Welcome to the international (English-speaking) blog of the Department of Computer Science at the University of Helsinki. Our blog invites views on research, education, student life, and other societal themes connected to our computer science fields. The intention is to build a forum as an open window for readers from inside and outside computer science. If you have any good ideas or articles to share on this blog, please feel free to contact us: cs-blog [ät] cs.helsinki.fi.
Computational Creativity - Now!
by Anna Kantosalo
Computational creativity is a new and hot topic in computer science. This year I had a wonderful opportunity to participate in the International Conference for Computational Creativity (ICCC) at the Jozef Stefan Institute in Ljubljana and present my first academic paper for the Discovery Research Group. The great community and fascinating topics of the conference made me want to share the highlights of my first conference experience with all of you!
This blog post offers a personal take on some of the most interesting topics of the conference. For a more detailed introduction to what computational creativity is, and how the Discovery Research Group studies it here at the University of Helsinki.
Put Your Computational Creativity Goggles On
What makes computational creativity interesting, is that our research goes beyond computers. Although a large part of our work goes to building simulations and developing algorithms, our study of creative behavior is not restricted to computers. As researchers we need a wider knowledge of what creative behavior is, and how it appears in single individuals or whole societies. It is also important to study art and think of important philosophical questions, such as what is creativity in the first place. Only then can we start building programs to act creatively.
This ideology can also be seen in at the ICCC - while the majority of papers focuses on presenting new algorithms and more complex experiments and programs for generating art like poems, songs or paintings, a great amount of the presentations also covered more multidisciplinary topics. These topics included talks about humans creating together with computers, simulations of the effects of social policies to creativity, and philosophical issues on evaluating the creativity of computer programs. The audience of the conference is also wide - in addition to computer scientists, designers and artists alike are welcome to attend.
Hot Topics of the Conference
My personal favourite among this year’s topics was co-creation, a category to which I contributed to with my own interaction design related presentation. I had the honor of being the first speaker of the whole conference. From my first-timer’s perspective this seemed quite a challenge, but it turned out to be a blessing, since I was able to concentrate on everyone else’s speeches after that. My own talk reflected on the topic of design processes for co-creative software from the viewpoint of interaction design [9]. It also presented our own interactive educational co-creation tool for writing poetry together with a computer.
From other contributions to co-creation I liked especially Robert Tubb’s take on different user strategies in communicating with a co-creative music composition program [16]. I was also impressed by Nicholas Davis’s talk on collaborative painting with computer colleagues [5]. His examples of computer and human working together were very impressive.
The topic of why and how computers are creative is one of the most interesting philosophical questions in the community. The first big question is, if we can assess the creativity of a computer without relying on its output. Both Simon Colton [3] and Diarmund O’Donoghue [12] presented new ideas for assessing creativity without the output. The second big question is related to metrics. When assessing creativity by output, we tend to pay a great deal of attention to how valuable and novel the output is. This year, Oliver Bown [2] and Kazjon Grace [7], argued that these measures are ill defined and demanded incorporating a human observer in the process.
The more “traditional” fields of computational creativity, such as music and poetry generation, had each their own session at the conference. As the highlights of the sessions I would like to mention Jukka Toivanen’s presentation on poetry generation based on news articles [15] . The presentation was kicked off by having a robot from the Jozef Stefan institute read out aloud one of the poems generated by Jukka’s program. The little artificial man gave a true performance, which won a heartfelt laughter from the audience. This way of presenting computer poetry seemed like a breakthrough; I remember thinking, how much more effective the poem felt when presented by a machine.
In the visual arts section, I was impressed by Dan Ventura’s presentation on how the DARCI system is gradually learning to convey the meaning of adjectives in its images [11]. In music, the presentation by François Pachet on creative transformation of notes to harmonisations with its examples was very captivating [13]. However for me the most interesting advances were taken in narrative: Guerrero Román’s talk on incorporating social norms into story generation [8], Rafael Pérez y Pérez’ presentation on a learning version of the Mexica system [14], and Tony Veale’s work on character development [17] were adding vividness and credibility to existing story generation systems.
Computer game enthusiasts will be happy to hear this year’s conference included a session on computational creativity in games. Games have been a very fruitful field for computational creativity to grow on, as the field hungers especially for content producing methods. Antonios Liapis presented a paper focusing on these ideas [10], he argued for mutual benefits as well as a more philosophical approach, in which the playing of a game is actually a creative event itself. This also produces some interesting possibilities for analysing gaming from a co-creation perspective. Michael Cook, the proud father of the famous game composing system Angelina, presented the most recent advances in Angelina’s work [4]. Most curious were his results from human evaluations of Angelina’s work at a game competition - they clearly showed that in games, humans tend to have a very positive outlook on the products made by a computer, whereas in many fields, such as poetry, the bias is negative.
