Creativity

The emergence of novelties and their rise and fall
in popularity is an ubiquitous phenomenon in human
activities. The coexistence of always popular
milestones with novel and sometimes ephemeral trends
pervades technological, scientific and artistic
production. By introducing suitable statistical
measures, we demonstrate that different systems of
human activities, i.e. the creation of hashtags in
Twitter, the interaction with online program code
repositories, the creation of texts and the
listening of songs on an on-line platform, exhibit
surprisingly similar properties. We then introduce
a general framework to explain those
regularities. We propose a simple mathematical model
based on the expansion into the adjacent possible,
that has been proven to be a very general and
powerful mechanism able to explain many of the
statistical patterns emerging in innovation
dynamics, to which we add two crucial elements. On
the one hand we quantify the idea that, while
exploring a conceptual or physical space, inertia
exists towards known already discovered elements. On
the other hand, we highlight the role of the
collective dynamics – where many users interact, in
a direct or indirect way in the emergence and
diffusion of novelties and innovations.

The study of the dynamics behind the emergence of
novelties and innovation is a relatively recent
field of study in complex systems, fostered by the
abundance of data about the creations and sharing of
artworks and about on-line activity in
general. Despite this recentness, many works have
been able to discover and characterise several
interesting statistical patterns related to the
emergence of new creative elements and a very
general mathematical framework describing the
collective process of discovering and sharing
novelties come out. However, still a lot has to be
discovered concerning the conditions, either
historical and social, fostering the emergence of
creative elements from a group of interacting
individuals. From a social perspective, many
hypotheses have been developed and tested
concerning the relations between individual like the
presence of ?weak ties? in social networks or the
?folding? of different social groups into a larger
one sharing a common goal. Complex Systems Science
has given little contributions to the
understanding of how the dynamics behind social
interactions contributes to foster the emergence of
creativity. This work of thesis is devoted to the
analysis of data collected during a collective
social experiment in which individuals were asked to
collaborate in the realisation of a set of LEGO
bricks sculptures. The participants to the
experiments were provided with particular RFID tags,
developed in the framework of the SOCIOPATTERNS
project, that enabled a quite precise mapping of the
social interactions occurring during their activity
within the experiment. The interaction with the LEGO
Sculptures were similarly mapped by means of other
RFID tags placed around the sculptures, and their
growth in volume has been recorded with the aid of
infra-red depth sensors. The RFID sensors allowed
for a reconstruction of the dynamical network of
social interactions between the participants in the
experiment. We looked for correlations between the
evolving structure of this social net- work and the
growing patterns of the sculptures, spotting the
local social structures more prone for a rapid
growth of the volume in small amounts of times and
in long term periods. In this way, we were able to
identify the social patterns more fruitful in terms
of ?local consensus? around the development of the
collective artwork, indicating a shared vision
around the actions to be performed on it. Moreover,
we were able to identify how the presence of
‘influential individuals’ characterised by means of
information spreading models favoured the growth of
the sculptures in the long-term. The novelty behind
the proposed approach could contribute to shed light
on the phenomena related to creativity and could be
useful in conceiving and designing new collecting
creativity experiments.

