalization cross-regional city-level tourist routes is proposed in this paper. For the multiconstraint multiobjective tourist route planning problem, if only the objective function is weighted, the global optimal solution may not be obtained, so this paper practices the enhanced greedy set of rules to elucidate the optimal tourist routes. In the actual tourism scene, tourists are more thinking about which scenic spot is suitable to go next when the current scenic spot is finished. The greed algorithm is one of the most commonly used algorithms to solve the problem of route planning. However, in this paper, aiming at practical problems, an improved greedy algorithm is used to recommend the top M routes with the highest scores for tourists to choose [14].

The improved greedy algorithm is described as follows: tourists start from the recommended hotel, only consider the top N scenic spots with the highest constraint score based on their current location, and only consider the scenic spots they have never visited. Because tourists have to return to the hotel, when tourists visit all the scenic spots and finally return to the hotel, the hotel constraint score of 10I is no longer considered. So we can get M routes, calculate the total score of each route, and finally only keep the top M routes with the highest total score for tourists to choose from. Set the objective function to maximize the total score of the route and the number of types of scenic spots under the condition of satisfying the user constraints. Taking advantage of the initial value sensitivity, regularity, universality, ergodicity, and other advantages of a chaotic map, the universal optimization regulator of the recommended procedure is performed, and as an outcome, the Logistic map is assembled, whose manifestation is mathematically given in the following equation:

(9)

wherein y_i ∈ [0,1] is the random number, μ is the group real-time intelligent recommendation control parameter of cross-regional city-level tourist routes with the normalized epidemic situation. Generally, it takes a value of 4, and a gradient particle swarm algorithm is constructed. d is assumed that in a dimensional gradient particle swarm search space, m represents the population composed of particles and S = {P₁, P₂, ⋯, P_m} represents the clustering center of tourists’ urban tourism interest preference intelligent search particles in the current dimensional solution space [16]. It represents the current optimization speed of the traversing particle

of the cross-regional city-level tourism route with the normalized epidemic situation, and

characterizes the paramount position which is practiced by the particle

itself. The personalized recommendation gradient particle swarm optimization expression of the cross-regional city-level tourism route with the normalized epidemic situation is obtained as follows:

(10)

wherein

is the transmission coefficient and correlation dimension feature quantity of tourists′ urban tourism interest preference mining on the cross-regional urban level tourism routes at the current moment and the next moment of the particle, respectively. In addition, the route with the highest total score is a learning factor for tourists, and W is generally between 25 and 25. Similarly, the variables C₁ and C₂ represent the search radius and global search threshold of gradient particle swarm optimization, respectively, when assuming taking random numbers between [0, 1]. It is the inertia weight of the route with the highest total score for tourists′ preference. Combined with gradient PSO, the step length of the optimization modification, which is iterative, of the recommended procedure of epidemic normalization cross-regional city-level tourist routes is performed [17]. This should be noted that for each interval the modification formula of the real-time intellectual recommendation system for the epidemic normalization of cross-regional city-level tourist routes is obtained as follows:

(11)

(12)

wherein the notation [W_min, W_max] is the range values for the inertia weight factor recommended for cross-regional city-level tourist routes indicating the normalization of the epidemic situation, generally taking (0.5, 0.6).

3.2. Implementation of Real-Time Intelligent Recommendation Algorithm for Tourism Route

The particle swarm optimization algorithm with multitime scale constraints is used to carry out adaptive learning of cross-regional city-level tourism route recommendations for epidemic normalization [14]. The convergence regulator of the recommendation development is comprehended by mining the geographic information data set of cities, and chaos is familiarized within the process of the optimization in terms of the inertia factor W. In the gradient PSO learning process of cross-regional city-level tourism route recommendation for epidemic normalization, search radius R₁ and R₂ are introduced, and the updated formula is as follows:

(13)

wherein R_i(t) ∈ (0,1), i = 1,2, rendering to the gathering of the tourists, the fuzzy feature measure of epidemic normalization cross-regional city-level tourism route distribution is pulled out, the chaos is familiarized into the two learning factors, as denoted by C₁ and C₂, and the learning formula for recommending epidemic normalization cross-regional city-level tourism route is updated as follows:

(14)

