CoFeel: Using Emotions for Social Interaction in Group Recommender Systems Yu Chen Pearl Pu Human Computer Interaction Group Human Computer Interaction Group Swiss Federal Institute of Technology Swiss Federal Institute of Technology CH-1015, Lausanne, Switzerland CH-1015, Lausanne, Switzerland yu.chen@epfl.ch pearl.pu@epfl.ch ABSTRACT information overload. Group recommender systems (GRSs) aim Group and social recommender systems aim to suggest items of to alleviate information overload by suggesting items to a group interest to a group or a community of people. One important issue of people. in such environment is to understand each individual’s preference Group recommendation problem is not only “the sum of and attitude within the group. Social and behavioral scientist have members” (Jameson, 2004). As the audiences move from evidenced the role of emotions in group work and social individuals to groups of people, challenges arise such as communication. This paper aims to examine the role of emotion aggregating preferences and arriving at equilibrium point of for social interaction in group recommenders. We implemented expectations. Picture yourself sitting together with your friends CoFeel, an interface that aims to provide emotional input in group and selecting a music playlist for a birthday party. The selection recommenders. We further apply CoFeel in a GroupFun, a mobile process does not only depend on the verbal indication on group music recommender system. Results of an in-depth field preferences and choices, but also on various non-verbal channels study show that by exchanging feelings with other users, CoFeel such as individuals’ emotion within the group. Social and motivates users to provide feedback on recommended items in a behavioral scientists have long been studying the social role of natural and enjoyable way. Results also show that emotions do emotion in group environment. Our goal is to set a basic serve as an effective and promising element to elicitate users’ understanding of using emotion for social interaction in group attitudes, and that they do have the potential to increase user recommenders with a particular focus on the following two engagement in a group. Based on suggestions collected from users, questions. we propose other potential recommendation domains of CoFeel. 1) What are the roles of emotional information in group recommender systems? Categories and Subject Descriptors 2) How to design such interface that is useful, easy to use H.5.2 [Information Interfaces and Presentation]: User and playful? Interfaces –Graphical user interfaces (GUI), User-centered design. H.5.3 [Information Interfaces and Presentation]: Group To answer these questions, we introduce CoFeel, an affective and Organization Interfaces - Organizational design, Web-based interface that allows users to provide emotional input in interaction recommender systems. We further implemented CoFeel in GroupFun, a mobile group music recommender system. The rest General Terms of the paper is organized as followed. Section 2 discusses existing Design, Human Factors work and how they related with our work, and particularly, why emotions play an important role in group and social environment. Keywords This is followed by design and usage of CoFeel interface in Group and Social Recommender Systems, Interface Design, Section 3 and how to apply CoFeel in GroupFun in Section 4. Mobile Interface, Affective Interface, Emotional Feedback After reporting the results of a small-scale qualitative user study in Section 5, this paper discusses further application scenarios in 1. INTRODUCTION Section 6 and concludes with limitations and future work. Nowadays, sharing, coordination, cooperation and communication among group members are becoming indispensible in online 2. RELATED WORK environment. Such groups can be constituted by families selecting a recipe together, colleagues working on same projects, and social 2.1 Social Interaction in Group club members planning a culture event. These are examples of Recommender Systems small groups, normally less than hundreds of people. In group Jameson studied some of the key user issues for group environment, group decision-making becomes a problem due to recommender systems (Jameson, 2004) and investigated several measures for promoting collaborating and coordination. These measures mainly aim at designing user interfaces to enhance Paper presented at the Workshop on Interfaces for Recommender mutual awareness. Mutual awareness in group recommender Systems 2012, in conjunction with the 6th ACM conference on Recommender Systems. Copyright @ 2012 for the individual papers is systems includes membership awareness, preference awareness held by the papers' authors. This volume is published and copyrighted by and decision awareness. its editors. Membership awareness allows users to check which users are in Interface@RecSys’12, September 13, 2012, Dublin, Ireland the group. Being aware of members in a group facilitates users to decide how to