Design and implementation of an intelligent recommendation system for tourist attractions: The integration of EBM model, Bayesian network and Google Maps

Hsu, Fang-Ming; Lin, Yu-Tzeng; Ho, Tu-Kuang

doi:10.1016/j.eswa.2011.09.013

Cited by 99 publications

(52 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…It reveals that MAE of the model in this paper decreases along with number of similar user group (5,10,15,20), value of MAE are 0.46、 0.32、0.21、0.17. MAE of traditional collaborative filtering method also decreased along with number of similar user group (5,10,15,20), value of MAE are 0.68、 0.47、 0.35、 0.26. When the number of similar users are the same, the value of MAE of the model in this paper is smaller than that of traditional collaborative filtering method, comparison of MAE is as figure 5.…”

Section: Princinple Of Personalized E-tourismmentioning

confidence: 62%

“…Due to the feature of tourism, researchers have conducted study of improving personalized recommendation method along with user feature and the whole industry including expanding user information dimension to come up with user model with the analysis of users' behavior [3] and solving the cold start and new user new need problem of personalized recommendation method [4] [5].However, multi-dimensional measurement of users' information is a must besides basic demographic information, historical behavior user preference and active situation also has influence on recommendation. Acquiring more factors of recommendation and lowing search cost of travelers, context theory has become one of the effective methods to conduct personalized product/service recommendation and accomplish interaction [6].…”

Section: Literature Reviewmentioning

confidence: 99%

“…, as (5). The multi-dimensional separate-weight user preference model is as (6) Cosine-based similarity is used to calculate user similarity in this paper.…”

Section: Princinple Of Personalized E-tourismmentioning

confidence: 99%

See 2 more Smart Citations

Personalized e-tourism attraction recommendation based on context

Wei

2013

2013 10th International Conference on Service Systems and Service Management

View full text Add to dashboard Cite

Based on analysis of mobile tourism users' multi-dimensional feature, the concept of context is introduced into user model modeling of mobile tourism. From the perspective of user and context, context theory and machine learning is used to accomplish user modeling in terms of tourism activities recommendation. The dimension of this model includes history behavior, current context, historical context and demographic factor. The problems of new user and similar recommendation and lack of weight are settled in this paper. According to the impact of multi dimension to user preference, user preference interfering is used to acquire user preference to accomplish multi-dimensional user model based on context model to contribute to improvement of traditional e-tourism recommendation and personalization and adaptability of platform.

show abstract

Section: Princinple Of Personalized E-tourismmentioning

confidence: 62%

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Personalized e-tourism attraction recommendation based on context

Wei

2013

2013 10th International Conference on Service Systems and Service Management

View full text Add to dashboard Cite

show abstract

“…To produce high-quality results, recommendation systems use data for all of the users and items, the quantities of which are extremely large (Carullo et al 2015). Determining how to efficiently mine the data and provide recommendations under the conditions of large data remains a key issue (Hsu et al 2012). Moreover, for many systems, with a moderate increase in the numbers of items and users, the computation complexity of the recommendation system dramatically increases, which is called not scalable.…”

Section: The New Challengesmentioning

confidence: 99%

Improved collaborative filtering with intensity-based contraction

Cui

Jiang

Liu

et al. 2015

J Ambient Intell Human Comput

View full text Add to dashboard Cite

Recommendation systems are essential tools for piquing consumers' interests and stimulating consumption in today's electronic commerce, and the quality of these systems depends on the employed filtering algorithms. Therefore, improving the performance of these algorithms is an important issue. In this paper, we design an intensity-based contraction (IC) algorithm that works in combination with other machine-learning algorithms in model-based collaborative filtering, which is currently the most popular filtering algorithm. The main challenges for this algorithm are sparseness of the database and lack of scalability. To demonstrate how IC is used, we implemented IC clustering as an example, which can effectively reduce the sparseness of the database and improve the efficiency. Moreover, we created a scalable IC on a MapReduce model, the scalability of which is demonstrated with actual experiments.

show abstract

“…Hsu et al (2012) propose a touristic trip forecasting and intelligent recommendation system. Also, they use Google Maps API to allow the user to adjust the geographic data according to personal needs.…”

Section: Fig 3 Hierarchical Task Analysis and Menu Itemsmentioning

confidence: 99%