Inma Garcia

On the design of individual and group recommender systems for tourism

Research highlights? Recommender system able to provide recommendations for groups of tourists. ? Group members must be equally satisfied with the recommendation. ? Two techniques are explored: intersection and aggregation. ? Intersection obtains better results, it brings together the preferences of all group members. This paper presents a recommender system for tourism based on the tastes of the users, their demographic classification and the places they have visited in former trips. The system is able to offer recommendations for a single user or a group of users. The group recommendation is elicited out of the individual personal recommendations through the application of mechanisms such as aggregation and intersection. The elicitation mechanism is implemented as an extension of e-Tourism, a user-adapted tourism and leisure application whose main component is the Generalist Recommender System Kernel (GRSK), a domain-independent taxonomy-driven recommender system.

e-Tourism: a tourist recommendation and planning application

e-Tourism is a tourist recommendation and planning application to assist users on the organization of a leisure and tourist agenda. First, a recommender system offers the user a list of the city places that are likely of interest to the user. This list takes into account the user demographic classification, the user likes in former trips and the preferences for the current visit. Second, a planning module schedules the list of recommended places according to their temporal characteristics as well as the user restrictions; that is the planning system determines how and when to perform the recommended activities. This is a very relevant feature that most recommender systems lack as it allows the user to have the list of recommended activities organized as an agenda, i.e. to have a totally executable plan.

A Group Recommender System for Tourist Activities

This paper introduces a method for giving recommendations of tourist activities to a group of users. This method makes recommendations based on the group tastes, their demographic classification and the places visited by the users in former trips. The group recommendation is computed from individual personal recommendations through the use of techniques such as aggregation, intersection or incremental intersection. This method is implemented as an extension of the e-Tourism tool, which is a user-adapted tourism and leisure application, whose main component is the Generalist Recommender System Kernel (GRSK) , a domain-independent taxonomy-driven search engine that manages the group recommendation.

Preference elicitation techniques for group recommender systems

A key issue in group recommendation is how to combine the individual preferences of different users that form a group and elicit a profile that accurately reflects the tastes of all members in the group. Most Group Recommender Systems (GRSs) make use of some sort of method for aggregating the preference models of individual users to elicit a recommendation that is satisfactory for the whole group. In general, most GRSs offer good results, but each of them have only been tested in one application domain. This paper describes a domain-independent GRS that has been used in two different application domains. In order to create the group preference model, we select two techniques that are widely used in other GRSs and we compare them with two novel techniques. Our aim is to come up with a model that weighs the preferences of all the individuals to the same extent in such a way that no member in the group is particularly satisfied or dissatisfied with the final recommendations.

A negotiation framework for heterogeneous group recommendation

Over the last years, some remarkable recommender systems for group of users have been developed. When using most of these systems, each group member communicates his/her preferences to the system, which obtains a group profile as the result of an equal weighting of the individual preferences. This way, no member is particularly dissatisfied with the recommendations. However, this is not a realistic situation, given that not all the members in a group act in the same manner. This paper deals with the problem of recommendation for a group of users, where, besides his/her own preferences, each user may have different expectations about the result of the recommendation and may exhibit a different behaviour with respect to the other group members. Moreover, all this information is private and may be revealed under certain circumstances. In this context, we have opted for building a multi-agent system, where an agent acts on behalf of one group member. We have implemented a UserAgent that can be configured in order to exhibit the behaviour desired by the corresponding user. Then, different UserAgents negotiate with the aim of building a group profile that satisfies their particular minimum requirements, while preserving some privacy. Moreover, we have designed a NegotiatorAgent, which governs the negotiation and may act as a mediator in order to facilitate the agreement. Finally, we have performed some experiments that show that this mechanism is able to give a response in this heterogeneous environment.

e-Tourism: A Tourist Recommendation and Planning Application

e-Tourism is a tourist recommendation and planning application to assist users on the organization of a leisure and tourist agenda. First, a recommender system offers the user a list of the city places that are likely of interest to the user. This list takes into account the user demographic classification, the user likes in former trips and the preferences for the current visit. Second, a planning module schedules the list of recommended places according to their temporal characteristics as well as the user restrictions; that is the planning system determines how and when to perform the recommended activities. This is a very relevant feature that most recommender systems lack as it allows the user to have the list of recommended activities organized as an agenda, i.e. to have a totally executable plan.

A Multi Agent Architecture for Tourism Recommendation

In this paper, we present a Multi Agent System aimed to support an user on the realization of different leisure and tourist activities in a city. The system integrates agents that cooperate to dynamically capture the user profile and to obtain a list of suitable and satisfactory activities for the user, by using the experience acquired through the interaction of the user and similar users with the system. Moreover, the system is also able to generate a time schedule of the list of recommended activities thus forming a real activity plan. This paper focuses on the architecture and functional behaviour of our system.

Approaches to preference elicitation for group recommendation

Recommendation can be defined as the problem of selecting, among a set of items, those ones that are likely of interest to the user. In case of a group of users, recommendations should satisfy, as far as possible, the preferences of all the group members. In order to elicit the group preferences, we present two different mechanisms: the first one consists in a voting procedure whereas the second is based on a negotiation procedure. In both cases, intelligent agents act on behalf of the group members. The experimental results show the pros and cons of both approaches and highlight which of the two mechanisms returns the highest-valued recommendation for the whole group in each case. Moreover, we also study which approach is able to reflect more easily the different behaviour of each user, which is also an important aspect in group recommendation.

The Generalist Recommender System GRSK and Its Extension to Groups

This paper presents a Generalist Recommender System Kernel (GRSK) and describes the differences of the recommendation process when it is applied to groups. The GRSK is able to work with any domain as long as the domain description is represented within an ontology. Several basic techniques like demographics, content-based or collaborative are used to elicit the recommendations, as well as other hybrid techniques. The GRSK provides a configuration process through which to select the techniques and parameters that best suit the particular application domain. The experiments will show the success of the GRSK in different domains. We also outline the changes and new techniques required by the GRSK when it is used in a group recommendation.