Figure - available from: Multimedia Tools and Applications
This content is subject to copyright. Terms and conditions apply.
Source publication
Travel products recommendation has become one of emerging issues in the realm of recommendation systems. The widely-used collaborative filtering algorithms are usually difficult to be used for recommending travel products due to a number of reasons, including (1) the content of travel products is very complex, (2) the user-item matrix is extremely...
Similar publications
The rapid development of the tourism industry and the Internet era has led to an increasingly severe problem of travel information overload, and travel recommendation methods are essential to solving the information overload problem. Traditional recommendation algorithms only target common travel scenarios during the daytime, combining the ratings...
Traveling has become a part of life at the time of the current urbanization. The growth rate of the internet and the availability of information allows travellers to access tourist information easier and faster. However, the individual’s ability to find the information that they want is not proportional to the complexity of the information. This re...
Citations
... With the rapid growth of the national economic level, people's quality of life and living standards have improved, and tourism consumption accounts for an increasing proportion of people's daily consumption [1]. With the rise of the E-commerce industry, the development of data mining technology, machine learning algorithms, the rapid development of intelligent tourism service technology, tourism E-commerce has become an important means for people to obtain tourism information and tourism product booking [2]. With the further improvement of people's quality of life and the further expansion of tourism demand, most of the tourism E-commerce platform tourism product retrieval is complex, the process is redundant, and the product recommendation interface is uniform, which is difficult to meet the needs of people's tourism consumption [3]. ...
Convenient and intelligent tourism product recommendation method, as the key technology of tourism E-commerce platform design, not only provides academic value to the research of tourism E-commerce platform, but also improves the efficiency of personalized recommendation of tourism products. In order to improve the quality of tourism recommendation, this paper proposes a tourism E-commerce platform design method based on HHO improved K-means clustering algorithm. Firstly, the Harris optimization algorithm is used to improve the K-means algorithm to construct a user-oriented tourism product recommendation strategy; then, combined with the XGBoost algorithm, an item-oriented tourism product recommendation strategy is proposed; secondly, the two strategies are mixed to construct a personalized tourism product recommendation model. Finally, the effectiveness of the proposed method is verified by simulation experiment analysis. The results show that the recommendation accuracy of the tourism E-commerce platform design method proposed in this paper reaches more than 90%, and the recommendation response time meets the real-time requirements, which can provide personalized tourism product recommendation for platform users and enhance the purchase of tourism products.
... The idea of predicting the temporal patterns' completion was applied with sequential patterns in several domains, such as complex activity recognition [8,20,21] and recommendations [22,23]. However, many events in daily applications, such as meteorology data, stock fluctuations, or patient data, are not instantaneous events. ...
... Prediction of sequential patterns' completion is useful in many tasks, such as complex activity recognition [8,20,21] and recommendations [22,23]. In code recommendations, it was suggested to use sequential patterns to mine coding patterns from the project repository, and once the developer's code coincides with a beginning of a sequential pattern, the pattern is suggested [23]. ...
... In code recommendations, it was suggested to use sequential patterns to mine coding patterns from the project repository, and once the developer's code coincides with a beginning of a sequential pattern, the pattern is suggested [23]. Zhu et al. [22] introduced a recommendation system for travel products that is based on frequent sequential patterns, in which the sequential patterns comprised of the visited web pages' semantic descriptions and their target products. A model based on frequent sequential patterns is trained to recommend the highest-scored target product given a user's click-stream. ...
In many daily applications, such as meteorology or patient data, the starting and ending times of the events are stored in a database, resulting in time interval data. Discovering patterns from time interval data can reveal informative patterns, in which the time intervals are related by temporal relations, such as before or overlaps. When multiple temporal variables are sampled in a variety of forms, and frequencies, as well as irregular events that may or may not have a duration, time intervals patterns can be a powerful way to discover temporal knowledge, since these temporal variables can be transformed into a uniform format of time intervals. Predicting the completion of such patterns can be used when the pattern ends with an event of interest, such as the recovery of a patient, or an undesirable event, such as a medical complication. In recent years, an increasing number of studies have been published on time intervals-related patterns (TIRPs), their discovery, and their use as features for classification. However, as far as we know, no study has investigated the prediction of the completion of a TIRP. The main challenge in performing such a completion prediction occurs when the time intervals are coinciding and not finished yet which introduces uncertainty in the evolving temporal relations, and thus on the TIRP’s evolution process. To overcome this challenge, we propose a new structure to represent the TIRP’s evolution process and calculate the TIRP’s completion probabilities over time. We introduce two continuous prediction models (CPMs), segmented continuous prediction model (SCPM), and fully continuous prediction model (FCPM) to estimate the TIRP’s completion probability. With the SCPM, the TIRP’s completion probability changes only at the TIRP’s time intervals’ starting or ending point. The FCPM incorporates, in addition, the duration between the TIRP’s time intervals’ starting and ending time points. A rigorous evaluation of four real-life medical and non-medical datasets was performed. The FCPM outperformed the SCPM and the baseline models (random forest, artificial neural network, and recurrent neural network) for all datasets. However, there is a trade-off between the prediction performance and their earliness since the new TIRP’s time intervals’ starting and ending time points are revealed over time, which increases the CPM’s prediction performance.
