=Paper= {{Paper |id=Vol-2140/paper7 |storemode=property |title=Theory-driven Recommendations: Modeling Hedonic and Eudaimonic Movie Preferences |pdfUrl=https://ceur-ws.org/Vol-2140/paper7.pdf |volume=Vol-2140 |authors=Marko Tkalčič,Bruce Ferwerda |dblpUrl=https://dblp.org/rec/conf/iir/TkalcicF18 }} ==Theory-driven Recommendations: Modeling Hedonic and Eudaimonic Movie Preferences== https://ceur-ws.org/Vol-2140/paper7.pdf

Theory-driven Recommendations: Modeling
Hedonic and Eudaimonic Movie Preferences

Marko Tkalčič1 and Bruce Ferwerda2
1
Faculty of Computer Science, Free University of Bozen-Bolzano, Piazza
Domenicani 3, I-39100, Bozen-Bolzano, Italy
marko.tkalcic@unibz.it
2
Department of Computer Science and Informatics, School of Engineering,
Jönköping University, P.O. Box 1026, SE-551 11, Jönköping. Sweden
bruce.ferwerda@ju.se

Abstract. Most of the research in recommender systems focuses on
data-driven approaches. In this paper we present our vision for com-
plementing data-driven approaches with model-driven ones. We present
a preliminary experimental set-up and we expose our research plan. In
the experimental set-up we acquired eudaimonic characteristics of movies
and user preferences. Furthermore, we performed a preliminary analysis
of the acquired data.

Keywords: recommender systems · personality · hedonic emotions ·
eudaimonic emotions

1 Introduction

Mainstream research in recommender systems is data-driven. Logs of user be-
haviour, such as clicks, ratings, purchases, are used in a variety of algorithms,
such as collaborative and feature-based approaches to generate recommenda-
tions [6]. In recent years we have worked on complementing these bottom-up,
data-driven approaches with top-down, model-driven approaches to recommen-
dations.
We view recommender systems as tools for helping users make better deci-
sions [7]. Psychology research has shown that human decisions are influence by
diverse factors, among others by personality and emotions [3]. In our past work
we focused mainly on these two models [8, 2, 10]. In this paper we present the
preliminary results of introducing a known psychological model, that of eudaimo-
nia in user modeling and recommender systems. The experience of consumption
of an item (listening to a song, watching a movie) does not have only hedonic
qualities (fun, enjoyment, relaxation) but also eudaimonic qualities, which are
related to meaning and purpose [4].

IIR 2018, May 28-30, 2018, Rome, Italy. Copyright held by the author(s).
2 Tkalčič and Ferwerda

2 Eudaimonic Modeling and Related Work
We assume that users differ in their need for eudaimonic experiences, i.e. some
people prefer to just have fun, while other people may prefer to spend their time
contemplating meaning and purpose. Similarly, we observe that movies differ in
the experience quality they induce. For example, the movies The Hangover and
La vita e’ bella are both comedies. But while the former is a shallow comedy with
a series of simple jokes the latter deals with deeper issues, such as the holocaust.
In positive psychology, happiness is often described through two opposite con-
cepts: hedonism and eudaimonism [1]: the hedonic view equates happiness with
pleasure, comfort, and enjoyment, whereas the eudaimonic view equates happi-
ness with the human ability to pursue complex goals which are meaningful to the
individual and society. Oliver and Raney [5] have carried out research to identify
whether there are distinct eudaimonic and hedonic motivations for consuming
entertainment. Through a series of studies they devised an instrument for mea-
suring the eudaimonic and hedonic qualities of entertainment experiences. They
showed that in addition to viewing movies for purposes of fun and pleasure,
individuals also turn to entertainment for purposes of greater insight and mean-
ingfulness. Wirth et al. [11] further extended Oliver’s work by analyzing what
are the hedonic and eudaimonic qualities of movies with good and bad endings
and found significant differences.

3 Work Plan
In order to devise personalization approaches using eudaimonia there are a lot
of steps to make, since it is an unexplored area. We foresee the following steps
need to be taken:
1. unobtrusive inference of eudaimonic and hedonic user preferences
2. automatic labeling of movies’ eudaimonic and hedonic qualities
3. a personalized recommender system that takes advantage of eudaimonic and
hedonic features

4 Data acquisition
We performed a user study to acquire data. We let the subjects choose movies
from a pool of popular movies. For the hypothetical context of choosing a movie
to watch alone on a Saturday evening, each subject had to choose the most
appropriate movie (the liked movie) and the least appropriate movie (the disliked
movie). The subjects were then asked to describe, for each of the two movies,
their viewing experience in terms of hedonic and eudaimonic experience using
an adaptation of the scale developed by [5]. After answering the movie-related
questions, the subjects filled in the ten-items personality questionnaire (TIPI).
We hand-picked the movies in order to have a mix of hedonic and eudaimonic
movies.
Hedonic and Eudaimonic Movie Preferences 3

Table 1. Excerpt of movie titles used in the experiment. The eudaimonic and hedonic
qualities are our subjective assessments

Title Eudaimonic Hedonic
Manchester by the sea Y N
Bad Moms N Y
Mad Max: Fury Road N Y
The Shawshank Redemption Y N
Inside Out Y Y

We ran the study through Amazon Mechanical Turk. After removing subjects
who did not pass a control question and removing outliers using the Mahalanobis
distance we had the answers of 84 subjects (M = 34.2 years, SD = 9.5 years,
29 females).