(15)

wherein i = 1,2, [C_min, C_max] is the preliminary population prototype for the gradient PSO. After a scenic spot is played, the top N scenic spots with higher scores are searched as the next scenic spots to be played according to the current scenic spots, and whether joining the scenic spot meets the time constraint and cost constraint of tourists is calculated. If not, the scenic spot is abandoned [18]. If all scenic spots no longer meet the time or cost constraint of tourists, the convergence control coefficient obtained after the tourists’ play is defined as illustrated mathematically in the following equation:

(16)

Wherein m is the number of particles in the recommended particle swarm of the cross-regional city-level tourism route with a normalized epidemic situation, F_i is the adaptability of particle i to tourists′ learning of urban tourism interest preference on the cross-regional city-level tourism route with a normalized epidemic situation, F_avg is the average adaptability of the particle swarm, and F is the recommended control objective function, which is used to limit the size of δ², and is expressed as follows:

(17)

If δ² < H, H is a given constant, the precocious judgment and adaptive processing of tourists’ characteristics mining on the cross-regional city-level tourist routes with a normalized epidemic situation are carried out. For the particles falling into a precocious state, the gradient reduction method is adopted to make them jump out of the local optimum, and the realization of this algorithm is described as follows:

(18)

(19)

wherein [V_min, V_max] is the range of particle velocity of self-adaptive recommendation for cross-regional city-level tourist routes with epidemic normalization [18]. By using the connotation regulation taking out and the local optimization controller methods, the updated formula of recommendation for cross-regional city-level tourist routes with epidemic normalization is obtained as follows:

(20)

In the above formulas, t = 1,2, ⋯, T and T represent the maximum iteration times of the population. The convergence regulator of the recommended method is comprehended by mining the geographic information data set of the city. In order to show the system better, it is supposed that the tourist chooses the hotel scenic spot as shown in Figure 1, and the tourist stays in Hotel H, and the tourist expects to get the maximum tourism income. According to the above scenic spots A, B, C, and D, which route should he choose to play, he can get the maximum income while meeting the time constraint of tourists for 8 hours, and the cost constraint is 300 yuan. Assume that n = 2, that is, when the greedy algorithm is used, each scenic spot expands downwards into two scenic spots, and the number of routes with the highest revenue output, at last, does not exceed three. According to the threshold judgment result, whether the convergence criterion is satisfied or not is judged and appropriate decisions are made.

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Schematic diagram of recommendation model of hotels and scenic spots selected by tourists.

The particle swarm optimization and self-adaptive optimization are carried out for the recommendation of epidemic normalization cross-regional city-level tourist routes combined with a gradient algorithm [19]. The convergence regulator of the recommended procedure is comprehended by mining the city′s geographic information data set, and the universality and ergodicity of individual tourists′ urban tourism interests and social characteristics are investigated. In fact, the PSO and the process of self-adaptive optimization are carried out for epidemic normalization cross-regional city-level tourist routes combined with gradient algorithm, so as to comprehend the real-time intelligent recommendation of tourists′ information on epidemic normalization cross-regional city-level tourist routes [20].

4. Simulation Experiment, Results, and Discussion

In order to validate the accuracy and convergence of the anticipated technique in the real-time intelligent recommendation of cross-regional city-level tourist routes for the normalization of the epidemic situation, a simulation experiment was conducted. The system includes eight functional modules, including login and registration, the introduction of hotel attractions, preference constraint selection, tourist route planning, hotel information management, route, itinerary management, user information management, and scenic spot information management. Users who want to use the intelligent planning system for tourist routes must register first, and then they can log in after successful registration. After entering the system, users can first choose the information of the scenic spots they are interested in to browse [21]. Secondly, users can come to the preference constraint selection module, enter their own personalized constraints, including travel time, travel expense budget, and the conditions of favorite hotels and scenic spots, and submit them to the background server.