behave and thus enhances trust in a group 48 recommender (Yu, Zhou, Hao, & Gu, 2006). representative conveys to the current online users some key Preference awareness enables users to be aware of the preferences aspects of its corresponding offline user’s responses to a proposed of other members. A user study on PolyLens reveals that users solution. This is one of the few work that employs non-verbal would like to see each other’s preference information, even at the channels in group environment. expense of some degree of privacy loss. Preference awareness in group recommender systems are categorized into three levels: 2.3 Emotion in Recommender Systems zero awareness, partial awareness and full awareness. Zero preference awareness means that users only know their own Musicovery 1 and Stereomood 2 have developed an interactive preferences, as shown in MusicFX (J. F. Mccarthy, 1998a). Zero interface for users to select music category based on their mood. preference awareness systems are simple but do not inspire user Musicovery classifies mood by two dimensions: dark-positive and trust. Partial awareness in group recommenders allows users to energetic calm. It uses highly interactive interface for users to apply preference information from other group members experience different emotion categories and their corresponding (Kudenko, Bauer, & Dengler, 2003). However, it is prone to music. However, such recommender does not support interaction social loafing, a phenomenon when people contribute less in a in social group environment. The main goal of studying social environment than when they work individually. In full recommender systems is to improve user satisfaction. However, preference awareness, users are aware of other members’ satisfaction is a highly subjective metric. Masthoff and Gatt preferences. One typical technique for is Collaborative Preference (Masthoff, 2005) have considered satisfaction as an affective state Specification (CPS) (Jameson, 2004), as presented in CATS, or mood based on the following aspects in socio- and psycho- PocketRestaurantFinder (J. F. Mccarthy, 1998b) and Travel theories: 1) mood impacts judgement; 2) retrospective feelings Decision Forum. CPS in group recommender systems enables can differ from feelings experienced; 3) expectation can influence persuasion, supports preference explanation and justification and emotion and 4) emotions wear off over time. However, they did reduces conflict. Decision awareness is important in helping users not propose any feasible methods to apply the above arrive at a final decision. Decision awareness is a status in which psychological theories. They also proved that in group users are aware of the decision making process of other members. recommender systems, members’ emotion can be influenced by Existing group recommender systems include the following each other, and this phenomenon is called emotional contagion. decision making styles: (1) zero awareness - simply translating the most highly rated solution into action without the consent of any 2.4 Emotions and Decision Making user (e.g. in MusicFX), (2) partial awareness - one or a selected Our everyday experiences leave little doubt that our emotions can set of representatives of the group are responsible for making the influence decisions we make. For instance, experiment results final decisions (e.g. INTRIGUE and PolyLens), and (3) full (Raghunathan & Pham, 1999) showed that in gambling decisions, awareness - arriving at final decision through face-to-face as well as job-selection decisions, sad individuals are biased in discussions (e.g., CATS) or mediated discussions (e.g., MIAU favor of high-risk and high-reward options, whereas anxious (Kudenko et al., 2003) and Travel Decision Forum). However, individuals are biased in favor of the opposite. On the other hand, none of the work addresses the role of emotion in decision- (Isen, 2001) reveals evidence that in most circumstances, positive making or group interaction. affect enhances problem solving and decision making, leading to cognitive processing that is not only flexible, innovative, but also 2.2 Interface in Group Recommender thorough and efficient. (Schwarz, 2000) has addressed the Systems influence of moods and emotions experienced at the time of “Group interfaces differ from single-user interfaces in that they decision making, affective consequences of decisions and the role depict group activity and are controlled by multiple users rather of anticipated and remembered affect in decision making. than single user” (Ellis, J. Gibbs, & Rein, 1991). Therefore, (Bechara, 2004) further proves the influence of emotions on decision-making from neurology. (Velásquez, 1997) and (Gratch interface adequacy has more requirements in group recommenders & Rey, 2000) also modeled emotion-based decision making. compared with individual recommenders. Flytrap (Crossen, Budzik, & Hammond, 2002) visualizes recommended items by 2.5 Social Role of Emotions using colors and locations. Songs