Group Recommender Systems in Tourism: From Predictions to Decisions Tom Gross University of Bamberg Kapuzinerstr. 16, 96047 Bamberg Germany tom.gross@uni-bamberg.de ABSTRACT Decision making is a core aspect of recommender systems, since the basic assumption is that the system provides suggestions Recommender systems help users to identify goods or services— helping users to make informed decisions [9]. For instance, typically by offering suitable items from a broad range of Jameson et al. have identified diverse patterns of humans making alternatives. They have successfully spread into many domains. a choice [5]. Decision making has also been discussed in the Tourism is a domain that has a huge potential for simplifying context of group recommender systems, but it has been pointed selections and decisions (e.g., on destinations; on itineraries; on out that ‘only a few studies that concentrate on accommodation; on cultural activities). In this position paper I decision/negotiation support in group recommender systems’ [2, discuss how groups of tourists can benefit from group p. 30]. recommender systems and give some examples. In this position paper I share two examples for our own work on group recommender systems: the AGReMo process model for CCS CONCEPTS recommendation and decision processes; and the MTEatSplore • Human-centered computing → Collaborative and social interactive tabletop applications for groups. computing devices 2 THE AGREMO PROCESS MODEL KEYWORDS The AGReMo (Ad-hoc Group Recommendations Mobile) process Recommender Systems, Group Recommender Systems, Decision model was conceived to serve as a blueprint for our group Process, Tourism Recommender Systems recommender systems that aim to support the full cycle of a recommender process starting with a preparation, followed by a decision, and leading to action. Since it has already been 1 INTRODUCTION published elsewhere [1], we just quickly glance at it. The increasing diversity of information, goods, and services offers As Figure 1 shows, AGReMo consists of three principal consumers a huge choice. At the same time finding the preferred phases: item can be challenging. Recommender systems help users The Preparation Phase kicks off the process by collecting all making choices by offering suitable items. They have spread into the required data that are needed to later estimate the predictions many domains (e.g., book recommendations; music and movie and generate recommendations. Each group member creates a recommendations) [9]. personal profile. In our case the process model originated from a Tourism is a domain with a huge potential for recommender group recommender systems for movies, so the individual users systems to help users to reduce the complexity of planning and created a profile and rated movies that they had already seen. deciding, since ‘planning a vacation usually involves searching After that the group members meet and elect an agent who for a set of products that are interconnected (e.g., transportation, interacts with the group recommender system (i.e., the assumption lodging, attractions) with limited availability, and where is that the whole group meets face-to-face and therefore only contextual aspects may have a major impact (e.g., spatiotemporal needs one system). Then the group can optionally specify group context)’ [7]. preferences and set preferred parameters. In our case of movie Group recommender systems support groups of users who recommendations the group can pre-select cinemas and movies in want to share information, experiences, or products. Private the region. The group members can furthermore also optionally travelling and touristic activities often happen in pairs or groups: specify vote weights (i.e., the default was that all group members people travel with a partner, people travel with family, people have equal influence on the recommendation generation, but the travel to meet friends, but also in business travelling colleagues group can assign stronger weights to a member, for instance, as a might travel together or travel to meet working partners or courtesy or due to different levels of expertise). The active agent colleagues. Here, group recommender systems are particularly then requests recommendations, and the system generates group suited, since ‘a group recommender is more appropriate and recommendations. useful for domains in which several people participate in a single In the Decision Phase the group members receive the activity’ [8, p. 199]. recommendations with the best prediction on top. The group RecTour 2017, August 27th, 2017, Como, Italy. 