The server in the background will obtain the personalized constraints of users, recommend the hotels and scenic spots with the highest scores for users according to the model established above, and then plan a number of tourist routes for users to show to them with the maximum profit. Users can also manage their past itineraries anytime and anywhere. Administrators have the authority to add, delete and check the information of hotel and scenic spot users, the quantity of tourists that were verified is 2,000, the trial function of the PSO is assumed as type and combination of the ZDT series test functions, the amount of all particles in a particular particle group in D-dimensional universe is 20, the recommended simulation time is 1.3 s, and the iteration is carried out for 6,000 times. The mining results and findings of tourists’ urban tourism interest preferences on the cross-regional city-level tourist routes with a normalized epidemic situation are shown in Figure 2.

With the method of this paper, the real-time recommendation method of tourist routes based on multitime scale constraints is adopted, and the result of route optimization recommendation is shown in Figure 3.

Taking the results of route optimization and recommendation in Figure 3 as the input test sample set for tourists’ urban travel interest preference for the epidemic-normalized cross-regional city-level tourism route, the adaptive learning of the epidemic-normalized cross-regional city-level tourism route recommendation was performed, and the curve of the convergence was acquired as made known in Figure 4.

According to the analysis of Figure 4, the self-adaptive learning ability of real-time intelligent recommendation of epidemic normalization cross-regional city-level tourist routes by this method is strong, and the universality and ergodicity of urban tourism interest preference and feature mining of epidemic normalization cross-regional city-level tourist routes are good. Additional test the errors of dissimilar approaches in recommending cross-regional city-level tourist routes for epidemic normalization, and get the comparison results as given away in Figure 5. The examination of the results in Figure 5 displays that the accuracy of this method in recommending the information of cross-regional city-level tourist routes with epidemic normalization is decent, and the nondivergence of the swarm aptitude optimization is significantly robust so that the real-time intelligent recommendation process of cross-regional city-level tourist routes with epidemic normalization can be avoided from falling into a local optimal solution.

On this basis, the formulation of tourists’ travel plans first needs to determine the tourist attractions they want to visit and the hotels they stay in. Based on this, this paper sets up a hotel attraction scoring model, sets tourists’ preference design for hotels and attractions, integrates the information of related hotel attractions, and according to the preference information input by tourists, puts a good hotel and attractions for tourists to choose. In this way, there is no need for tourists’ past travel data, so there is a problem of “moving after a cold.” This paper also establishes a multi-constraint and multi-objective tourism route planning model. After tourists have determined their tourist attractions and hotels, the next problem they face is how to determine their tour order [22]. According to the actual needs of tourists, this paper puts forward a tourism route planning model, which sets the starting point and ending point of tourism to be hotels, practices the enhanced greedy procedure and set of rules to solve the model, and finally, obtains the most profitable tourism routes. The model fully considers the realistic factors and has certain rationality and usability.

5. Conclusions and Future Work

According to tourists’ historical visit records and preferences, personalized recommendations and resource optimization scheduling can be carried out to promote the information use of cross-regional city-level tourist routes with the normalized epidemic situation. This paper proposes a real-time intelligent recommendation algorithm for cross-regional city-level tourist routes with the normalized epidemic situation based on multitime scale constraints. Under the training of limited samples, the tourist correlation model of epidemic normalization cross-regional city-level tourist routes is constructed, and the correlation features of epidemic normalization cross-regional city-level tourist routes recommendation information are extracted. In addition, subsequently, the convergence governor of the recommended procedure is comprehended by mining the city′s geographic information data set, and the universality and ergodicity of tourists’ individual urban tourism interest preferences and social characteristics are examined. In fact, the particle swarm evolution and dynamic optimization of epidemic normalization cross-regional city-level tourist routes recommendation are performed by combining the gradient process, so as to comprehend the real-time intelligent endorsement of epidemic normalization cross-regional city-level tourist routes. The research shows that the accuracy of information recommendation of epidemic normalization cross-regional city-level tourist routes by the anticipated process is decent, and the antidivergence of the swarm intelligence optimization is significantly robust so that the real-time intelligent recommendation process of epidemic normalization cross-regional city-level tourist routes can be avoided from dwindling into local optimal solution, and the intelligence and global stability of epidemic normalization cross-regional city-level tourist routes recommendation are enhanced. In fact, this has virtuous application significance in a real-time intellectual recommendation, as well as, in the personalized learning of epidemic normalization cross-regional city-level tourist routes [15].