personalized for different users (Keltner, 1999) integrate claims and findings concerning the are displayed with different colors, and the closer the songs are to social functions of emotions at the individual, dyadic, group, and the center, the more likely they will be played. PolyLens (Connor, cultural levels of analysis. On dyadic level (a group of two), Cosley, Konstan, & Riedl, 2001) supports three models of emotional expressions help individuals know others’ emotions, visualizing recommendation UI. Group-only interface only beliefs and intensions, and thus rapidly coordinating social displays movies from group recommendation. Composite interactions. Emotional communication also evokes interface displays a list of recommended movies with both group complementary and reciprocal emotions in others that help and individual member predictions. Individual-focused interface individuals respond to significant social events. Emotions serve as shows the items for other individual users’ preferences. CATS (K. incentives or deterrents for other individuals’ social behavior. On Mccarthy et al., 2006) offers users personal space and group group level, emotions have claimed to help individuals solve the space. In group space, each user has a snowflake with a different problem of identifying group members. Displaying emotions may color and the size of snowflake indicates preferences of individual help individuals define and negotiate group-related roles and users. This allows users to check the interest of other users for a status. Collective emotional behavior may also help group particular resort. Additionally, each icon presents a resort, and its members negotiate group-related problems. Study results from size grows or shrinks in accordance with the preference of the whole group. 1 Travel Decision Forum (Taylor, Ardissono, Goy, & Petrone, Musicovery. http://musicovery.com/ 2003) introduces an animated character for each group member 2 Stereomood. http://www.stereomood.com/ currently not available for communication. By responding with speech, facial expressions, and gesture to proposed solutions; a 49 (Ketelaar & Tung Au, 2003) are discussed in terms of an “affect- as-information” model, which suggests that non-cooperating individuals who experience the negative state associated with guilt in a social bargaining game may be using this feeling state as “information” about the future costs of pursuing an uncooperative strategy.(Bowles & Gintis, 2002) suggest that prosocial emotions, such as shame, guilt, (K. Mccarthy et al., 2006)pride, regret, and joy, play a central role in sustaining cooperative relations, including successful transactions in the absence of complete contracting. (Hareli & Rafaeli, 2008) propose that organizational dyads and groups inhabit emotion cycles: emotions of an individual influence the emotions, thoughts and behaviors of others; others’ reactions can then influence their future interactions with the individual expressing the original emotion, as well as that individual’s future emotions and behaviors. (Barsade, 2001) proved that the leaders transfer their moods to group members and that leaders’ moods impact the effort and the coordination of groups. (Hancock et al., 2008) have investigated Figure 1. Geneva Emotion Wheel (Scherer, 2005) emotion contagion and proved that emotions can be sensed in text-based computer mediated communications. We adopt Scherer’s color wheel style and choose 8 emotions for CoFeel Emotion Plate: excited, joyful, surprised, calm, sad, fear, 3. CoFeel: Providing Emotional Input distressed, aroused, as is shown in Figure 2. Each emotion class provides a scale from 1 to 5 indicating the intensity of the emotion. 3.1 Design Goals In order to enhance user engagement in interacting with the As the first step to investigate the social role of emotions, we CoFeel, we design each emotional position as a hole and a ball is design an interface that helps users to provide emotional input. rolling on the surface of emotion plate. Users interact with the Since this input is also users’ feedback, we cross-use “emotional plate by placing the ball in the hole that corresponding to the input” in this paper. We refer to the guidelines for designing emotional state. The aim of using the plate-hole-ball metaphor is recommender systems, proposed by (Pu, Chen, & Hu, 2011). to enhance user affordance to interact with the interface. Designing CoFeel should meet the following design principles. 1. Usefulness. Users are able to provide emotional feedback using CoFeel. 2. Ease to use. Users find CoFeel easy to use and easy to learn. 3. Playfulness. Users find it fun, playful and entertaining to use CoFeel. 