5 Results

A deeper analysis of the data acquired is going to be presented at the UMAP
2018 conference [9]. Here we pick two specific aspects: (i) bimodal distribution
of eudaimonic reasoning and (ii) movie characteristics.
Users divided themselves into two clusters in terms of the eudaimonic qual-
ities of the liked movies (see Fig. 1): some users liked movies with high eudai-
monic qualities (scores > 3) while some users liked movies with low eudaimonic
qualities (scores < 3), which is reflected in the bimodal shape of the histogram
in Fig. 1. We conjecture that this bi-modal shape is due to user being either
pleasure-seekers or meaning-seekers.
We clustered the movies into three categories: hedonic-only, eudaimonic-only
and mixed. We performed the Wilcoxon rank sum test in order to test the
hypothesis of the mean reported hedonic and eudaimonic quality being equal.
Examples from all three clusters are reported in Tab. 2.

Table 2. Examples from clusters of movies. The left column shows movies that have
a stronger eudaimonic quality, the right column shows movies with a stronger hedonic
quality and the mid column shows movies that are equally hedonic and eudaimonic

Eudaimonic Mixed Hedonic
Arrival La La Land Deadpool
Passengers Hidden Figures Mad Max: Fury Road
The Girl on the Train Bad Moms
Fifty Shades of Gray
4 Tkalčič and Ferwerda

I liked the movie because it challenged my way of seeing the world.

20
count

1 2 3 4 5
dfClean$eud_1_1

Fig. 1. Distribution of eudaimonic qualities of liked movies. The variable reported in
this figure is the agreement with the statement I liked this movie because it challenged
my way of seeing the world.

6 Future Work

The results of our study indicate that eudaimonic characteristics are useful for
accounting variance in user preferences and for characterizing movies.
We plan to proceed the with the three steps: (i) unobtrusive inference of
eudaimonic/hedonic preferences, (ii) automatic labeling of movies and (iii) per-
sonalized recommendations of movies.
For the unobtrusive inference we are designing an experiment. In addition to
the variables acquired in the experiment reported above, we will collect also the
users’ social media data. We plan to ask for Facebook likes, twitter activity and
Instagram activity. Using features extracted from social media activity we plan
to do a predictor of the user preferences for movies in the hedonic/eudaimonic
space. The recent scandal with Facebook data integrity may pose a further prob-
blem in devising such a method.
For the automatic labeling of movies we plan to use movie subtitles for feature
generation. We foresee the usage of NLP techniques and generate features using
TF-IDF and embeddings. In order to get ground truth movie labels we plan to
crowd-source the labeling of a pool of movies.
Hedonic and Eudaimonic Movie Preferences 5

For the personalized recommendation part we plan to use content-based rec-
ommendation methods that take advantage of the eudaimonic and hedonic fea-
tures.

References
1. Antonella Delle Fave, Fausto Massimini, and Marta Bassi. Hedonism and Eu-
daimonism in Positive Psychology, pages 3–18. Springer Netherlands, Dordrecht,
2011.
2. Bruce Ferwerda, Marko Tkalcic, and Markus Schedl. Personality Traits and Music
Genres. In Proceedings of the 25th Conference on User Modeling, Adaptation and
Personalization - UMAP ’17, pages 285–288, New York, New York, USA, 2017.
ACM Press.
3. Jennifer S. Lerner, Ye Li, Piercarlo Valdesolo, and Karim S. Kassam. Emotion and
Decision Making. Annual Review of Psychology, 66(1):799–823, 2015.
4. Elisa D. Mekler and Kasper Hornbæk. Momentary Pleasure or Lasting Meaning?:
Distinguishing Eudaimonic and Hedonic User Experiences. Proceedings of the 2016
CHI Conference on Human Factors in Computing Systems - CHI ’16, pages 4509–
4520, 2016.
5. Mary Beth Oliver and Arthur A. Raney. Entertainment as Pleasurable and Mean-
ingful: Identifying Hedonic and Eudaimonic Motivations for Entertainment Con-
sumption. Journal of Communication, 61(5):984–1004, 2011.
6. Francesco Ricci, Lior Rokach, and Bracha Shapira, editors. Recommender Systems
Handbook. Springer US, Boston, MA, 2015.
7. Francesco Ricci, Lior Rokach, and Bracha Shapira. Recommender Systems: Intro-
duction and Challenges. In Recommender Systems Handbook, volume 54, pages
1–34. Springer US, Boston, MA, 2015.
8. Marko Tkalcic and Li Chen. Personality and Recommender Systems. In Francesco
Ricci, Lior Rokach, and Bracha Shapira, editors, Recommender Systems Handbook,
volume 54, pages 715–739. Springer US, 2nd edition, 2015.
9. Marko Tkalcic and Bruce Ferwerda. Eudaimonic Modeling of Moviegoers. In
UMAP 2018. ACM, 2018.
10. Marko Tkalcic, Ante Odic, Andrej Kosir, and Jurij Tasic. Affective Labeling in a
Content-Based Recommender System for Images. IEEE Transactions on Multime-
dia, 15(2):391–400, feb 2013.
11. Werner Wirth, Matthias Hofer, and Holger Schramm. Beyond Pleasure: Explor-
ing the Eudaimonic Entertainment Experience. Human Communication Research,
38(4):406–428, 2012.