... However, on one hand, exploring some suitable travel packages from immense amounts of travel data may be time-consuming for users. On the other hand, Online Travel Agencies (OTAs) are eagerly seeking the novel use of recommender techniques to release the potential for business from those data and serve tourists in a highly personalized manner [9,10,33,70,71]. ...
... There has been extensive research on travel packages recommendation in the personalized recommendation literature over the past several years [9,10,14,19,33,52,[68][69][70][71]. In general, the interaction matrix between users and travel packages extracted from travel data is much sparser than that of general merchandises, such as movies, books, musics, and so on [33]. ...
... Although a wide array of studies fall within the aforementioned field, this work is highly related to the second sub-stream: travel products or packages recommendations. Since the clickstream and purchase records provided by OTA are available for researchers, some prior studies [9,10,19,33,47,70] have shown that travel data has some unique characteristics compared with traditional products data, such as extremely sparse data, involving a large number of cold-start users and unavailable ratings. Therefore, traditional recommendation methods such as CF [4] or Matrix Factorization (MF) [27] are hard to achieve good performance on travel data. ...
Session-based travel packages recommendation aims to predict users’ next click based on their current and historical sessions recorded by Online Travel Agencies (OTA). Recently, an increasing number of studies attempted to apply Graph Neural Networks (GNN) to the session-based recommendation and obtained promising results. However, most of them do not take full advantage of the explicit latent structure from attributes of items, making learned representations of items less effective and difficult to interpret. Moreover, they only combine historical sessions (long-term preferences) with a current session (short-term preference) to learn a unified representation of users, ignoring the effects of historical sessions for the current session. To this end, this paper proposes a novel session-based model named STR-VGAE, which fills subtasks of the travel packages recommendation and variational graph auto-encoders simultaneously. STR-VGAE mainly consists of three components: travel packages encoder , users behaviors encoder , and interaction modeling . Specifically, the travel packages encoder module is used to learn a unified travel package representation from co-occurrence attribute graphs by using multi-view variational graph auto-encoders and a multi-view attention network. The users behaviors encoder module is used to encode user’ historical and current sessions with a personalized GNN, which considers the effects of historical sessions on the current session, and coalesce these two kinds of session representations to learn the high-quality users’ representations by exploiting a gated fusion approach. The interaction modeling module is used to calculate recommendation scores over all candidate travel packages. Extensive experiments on a real-life tourism e-commerce dataset from China show that STR-VGAE yields significant performance advantages over several competitive methods, meanwhile provides an interpretation for the generated recommendation list.
... In the tourism industry, it has been repeatedly verified that the recommendation of tourism products is remarkably different from providing traditional items such as movies or books [23,24,32,46,55]. Specifically, the sparsity of the useritem interaction matrix in the context of tourism is extremely high in general [41,53], since the relatively expensive and time-consuming tourism products result in infrequent searching and purchasing. ...
Recently, the recommender system has been raised as one of the essential research topics in smart tourism. The massive card transaction data generated in the tourism industry has become an important resource that implies tourist consumption behaviors and patterns. However, there are challenges such as the high absence possibility of explicit feedback, which is the basis of traditional collaborative filtering techniques, and the consideration of auxiliary factors (e.g., temporal, spatial, and demographic information) that could improve the recommendation performances. In this paper, we introduce TPEDTR, a novel approach using card transaction data to recommend tourism services. It consists of two main components: (i) temporal preference embedding (TPE) models tourist groups’ interactions with services chronologically to obtain their representation vectors. And (ii) deep neural network-based tourism recommendation (DTR) uses the vectors and auxiliary factors as inputs to provide tourist services. To evaluate the TPEDTR, a dataset of card transactions that happened in Jeju island, one of the most famous attractions in South Korea, over eight years is used. Experimental results demonstrate the efficacy of the proposed method and the positive effectiveness of introducing additional information on recommendation performances.
... In the literatures, a large number of personalized recommendation methods have sprang up for travel package recommendation over the past several years [4][5][6][7][8][9]. These methods can effectively capture the non-linear relationship between users and travel packages. ...