3.2 What is it? CoFeel aims to enhance group experience by enhancing self- presence and mutual awareness within a group. By exchanging feelings with other users, CoFeel aims to motivate users to provide feedback on recommended items in a natural and easy way. It is implemented as an infrastructure, which can be easily extended to various group recommendation domains. We choose Geneva Emotion Wheel (GEW) introduced by Scherer (Scherer, 2005) for users to label emotions, i.e., attitude to recommended items, see Figure 1. Using GEW to label emotion has two advantages: natural tagging of discrete categorical words and the possible mapping of these labels to a two-dimensional space (valance-arousal). In each emotion, users can choose different sized circle. As such, users can assign different intensity Figure 2. Interface of CoFeel Emotion Plate values to the emotion they choose. 3.3 How to use it? We implement CoFeel emotion plate on mobile phones. Since we have chosen the metaphor of a plate, it is natural that a ball can roll around the surface. Users can select the emotion, i.e., place the ball, by tilting the plate surface. Once users confirm an emotion, they can simply click a ‘track’ button, which is around the emotional plate, see Figure 3. The phone detects user movement and direction of surface plate using sensors on mobile phones, i.e., accelerometer and gyroscope. We designed this way in order to make the proces more fun and engaging. We have also 50 filtered out constant accelormeter data when users are walking, 4. PROTOTYPE travelling and etc. In this way, users can input their emotion in a stable way. 4.1 GroupFun: a music recommender system In order to test the applicability of CoFeel, we implemented GroupFun, a mobile group music recommender system. Its function is to come up with common playlists for user created groups. Users can create groups and share their music taste with their group members by rating songs in GroupFun. When GroupFun generates a common playlist for a group, the criterion is to take into account the music taste of all of its contributing members. Figure 4 shows the group function of GroupFun. Users can use CoFeel for two purposes: 1) providing emotional feedback to a song and 2) leaving mood traces on the timeline of a song. Figure 4. Group function of GroupFun 4.2 Providing Emotional Feedback to a Song Emotional feedback can be used as an explanation interface for rating. Users can choose the emotion category and its intensity using CoFeel, see Figure 5. As we introduced in Section 3.3, users hold the phone and roll around the indicator ball around the Figure 3. Interacting with CoFeel Emotion Plate surface of emotion plate, as is shown in Figure 6. 51 Figure 7. Visualzing friend’s emotional feedback in GroupFun 4.3 Emotional Traces in Timeline of a Song Figure 5. Providing emotional feeback to a whole song Users can also leave emotional traces throughout the timeline of a song. Figure 8 is an example way to visualize the traces as music score. User emotions are distinguished by different colors, corresponding with colors in CoFeel. Intensities of emotions correspond to the line. The position of dots in the lines represents the relative position of the moment when user leaves emotional comments. For example, a user is listening to “Paradise” from Coldplay. The last two red dots represent users’ emotion towards the end of the song: aroused with different levels of intensities. Figure 6. Interacting with CoFeel in GroupFun After selecting, emotional feedback is recorded with the song, as is shown in Figure 7. The color dots right to the title of a song indicates the type and intensity that users have chosen, which correspond to the colors in CoFeel. The intensity of emotions is visulized with transparency of circles. For example, the song ‘We will rock you’, is rated as an ‘exciting’ song, with the level of 3 out 5. Figure 8. Leaving emotional comments in a timeline for a song 52 in the field of social psychiatry and interviewed them for feedback in emotional design. They first briefly play around with GroupFun 5. Experiment and CoFeel then commented on the design from the theoretical function point of view. 5.1 Goals To the best of our knowledge, our work is the first to propose 5.3 Results providing emotional feedback in group recommender systems. Step 1: Evaluate with normal users Therefore, the main purpose of evaluation is not to prove its We summarize the demographic information as below in Table 1. superiority to other means of feedback or replace them. Rather, we aim to understand users’ opinions towards emotional feedback ID User 1 User 2 User 3 User 4 and the design of CoFeel interface, including their degree of Occupation Student Student Student Consultant acceptance and suggestions. To be more specific, we aim to Gender Male Female Male Female investigate two research questions: Age 22 26 25 32 1) Is emotional feedback useful for social interaction in group recommender systems? Music exp. >12 h/day >8h/day 2 h/day 2-4 h/day 2) Has CoFeel successfully been designed as an effective (App.) and playful interface to provide emotional feedback? Devices for Mobile Computer, Computer Mobile listening phones and car, MP3 phones 5.2 Design and Procedure laptop player In order to answer the above questions, we carried out a small- Music Studying, Working, Relaxing