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This study was supported by Research on the development trend and high-quality development path of cultural tourism integration in Shaanxi, funded by Philosophy and Social Sciences Prosperity Fund of Xi’an University of Science and Technology (No. 2021SZ03).

Open Research

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

References

1 Wen, Yu T., Yeo Peng H., and wang S. W., Efficient keyword-aware representative travel route recommendation, IEEE Transactions on Knowledge and Data Engineering. (2017) 29, no. 8, 1639–1652, https://doi.org/10.1109/tkde.2017.2690421, 2-s2.0-85029041926.
10.1109/TKDE.2017.2690421
Web of Science® Google Scholar
2 Luo G., Wang J., and Wang X., Personalized travel route recommendation using collaborative filtering based on GPS trajectories, International Journal of Digital Earth. (2017) 11, no. 3, 284–307, https://doi.org/10.1080/17538947.2017.1326535, 2-s2.0-85019593735.
10.1080/17538947.2017.1326535
Web of Science® Google Scholar
3 Zheng M.-C., How a map with a tour route recommendation promotes circuitous tourism, Journal of Asian Architecture and Building Engineering. (2015) 14, no. 2, 363–370, https://doi.org/10.3130/jaabe.14.363, 2-s2.0-84929330927.
10.3130/jaabe.14.363
Web of Science® Google Scholar
4 Cao J. X., Dong D., and Xu S., A k-core based algorithm for influence maximization in social networks [J], Chinese Journal of Computers. (2015) 38, no. 2, 238–248.
Google Scholar
5 Wang C., Chen W., and Wang Y., Scalable influence maximization for independent cascade model in large-scale social networks, Data Mining and Knowledge Discovery. (2012) 25, no. 3, 545–576, https://doi.org/10.1007/s10618-012-0262-1, 2-s2.0-84865609190.
10.1007/s10618-012-0262-1
Web of Science® Google Scholar
6 Moosavi H. and Fateme S., A new accessibility based team orienteering approach for urban tourism routes optimization (A Real Life Case) [J], Computers & Operations Research. (2022) 5, no. 4, 12–18.
Google Scholar
7 Chen Li, Retraction Note: air pollution atmospheric environment detection and global tourism line planning management based on genetic algorithm [J], Arabian Journal of Geosciences. (2021) 14, no. 23, 22–29, https://doi.org/10.1007/s12517-021-09114-8.
10.1007/s12517-021-09114-8
Google Scholar
8 Gavilanes Montoya A. V. and RamónVizuete D. D. C., A nature tourism route through GIS to improve the visibility of the natural resources of the altar volcano, sangay national park, Ecuador, Land. (2021) 10, no. 8, https://doi.org/10.3390/land10080884.
10.3390/land10080884
Web of Science® Google Scholar
9 Liao Z. and ZhengLi W., Rough approximation-based approach for designing a personalized tour route under a fuzzy environment, Information Sciences. (2021) 575, no. 34, 338–354, https://doi.org/10.1016/j.ins.2021.02.007.
10.1016/j.ins.2021.02.007
Web of Science® Google Scholar
10 Zyrianov A. I. and Zyrianova I. S., Planning of the interregional tourist route in the urals, Quaestiones Geographicae. (2021) 40, no. 2, 109–118, https://doi.org/10.2478/quageo-2021-0018.
10.2478/quageo-2021-0018
Web of Science® Google Scholar
11 Design of nature tourism route in chimborazo wildlife reserve, Ecuador, International Journal of Environmental Research and Public Health. (2021) 18, no. 10, 5293–5298, https://doi.org/10.3390/ijerph18105293.
10.3390/ijerph18105293
PubMed Web of Science® Google Scholar
12 Lee H. J. and LeeChoi C. K., Sequence-based travel route recommendation systems using deep learning, Korean Institute of Smart Media. (2020) 9, no. 1, 45–50, https://doi.org/10.30693/smj.2020.9.1.45.
10.30693/SMJ.2020.9.1.45
Google Scholar
13 Ulukus S., Yener A., Erkip E., Simeone O., Zorzi M., Grover P., and Huang K., Energy harvesting wireless communications: a review of recent advances, IEEE Journal on Selected Areas in Communications. (2015) 33, no. 3, 360–381, https://doi.org/10.1109/jsac.2015.2391531, 2-s2.0-84925863934.