The recent spread of social networks and ICT systems
has allowed for a huge availability of data on
social phenomena and collective behaviour. This has
induced a deep change in social dynamics field, that
moved from an essentially theoretical approach to a
strongly data driven one. In such framework, the
present work aims at exploring the collaboration
dynamics and the organisational structures within
the GitHub platform. Moreover, the purpose is using
success and popularity as feedbacks to check whether
some particular structures exist that are associated
with more efficiency, better results and
subsequently more innovative features in the
development of the code. GitHub is based on the Git
revision control system and is currently the most
important platform for open source coding, counting
millions of repositories and active users. Moreover,
the complete timeline of GitHub activity is publicly
accessible on the GitHub Archive website. GitHub is
therefore a particularly suitable system to observe
and analyse collective social behaviours and
collaborative dynamics. The collaboration among
users fosters an uninterrupted flow of new ideas
which actualise in many different events such as the
creation of new projects and updating of existing
ones through code modifications. The analysis
required a preliminary selection of the data
downloaded from GitHub Archive in order to create a
database containing all the necessary information
about projects activity. The analysis carried out on
this database was mostly inspired by previous
research on innovation dynamics in the framework of
complex systems. Every project was mapped in a
network structure in order to observe dynamically
the development and the modifications of the
code. Some metrics were defined that could estimate
the collaboration degree among users and the
organization of the workload within the developing
branches. Other metrics were chosen in order to
evaluate both the success and the popularity reached
by a project and its potential
innovation. Correlation analysis between the metrics
and the indexes above mentioned allow for some
evaluations about the interdependence between
attention received and structural features of the
projects. This thesis work follows up several
quantitative analyses on GitHub presented in
literature and proposes a new visualisation of
internal structures and collaborative dynamics
within GitHub projects. Moreover, identifying
successful patterns could help in highlighting the
most influential and pioneering projects and
encouraging their development.

It is common opinion that many innovations are
triggered by serendipity whose notion is associated
with fortuitous events leading to unintended
consequences. One might argue that this
interpretation is due to the poor understanding of
the dynamics of innovations. Very little is known,
in fact, about how innovations proceed and samples
the space of potential novelties. This space is
usually referred to as the adjacent possible, a
concept originally introduced in the study of
biological systems to indicate the set of
possibilities that are one step away from what
actually exists. In this paper we focus on the
problem of defining the adjacent possible space, and
analyzing its dynamics, for a particular system,
namely the cultural system of the network of movies.
We synthesized to this end the graph emerging from
the Internet Movies Database (IMDb) and looked at
the static and dynamical properties of this
network. We deal, in particular, with the subtle
mechanism of the adjacent possible by measuring the
expansion and the coverage of this elusive space
during the global evolution of the system. Finally,
we introduce the concept of adjacent possibilities
at the level of single node and try to elucidate its
nature by looking at the correlations with
topological and user annotation metrics.

The dynamics of political votes has been widely
studied, both for its practical interest and as a
paradigm of the dynamics of mass opinions and
collective phenomena, where theoretical predictions
can be easily tested. However, the vote outcome is
often influenced by many factors beyond the bare
opinion on the candidate, and in most cases it is
bound to a single preference. The voter perception
of the political space is still to be elucidated. We
here propose a web experiment (laPENSOcos`ı) where
we explicitly investigate participants’ opinions on
political entities (parties, coalitions, individual
candidates) of the Italian political scene. As a
main result, we show that the political perception
follows a Weber-Fechner-like law, i.e., when ranking
political entities according to the user expressed
preferences, the perceived distance of the user from
a given entity scales as the logarithm of this
rank.