Travelling, scale qualitative user experiment, with emphasis on learning from context designing, cooking, meditation, users through active listening, inspection and observation. In walking driving, music addition to normal users, we also showed GroupFun to domain before lessons experts. Based on the above two types of interviewed users, we sleep divide the experiment to two steps. Sharing Spotify, Facebook, Google+, CDs, music with Facebook Twitter, Facebook, DVDs Step 1: Evaluate with normal users friends Google + Email The goal of evaluate with normal users is to observe how they interact with CoFeel, particularly whether they have encountered any usability problems. However, we evaluate CoFeel interface Table 2. Demographic information of interviewed users using GroupFun, without explicitly telling users what we were evaluating and observing. From the interview process, we discovered some interesting Four users participated in the experiment. Each user is distributed phenomenon. with an Android phone installed with GroupFun. Before experiment, we assigned each participant with a specific group 1. They hardly notice that music is more frequent in their life than with 11 members. The 11 members come from his/her Facebook their perception. When asked how often they listen to music, 3 out friends. Each group is recommended with a music playlist. Since of the 4 interviewed users answered: not very often. However, the accuracy of recommendation is out of scope of this paper, we when we ask them to recall the last song they listened to recently, use choose most popular songs, i.e., top 40 songs in the they finally discover much more scenarios and time that they experiment week. listen to music. This implies that users tend to use listening to music as background tasks. Before exposing users with application and systems, we ask the following questions to warm them up. 2. The methods they listening to music tend to be mobile and pervasive. From user evaluation, we found that 3 out of 4 users 1) How often do you listen to music? listen to music on the go. Such mobile devices can be smart 2) In which context do you listen to music? phones, mp3 players, laptops, in-car entertaining system and etc. 3) Which kind of device do you use to listen to music? 3. They choose music based on different context. When asked what types of music they listen to. Their answers usually start 4) What do you think about the relation between music and with “er”, “well, depends…”. Then they elaborate how they emotion? choose music in different contexts, e.g., studying, driving, 5) Do you share music among friends? cooking etc. During the experiment, the participants explore and experience 4. They are intrinsically willing to share music among friends. GroupFun freely, with particular focus on CoFeel interface. We Surprisingly, all interviewed users share and discuss about songs observe how they interact with GroupFun and CoFeel, the whole among their friends. As one user mentioned, “I share a song with process of which is recorded. In the meantime, they can ask any friends, either because I like it, or I think my friend may like it, or questions and raise their concerns. After the experiment, we ask it include our shared memory, or it suits the current context.” for users’ comments. We further observe users when they are playing around CoFeel Step 2: Interview domain expert emotion plate in GroupFun. Not surprisingly, we observed some Different from experiment with normal users, the goal of common phenomenon during their interaction with system. interviewing domain experts is to understand the role of emotions 1. During the whole process they interact with GroupFun, they in social and group environment and whether CoFeel contributes spend the majority of time exploring CoFeel, out of curiosity and to this purpose. Additionally, the focus shifts from observation to fun. listening for their feedback and suggestions. We invited a doctor 53 2. The first time when they saw the interface, their mental model 5.4 Implications of choosing emotion is by clicking. After few seconds, they We summarize the findings from the above user study about realized how the ball is moving. providing emotional feedback in group recommender systems. 3. They learnt to use CoFeel to keep track of their mood in very 1. Providing emotional feedback enhances mutual awareness of short time. user preferences within a group. Users know the reasons their This implies that given the fact that CoFeel is a novel interface, friends like a song. users enjoy playing with it and can learn how to use it in short 2. A well-designed interface for emotional feedback offers social time. affordance and invites users engagement in the system. When After using interacting with the system, we further interviewed users know the items their friends like and the reasons of liking, them for feedback on the design of CoFeel and its usage in they are more likely to experience the recommended items, i.e., GroupFun. We received both many encouraging and promising music. This encourages users to be more engaged in the system. comments as well as suggestions. 