... Firstly, unlike traditional e-commerce recommendation systems, the description information of travel package is complicated, making difficulties for researchers to directly incorporate it in existing recommendation systems. Most of the previously proposed methods [8,14] build topic models or sentiment models to learn package classification based on various package information. Intuitively, there are significant differences in the degree of travel entity words attention for different travel information. ...
... POI/travel route recommendation methods focus on understanding users' travel behaviors by mining the human mobility data in daily life and predicting the next location he/she may visit or to generate itinerary under trip constraints. Examples of tourism-oriented Recommendation models include the stochastic approach [19], support vector machine-based models [20][21][22], topic models [8,23]. However, users' preferences may vary dramatically with respect to the geographical regions due to different urban compositions and cultures. ...
Travel package recommendation is a critical task in the tourism e-commerce recommender systems. Recently, an increasing number of studies proposed various travel package recommendation algorithms to improve Online Travel Agencies (OTAs) service, such as collaborative filtering-based, matrix factorization-based and neural network-based methods. Despite their value, however, the main challenges that incorporating complex descriptive information of the travel packages and capturing complicated users' long-term preferences for fine-grained travel package recommendation are still not fully resolved. In terms of these issues, this paper propose a novel model named Neural Attentive Travel package Recommendation (NATR) for tourism e-commerce by combining users' long-term preferences with short-term preferences. Specifically, NATR mainly contains two core modules, namely, travel package encoder and user encoder. The travel package encoder module is developed to learn a unified travel package representation by an attentive multi-view learning approach including word-level and view-level attention mechanisms. The user encoder module is designed to study long-term and short-term preference of the user by Bidirectional Long Short-Term Memory (Bi-LSTM) neural networks with package-level attention mechanism. In addition, we further adopt a gated fusion approach to coalesce these two kinds of preferences for learning high-quality the user's representation. Extensive experiments are conducted on a real-life tourism e-commerce dataset, the results demonstrate the proposed model yields significant performance advantages over several competitive methods. Further analyses from different attention weights provide insights of attentive multi-view learning and gated fusion network, respectively.
... They utilize association rules to identify similarity of users and item, respectively. When temporal information is added onto transactions, the sequential patterns [Hariri et al. 2012;Zhu et al. 2017Zhu et al. , 2016 as well as association rules [Nakagawa and Mobasher 2003] are widely used for recommendation, especially on Web usage data. In these studies, the key problem is to capture user's recent preference during an active session, where the sliding window is often used for controlling the size of session to be matched. ...
The association-rule-based approach is one of the most common technologies for building recommender systems and it has been extensively adopted for commercial use. A variety of techniques, mainly including eligible rule selection and multiple rules combination, have been developed to create effective recommendation. Unfortunately, little attention has been paid to the scalability concern of rule-based recommendation methods. However, the computational complexity of rule-based methods shall increase drastically with the growth of both online customers and rules, which are usually several millions in typical e-commerce platforms. Moreover, the dynamic change of users’ actions requires rule-based methods make recommendations in nearly real-time, which further highlights the scalability issue of rule-based recommender systems. In this article, we present a distributed framework that can scale different association-rule-based recommendation methods in a unified way. Specifically, based on the summarization of existing rule-based approaches, a generic tree-type structure is defined to store separate kinds of patterns, and an efficient algorithm is designed for mining eligible patterns along with computing recommendation scores. To handle the ever-increasing number of online customers, a distributed framework is proposed, where two load-balanced strategies for partitioning tree are put forward to fit sparse and dense data, respectively. Extensive experiments on five real-life data sets demonstrate that the efficiency of association-rule-based recommender systems can be significantly improved by the proposed framework.
... A major topic in addition to user interest in apparel recommendations is the incorporation of up-to-date fashion trends and seasonality into the recommendations, which requires a model that captures sequential and temporal dynamics from user behaviors. The most popular method for representing such dynamics is a graph, although some studies [24,25,26] suggested models that reflect item sequence without using graphs. Several studies [27,28,29] demonstrated that link analysis methods using Markov chains, such as personalized PageRank [30], can be effectively applied to personalized recommendation considering a sequential pattern based on implicit feedback data. ...
Although fashion-related products account for most of the online shopping categories, it becomes more difficult for users to search and find products matching their taste and needs as the number of items available online increases explosively. Personalized recommendation of items is the best method for both reducing user effort on searching for items and expanding sales opportunity for sellers. Unfortunately, experimental studies and research on fashion item recommendation for online shopping users are lacking. In this paper, we propose a novel recommendation framework suitable for online apparel items. To overcome the rating sparsity problem of online apparel datasets, we derive implicit ratings from user log data and generate predicted ratings for item clusters by user-based collaborative filtering. The ratings are combined with a network constructed by an item click trend, which serves as a personalized recommendation through a random walk. An empirical evaluation on a large-scale real-world dataset obtained from an apparel retailer demonstrates the effectiveness of our method.