Studies in literature and narrative have begun to
argue more forcefully for considering human
evolution as central to understanding stories and
storytelling more generally (Sugiyama, 2001;
Hernadi, 2002). However, empirical studies in
language evolution have focused primarily on
language structure or the language faculty, leaving
the evolution of stories largely unexplored
(although see Von Heiseler, 2014). Stories are
unique products of human culture enabled principally
by human language. Given this, the dynamics of
creativity in stories, and the traits which make
successful stories, are of crucial interest to
understanding the evolution of language in the
context of human evolution more broadly. The current
work aims to illuminate how stories emerge, evolve,
and change in the context of a collaborative
cultural effort. We present results from a novel
experimental paradigm centered around a story game
where players write short continuations (between 60
and 120 characters) of existing stories. These
continuations then become open to other players to
continue in turn. Stories are subject to player
selection, allowing for variation and speciation of
the resulting narratives, and evolve as a result of
collaborative effort between players. The game
starts with a seed of over 60 potential stories, and
players choose which stories to continue, providing
a player-driven story selection mechanism. In this
way, stories which are creative, intriguing, and
open ended spawn more stories, and eventually lead
to longer story paths as play continues. The game
also introduces further limitations by constraining
a players’�� view of the story path: players have
access only to a story and its parent, meaning
knowledge of the existing narrative is limited. We
present data from hundreds of players and stories,
creating large story trees which explore the space
of different possible narratives which grow out of a
confined set of starting points. This data allows us
to investigate several aspects of the growing story
trees to illuminate not only what makes a story
successful, but how creative stories trigger new
stories, and what makes individual storytellers
successful. Given the selection mechanism central to
game play, we identify the most successful stories
by their number of offspring. Particularly
successful storytellers emerge measured both by how
many children their stories have spawned, and also
how long their story path extends. We also show that
coherent stories often emerge, despite the fact that
they are authored by several different players, and
any given player only sees a limited snapshot of the
story path. We contextualise the results of the game
and connect it to language evolution in two
ways. First, we look for detectable triggers of
innovation and creativity within the story trees,
and identify these as expanding the adjacent
possible (e.g., new adaptations open the space of
other possible adaptations in the future; Tria,
Loreto, Servedio, & Strogatz, 2014). We argue that
this concept can be extended to stories, using
evidence from the game bolstered by evidence from
more traditional literature (the Gutenberg
Corpus). Second, we frame the results in terms of
recurring themes found in storytelling
cross-culturally (Tehrani, 2013). We suggest that
the most successful triggers of innovation in
stories combine original novelty and a firm
grounding in existing recurring story frameworks in
human culture. This indicates that much like other
cultural and biological systems, stories are subject
to competing pressures for stability and
conservation on the one hand, and innovation and
novelty on the other.

The understanding and the characterisation of
individual mobility patterns in urban environments
is important in order to improve liveability and
planning of big cities. In relatively recent times,
the availability of data regarding human movements
have fostered the emergence of a new branch of
social studies, with the aim to unveil and study
those patterns thanks to data collected by means of
geolocalisation technologies. In this paper we
analyse a large dataset of GPS tracks of cars
collected in Rome (Italy). Dividing the drivers in
classes according to the number of trips they
perform in a day, we show that the sequence of the
travelled space connecting two consecutive stops
shows a precise behaviour so that the shortest trips
are performed at the middle of the sequence, when
the longest occur at the beginning and at the end
when drivers head back home. We show that this
behaviour is consistent with the idea of an
optimisation process in which the total travel time
is minimised, under the effect of spatial
constraints so that the starting points is on the
border of the space in which the dynamics takes
place.

Coping with the complexities of the social world in
the 21st century requires deeper quantitative and
predictive understanding. Forty-three
internationally acclaimed scientists and thinkers
share their vision for complexity science in the
next decade in this invaluable book. Topics cover
how complexity and big data science could help
society to tackle the great challenges ahead, and
how the newly established Complexity Science Hub
Vienna might be a facilitator on this path.

The comprehension of vehicular traffic in urban
environments is crucial to achieve a good management
of the complex processes arising from people
collective motion. Even allowing for the great
complexity of human beings, human behavior turns out
to be subject to strong constraints—physical,
environmental, social, economic—that induce the
emergence of common patterns. The observation and
understanding of those patterns is key to setup
effective strategies to optimize the quality of life
in cities while not frustrating the natural need for
mobility. In this paper we focus on vehicular
mobility with the aim to reveal the underlying
patterns and uncover the human strategies
determining them. To this end we analyze a large
dataset of GPS vehicles tracks collected in the Rome
(Italy) district during a month. We demonstrate the
existence of a local optimization of travel times
that vehicle drivers perform while choosing their
journey. This finding is mirrored by two additional
important facts, i.e., the observation that the
average vehicle velocity increases by increasing the
travel length and the emergence of a universal
scaling law for the distribution of travel times at
fixed traveled length. A simple modeling scheme
confirms this scenario opening the way to further
predictions.

We introduce a model for music generation where
melodies are seen as a network of interacting
notes. Starting from the principle of maximum
entropy we assign to this network a probability
distribution, which is learned from an existing
musical corpus. We use this model to generate novel
musical sequences that mimic the style of the
corpus. Our main result is that this model can
reproduce high-order patterns despite having a
polynomial sample complexity. This is in contrast
with the more traditionally used Markov models that
have an exponential sample complexity.