10.1109/JSAC.2015.2391531
Web of Science® Google Scholar
14 Zhao N., Zhang S., Yu F. R., Chen Y., Nallanathan A., and Leung V. C. M., Exploiting interference for energy harvesting: a survey, research issues, and challenges, IEEE Access. (2017) 5, no. 5, 10403–10421, https://doi.org/10.1109/access.2017.2705638, 2-s2.0-85028749287.
10.1109/ACCESS.2017.2705638
Web of Science® Google Scholar
15 Xiaolu W., Design of travel route recommendation system based on fast Spark artificial intelligence architecture, International Journal of Industrial and Systems Engineering. (2021) 38, no. 3, 328–345, https://doi.org/10.1504/ijise.2021.116237.
10.1504/IJISE.2021.116237
Google Scholar
16 Zheng X., Luo Y., Sun L., Yu Q., Zhang, and Chen S., A novel multi-objective and multi-constraint route recommendation method based on crowd sensing, Applied Sciences. (2021) 11, no. 21, 10497, https://doi.org/10.3390/app112110497.
10.3390/app112110497
CAS Google Scholar
17 Alonso Tabares D., Mora-Camino F., and Drouin A., A multi-time scale management structure for airport ground handling automation, Journal of Air Transport Management. (2021) 90, no. 2021, 101959, https://doi.org/10.1016/j.jairtraman.2020.101959.
10.1016/j.jairtraman.2020.101959
Web of Science® Google Scholar
18 Zhou T., Wu W., Peng L., Zhang M., Li Z., Xiong Y., and Bai Y., Evaluation of urban bus service reliability on variable time horizons using a hybrid deep learning method, Reliability Engineering & System Safety. (2022) 217, 108090, https://doi.org/10.1016/j.ress.2021.108090.
10.1016/j.ress.2021.108090
Web of Science® Google Scholar
19 Li B., Chen Y., Wei W., Wang Z., and Mei S., Online coordination of LNG tube trailer dispatch and resilience restoration of integrated power-gas distribution systems, IEEE Transactions on Smart Grid. (2022) 13, no. 3, 1938–1951, https://doi.org/10.1109/tsg.2022.3141548.
10.1109/TSG.2022.3141548
Web of Science® Google Scholar
20 Petersen N. C., Rodrigues F., and Pereira F. C., Multi-output bus travel time prediction with convolutional LSTM neural network, Expert Systems with Applications. (2019) 120, 426–435, https://doi.org/10.1016/j.eswa.2018.11.028, 2-s2.0-85057741671.
10.1016/j.eswa.2018.11.028
Web of Science® Google Scholar
21 Qin H. and Zhang H., Intelligent traffic light under fog computing platform in data control of real-time traffic flow, The Journal of Supercomputing. (2021) 77, no. 5, 4461–4483, https://doi.org/10.1007/s11227-020-03443-3.
10.1007/s11227-020-03443-3
Web of Science® Google Scholar
22 Jin J., Guo H., Xu J., Wang X., and Wang F. Y., An end-to-end recommendation system for urban traffic controls and management under a parallel learning framework, IEEE Transactions on Intelligent Transportation Systems. (2021) 22, no. 3, 1616–1626, https://doi.org/10.1109/tits.2020.2973736.
10.1109/TITS.2020.2973736
Web of Science® Google Scholar

Citing Literature

All articles

A Real-Time Tourism Route Recommendation System Based on Multitime Scale Constraints

Abstract

1. Introduction

2. Model for Tourism Recommendation and Extraction of Correlation Features

2.1. Recommended Information Transfer Model of Urban Tourist Routes

2.2. Extraction of Correlation Features of Recommended Information of Tourist Routes

3. Optimization of the Recommendation Algorithm

3.1. The Gradient PSO Procedure

3.2. Implementation of Real-Time Intelligent Recommendation Algorithm for Tourism Route

4. Simulation Experiment, Results, and Discussion

5. Conclusions and Future Work

Conflicts of Interest

Acknowledgments

Open Research

Data Availability

References

Citing Literature

Figures

References

Information

About Wiley Online Library

Help & Support

Opportunities

Connect with Wiley