... In order to help people to quickly retrieve their desirable items from the massive online resources, the recommender system [1,2] has been designed and popularized in different applications. It works to recommend different resources to the target users, and has achieved remarkable success in many systems, e.g., product recommendation systems [3,4] , Netflix video recommendation system [5] , MovieLens [6] , road network services [7][8][9] and social network [10][11][12] . ...
... In this work, the proposed recommendation algorithm will be evaluated on a practical digital publishing system RAYS. 4 The dataset contains a collection of 18,598 tagged multimedia resource items, including books, articles, videos, etc. The system is supporting 805398 subscribers by the end of June 2016. ...
... Here precision is the ratio of the items in the recommendation list hitting the user's item collection, and the recall rate shows the proportion of recommended items in the user's actual item collection. 4 http://dcrays.cn/en/index _ en.html . ...
Along with the fast growing web-based applications, the recommender system is now attracting much attention due to its core function that matches the target users’ interest with the potential resources from the massive online information. Since the recommender system is a user centric application, in this work, we propose a recommendation framework based on user interaction, so as to explore the user's real-time interest from the instant feedback. Naturally, we utilize the tag information assigned to different resources as the medium for user interaction. During the interaction, the most effective tags will be provided for users to choose, and the chosen tag words will be considered as the personalized preference and utilized to dynamically adjust the recommendation list during the process. However, the interaction procedure may cause the problem of potential false dismissal during the candidate filtering. In this work, we propose to analyze the association between different tags, and utilize the tag co-occurrence to refine the recommendation candidate, so as to avoid false dismissal. To generate the recommendation list from the filtered candidates, we design the representation of user and resource characteristics based on tag information and user historical behavior. We distinguish the significance of each tag word for the corresponding resource item, so as to precisely describe the item feature. Probabilistic matrix factorization is employed in our work to overcome the rating sparsity, which is enhanced by embedding the similar user and resource information. The experiments on real-world datasets demonstrate that the proposed algorithm can achieve more accurate predictions and higher recommendation efficiency.
... Firstly, users generally purchase travel products at a relatively low frequency compared with listening to music or watching movies. This may cause the more serious sparsity problem of travel data [16,17,37]. Secondly, due to the relatively high costs both in time and money for a travel, users are more cautious and generally take into account more factors when making such decisions. ...
... De et al. [8] made good use of content-based, collaborative filtering and knowledge-based methods to make recommendations for individuals and groups. Zhu et al. [37] tackled cold-start and sparsity problems through mining the travel products' topic sequential patterns of web logs. Lucas et al. [18] took advantage of classification and association to enhance recommendation quality. ...
The great quantity of travel products available online has increased demand for travel product recommendation system. Due to the relatively high value and time cost of travel products, users consider more factors (personal preference, social preference and seasonality factor etc.) in making this type of low-frequent purchase decisions, compared to other products (e.g. music, movies or news). Thus, recommending travel products generally faces sparsity and complexity problems. In this study, we propose a two-stage multiple-factor aware method named TSMFA. In the topic stage, a user-topic matrix is constructed using travel products’ topic attributions to alleviate sparsity problem, while a preference-aware topic selection is introduced to consider both social and personal preference in recommendation. In the product stage, seasonal prevalence is employed to adjust the recommended product order to incorporate seasonality factor. The proposed method is validated with real transaction dataset from a leading OTA (Online Travel Agent) website in western China. The experimental results demonstrate that it outperforms the state-of-the-art recommendation methods in terms of effectiveness and usefulness.
... Performance of recommendation systems have a huge impact on the commercial success of these companies in terms of revenue and user satisfaction [5]. Some methods have been widely used for general recommendation, such as Content-Based (CB), CF and hybrid solution [6]. The quality of recommendations and usability of six online recommender systems are examined and the results show that RS provided better recommendations [7]. ...
In online e-learning, there is a lot of shared information for learners, but discovering the learner's interest is more difficult because of information overload. Traditional approaches are used to generate recommendations for learners by using content features and ratings of Collaborative Filtering (CF) and content based approaches. The personalized and accurate learning resource recommendations are presented by incorporating the sequential access patterns and learner's context into the Recommender System (RS). In this research, semantic Content-Based filtering and learner's Negative Ratings (CBNR) is presented to recommend the interesting messages to the learners. The experiments are evaluated to validate the learners' performance and recommendation accuracy of the proposed system with existing e-learning recommendation system. Furthermore, the obtained outcome proves that the learning performance has been increased in terms of accuracy and quality of recommendations when compared with the similar recommendation techniques.