3. Social interaction in turn strengthens users’ sense of social Overall, users were excited to talk about CoFeel emotion plate. As belonging and enhances their emotional state. users commented: “The plate reminded me of a game I played when I was young, very intuitive and entertaining.” “It is simply 6. LIMITATIONS AND DISCUSSIONS artistic and charming.” “I like the visual effect. It is beautiful”. This work has some limitations that we would like to continue in From the received comments, users are generally impressed by the the future. First of all, CoFeel collects explicit emotions reported visual effect of CoFeel. by users. Sometimes, users are not aware of their emotional attitude. Thus we also aim to consider users’ implicit emotional When asked whether CoFeel, i.e., emotional feedback, is useful in feedback. Additionally, the study is limited within individuals GroupFun, all of them agree it is useful. “It is interesting way to with manipulated friend groups instead of users within a group. comment on a song.” “In this way, my friend understand why I Furthermore, as an in-depth qualitative user study, we only invited like this song and I also know their styles and favorite songs a few users and domain experts. In order to further validate our better.” “I used the emotional re-tweeting function in one micro- hypotheses, we need more groups and users and conduct larger blogging system, which is a fast and convenient way to express scale user studies for quantitative analysis. It would also be multi-dimensional meanings.” “Sometimes I don’t know how to interesting to let users use GroupFun with their friends in real life express my feeling and comments for a song. They are abstract and observe their behavior and attitude. and I’m a person of few words. Emotional feedback looks like I’m choosing my comments from a set of words. It is a take-away Despite of the limitations, using emotions for social interaction style. Everything is predefined and very quick.” implies a much broader usage context. CoFeel not only applies in At the mean time, they suggest further application scenarios for music recommender systems but also various other domains. using CoFeel in social interaction. “It will be interesting to see a Based on feedback received from interviewed users, we propose music messaging system where people communicate emotions via the following example domains where emotional feedback can be music.” “What about an interface for mixed emotions?” “Re- useful: movies, tourists, product, hotels, food and etc. One thing tweeting a song attached with emotions would be cool!” in common in the above domain is the capability for the items to elicit emotions. This has been cross validated by social and From the qualitative analysis above, we conclude that CoFeel has behavioral scientists. fulfilled the goals we have set in Section 3.1: usefulness, ease of use and playfulness. 7. CONCLUSIONS Step 2: Interview domain expert We hypnotize that using emotion to enhance social interaction in group recommenders. We have implemented CoFeel, with the Furthermore, we interviewed a doctor in children and adolescent goal of designing an interface that is easy to use and enjoyable for psychiatry. From mental health perspective, he pointed out that users to leave emotional attitude. We further applied CoFeel in discussing with friends with/using music is also used to enhance GroupFun, a group music recommender system in mobile phones. people’s mental health. This process is called music therapy. CoFeel can be used in two modes in GroupFun: elicitation of Music and mood is by nature connected. Meanwhile, encouraging emotional attitude towards a whole song or emotional traces in the discussion about mood among a social group also brings benefit to timeline of a song. We then conducted an in-depth qualitative enhance users’ mental state, under the condition that the process experiment with users, observing their interaction with GroupFun should be fun. This method is also known as social therapy. He and CoFeel, followed by interviews with them. Besides normal also commented on GroupFun with CoFeel as followed. “Your users, we also showed our prototype to domain experts and software, I find it very interesting, especially the idea of self- received positive feedback from them, both theoretically and regulation by the music and the group's involvement even if it is a practically. Results show that providing emotional feedback not virtual interaction. In short, fun and social group, they are two only enhances mutual awareness of user preferences, but also very important elements, not just for people with depression, but encourages social interaction. In essence, providing such social also for everyone who is interested in this type of language. Every affordance using emotions in group environment in turn promotes day, we all have moments of frustration and we all seek for self- users’ enthusiasm in interacting with system. Based on discussion solutions and be content with a group that gives us support and with users, we are more convinced that emotional feedback, i.e., sense of belonging.” CoFeel, applies not only in music domain, but also in many others, From the interview results, we find that theoretically providing such as travel, movie and product recommendations. emotional feedback has a positive effect on encouraging group interaction and engagement. A further discovery is that social interaction that takes place within a group also enhances user mood and mental state. 