40 Copyright held by the author(s). Figure 1. The AGReMo process model. Source: [1]. recommendations are ranked according to the least misery respective restaurant. Figure 2 shows a scenario of the multi-user aggregation strategy (i.e., maximising the minimal prediction in multi-touch interaction with the restaurant recommendations. the group) [6]. The group can optionally retrieve details for each Further details on the design process and results can be found it recommendation, and can also go through suggestions with lower [3]. recommendations. They can discuss face-to-face and eventually come to a conclusions. In the Action Phase the group would go to the cinema together and each member is then asked to rate the watched movie to further develop their own profile. The group might also dissolve if no consensus can be reached. This model was then instantiated in multiple apps (e.g., on app for Android; another app for iOS) and explored with users and based on real-time movie data that were retrieved from our project partner moviepilot [4]. 3 THE MTEATSPLORE INTERACTIVE TABLETOP APPLICATION FOR GROUPS In a different project on a group recommender system we explored the suitability and affordances of interactive tabletops for Figure 2. MTEatSplore scenario showing the multi-user multi- supporting groups of users in choosing from a set of touch interaction with the restaurant recommendations. recommendations generated and presented by the tabletop app. Source: [3]. Here the primary focus was on a concept for the user interaction and user interface for the group decision phase. We 4 CONCLUSIONS started by developing paper prototypes that allows each team In this position paper I have suggested that in the tourism domain member to pick their personal favourite restaurant and to suggest it is often groups of users who travel together or meet during trips it to the group through a drag-and-drop gesture towards the centre and can benefit from group recommender systems that suggest of the table. The table then clusters and aggregates and counts items of information, services, or goods that are relevant to the nominations as well as allows the group members to drill down whole group. Group recommender systems in tourism face similar for textual as well as visual background information for the RecTour 2017, August 27th, 2017, Como, Italy. 41 Copyright held by the author(s). Group Recommender Systems for Tourism [2] Delic, A., Neidhardt, J. and Nguyen, T.N. Resaerch Methods for Group Recommender Systems. In Workshop on Recommenders in Tourism - RecTour 2016; Co-Located with the 10th ACM Conference on Recommender Systems - challenges and can benefit from contributions and solutions from RecSys 2016 (Sept. 15, Boston, MA). 2017. pp. 30-37. other domains. [3] Fetter, M. and Gross, T. Engage! Empower! Encourage! - Supporting Mundane Group Decisions on Tabletops. In Proceedings of the 2nd International In the Workshop on Recommenders in Tourism at the 11th Conference on Distributed, Ambient, and Pervasive Interactinos - DAPI 2014 ACM Conference on Recommender Systems I would love to (June 22-27, Crete, Greece). Springer-Verlag, Heidelberg, 2014. pp. 329-336. [4] Gross, T. Supporting Informed Negotiation Processes in Group Recommender discuss ideas and concepts for future work on the whole process Systems. i-com - Journal of Interactive Media 14, 1 (Jan. 2015). pp. 53-61. of group recommender systems—including technical aspects on [5] Jameson, A., Willemsen, M.C., Felfernig, A., de Gemmis, M., Lops, P., Semeraro, G. and Chen, L. Human Decision Making and Recommender how to generate recommendations that reach broad acceptability Systems. In Ricci, F., Rokach, L. and Shapira, B., eds. Recommender Systems as well as conceptual aspects of group decision making based on Handbook. (2nd ed.). Springer-Verlag, Heidelberg, 2015. pp. 611-648. group interaction with recommendations. [6] Masthoff, J. Group Recommender Systems: Combining Individual Models. In Ricci, F., Rokach, L., Shapira, B. and Kantor, P.B., eds. Recommender Systems Handbook. Springer-Verlag, Heidelberg, 2011. pp. 677-702. ACKNOWLEDGMENTS [7] Neidhardt, J., Fesenmaier, D., Kuflik, T. and Woerndl, W. Workshop on Recommenders in Tourism. https://recsys.acm.org/recsys17/rectour/, 2017. I would like to thank the members of the Cooperative Media Lab. (Accessed 22/6/2017). Part of the work has been funded by the German Research [8] O'Connor, M., Cosley, D., Konstan, J.A. and Riedl, J. PolyLens: A Recommender System for Groups of Users. In Proceedings of the Seventh Foundation (DFG GR 2055/2-1). Thanks to the anonymous European Conference on Computer-Supported Cooperative Work - ECSCW reviewers for valuable comments. 2001 (Sept. 16-20, Bonn, Germany). Kluwer Academic Publishers, Dordrecht, 2001. pp. 199-218. [9] Ricci, F., Rokach, L. and Shapira, B. Introduction to Recommender Systems REFERENCES Handbook. In Ricci, F., Rokach, L. and Shapira, B., eds. Recommender Systems [1] Beckmann, C. and Gross, T. AGReMo: Providing Ad-Hoc Groups with On- Handbook. (2nd ed.). Springer-Verlag, Heidelberg, 2015. pp. 1-34. Demand Recommendations on Mobile Devices. In Proceedings of the European Conference on Cognitive Ergonomics - ECCE 2011 (Aug. 24-26, Rostock, Germany). ACM, N.Y., 2011. pp. 179-183. RecTour 2017, August 27th, 2017, Como, Italy. 42 Copyright held by the author(s).