54 8. ACKNOWLEDGEMENT repeated social bargaining games: An affect-as-information We thank Swiss National Science Foundation for supporting this interpretation of the role of emotion in social interaction. work and project. We also thank all participants for their interest Cognition & Emotion, 17(3), 429-453. in our project, their valuable time and suggestions. doi:10.1080/02699930143000662 [14] Kudenko, D., Bauer, M., & Dengler, D. (2003). Group 9. REFERENCES Decision Making through Mediated Discussions, 238-247. [1] Barsade, S. G. (2001). Working Paper Series OB [15] Masthoff, J. (2005). The Pursuit of Satisfaction : Affective Organizational Behavior “ The Ripple Effect : Emotional State in Group Recommender Systems. User Modelling. Contagion In Groups ”, (August). [16] Mccarthy, J. F. (1998a). MUSICFX : An Arbiter of Group [2] Bechara, A. (2004). The role of emotion in decision-making: Preferences for Computer Supported Collaborative Workouts. evidence from neurological patients with orbitofrontal ACM conference on Computer supported cooperative work. damage. Brain and cognition, 55(1), 30-40. doi:10.1016/j.bandc.2003.04.001 [17] Mccarthy, J. F. (1998b). PocketRestaurantFinder: A Situated Recommender System for Groups (pp. 1-10). [3] Bowles, S., & Gintis, H. (2002). Prosocial emotions. [18] Mccarthy, K., Salam, M., Coyle, L., Mcginty, L., Smyth, B., [4] Connor, M. O., Cosley, D., Konstan, J. A., & Riedl, J. (2001). & Nixon, P. (2006). CATS : A Synchronous Approach to PolyLens : A Recommender System for Groups of Users. Collaborative Group Recommendation. American ECSCW 2001 (pp. 199-218). Association for Artificial Intelligence (pp. 86-91). [5] Crossen, A., Budzik, J., & Hammond, K. J. (2002). Flytrap: [19] Pu, P., Chen, L., & Hu, R. (2011). A user-centric evaluation intelligent group music recommendation. the 7th framework for recommender systems. Proceedings of the international conference on Intelligent user interfaces (pp. fifth ACM conference on Recommender systems - 184-185). RecSys ’11, 157. New York, New York, USA: ACM Press. [6] Ellis, C., J. Gibbs, S., & Rein, G. . (1991). Groupware: some doi:10.1145/2043932.2043962 issues and experiences. Commun. ACM, 34(l). [20] Raghunathan, R., & Pham, M. (1999). All Negative Moods [7] Gratch, J., & Rey, M. (2000). Modeling the Interplay Are Not Equal: Motivational Influences of Anxiety and Between Emotion and Decision-Making. Sadness on Decision Making. Organizational behavior and [8] Hancock, J., Gee, K., & Ciaccio, K. (2008). I’m sad you're human decision processes, 79(1), 56-77. sad: emotional contagion in CMC. Proceedings of the 2008 doi:10.1006/obhd.1999.2838 ACM, 295-298. Retrieved from [21] Scherer, K. R. (2005). What are emotions? And how can they http://dl.acm.org/citation.cfm?id=1460611 be measured? Social Science Information, 44(4), 695-729. [9] Hareli, S., & Rafaeli, A. (2008). Emotion cycles: On the Retrieved from social influence of emotion in organizations. Research in http://ssi.sagepub.com/cgi/content/abstract/44/4/695 Organizational Behavior, 28, 35-59. [22] Schwarz, N. (2000). Emotion, cognition, and decision doi:10.1016/j.riob.2008.04.007 making. Cognition & Emotion, 14(4), 433-440. [10] Isen, A. M. (2001). An Influence of Positive Affect on doi:10.1080/026999300402745 Decision Making in Complex Situations: Theoretical Issues [23] Taylor, P., Ardissono, L., Goy, A., & Petrone, G. (2003). With Practical Implications. Journal of Consumer INTRIGUE: Personalized recommendation of tourist Psychology, 11(2), 75-85. attractions for desktop and handset devices. Applied doi:10.1207/S15327663JCP1102_01 Artificial Intelligence, 37-41. [11] Jameson, A. (2004). More Than the Sum of Its Members : doi:10.1080/08839510390225050 Challenges for Group Recommender Systems, 48-54. [24] Velásquez, J. D. (1997). Modeling Emotion-Based Decision- [12] Keltner, D. (1999). Social functions of emotions at four Making. levels of analysis. Cognition & Emotion, (July 2012), [25] Yu, Z., Zhou, X., Hao, Y., & Gu, J. (2006). TV Program 37-41. Retrieved from Recommendation for Multiple Viewers Based on user Profile http://www.tandfonline.com/doi/abs/10.1080/026999399379 Merging. User Modeling and User-Adapted Interaction, 168 16(1), 63-82. doi:10.1007/s11257-006-9005-6 [13] Ketelaar, T., & Tung Au, W. (2003). The effects of feelings of guilt on the behaviour of uncooperative individuals in 55