On the creative society simulation front, this year’s conference boasted two interesting contributions: Liane Gabora discussed the effect of implementing social regulation in a society of artificial artists [6] and Oliver Bown discussed the effect of allowing for the cumulation of wealth acquired through creative behavior [1]. Gabora’s presentation showed, how enforcing a self regulation rule at the level of each agent produced two distinct classes of artists: the creators and imitators. This interesting finding goes to show similar societal mechanisms may contribute to the distinction between a creative and non-creative segment within our own society. Bown’s presentation then again focused on the question of why creative behavior has been useful from an evolutionary perspective.
While this blog post can not cover all the papers in the conference, you can look them up online. Almost all talks were taped (http://www.twitch.tv/computationalcreativity) and the whole proceedings is available online for free (http://computationalcreativity.net/iccc2014/proceedings/).
Extracurricular Activities
This year’s conference did not approach creativity solely from a scientific point of view, but also served as a place to present art with a computational focus. The conference was accompanied by an exhibition featuring works from artists participating in the conference. There was also a panel discussion between artists and scientists on computational creativity and its future. Especially artistic responsibility and ownership were discussed. This topic is especially interesting, since it seems to come up often enough when I’m discussing computer poetry in class or with friends.
The first day of the seminar also held an introductory tutorial to present the highlights of computational creativity. The educative theme was continued, when Tony Veale launched his textbook on computational creativity (http://robotcomix.com/comix/Catalogue/mobile/) the next day. Tony’s fully illustrated comic-style book seems like something I’d even love to read at bedtime!
This year’s invited speaker, professor Oliver Deussen also brought an interesting perspective to computational creativity. He presented the work of his research group on a painting robot talking about physically rendering non-photorealistic images. The robot, which according to Deussen is not creative, was a marvellous piece of work enabling numerous possibilities for further development of creative machines. I daresay everyone in the conference was keen to get their hands on the robot and put it to use in producing its own creative material.
Top trends in the hallway discussions were Twitterbots, and multidisciplinary approaches to computational creativity. The former was seen as a great way to disseminate the products of computational creativity in a compact and accessible way. The second as a way to broaden the spectrum of computational creativity research and help it have a genuine impact on more specialised fields. At the same time both of these are likely to get more and more gifted individuals interested in computational creativity and contribute to this new and interesting field.
Next year the ICCC travels to Sundance in Utah, where our own professor Hannu Toivonen is going to be the program chair. I hope to go there and see the latest trends for myself, but if not possible, I’m sure to go there online!
References
[1] Bown, O. (2014) A Model of Runaway Evolution of Creative Domains. 5th International Conference on Computational Creativity
[2] Bown, O. (2014) Empirically Grounding the Evaluation of Creative Systems: Incorporating Interaction Design. 5th International Conference on Computational Creativity
[3] Colton, S., Pease, A., Corneli, J., Cook, M. & Llano, M. T. (2014) Assessing Progress in Building Autonomously Creative Systems. 5th International Conference on Computational Creativity
[4] Cook, M. & Colton, S. (2014) Ludus Ex Machina: Building A 3D Game Designer That Competes Alongside Humans. 5th International Conference on Computational Creativity
[5] Davis, N., Popova, Y., Sysoev, I., Hsiao, C., Zhang, D. and Magerko, B., (2014), Building Artistic Computer Colleagues with an Enactive Model of Creativity. 5th International Conference on Computational Creativity
[6] Gabora, L. & Tseng, S. (2014) The Social Impact of Self-Regulated Creativity on the Evolution of Simple versus Complex Creative Ideas. 5th International Conference on Computational Creativity
[7] Grace, K. & Maher, M. L. (2014) What to Expect when you’re Expecting: The Role of Unexpectedness in Computationally Evaluating Creativity. 5th International Conference on Computational Creativity
[8] Guerrero Román, I. & Pérez y Pérez, R. (2014) Social Mexica: A Computer Model for Social Norms in Narratives. 5th International Conference on Computational Creativity
[9] Kantosalo, A., Toivanen, J. M., Xiao, P., & Toivonen, H. (2014) From Isolation to Involvement: Adapting Machine Creativity Software to Support Human-Computer Co-Creation. 5th International Conference on Computational Creativity
[10] Liapis, A., Yannakakis, G. N., & Togelius J. (2014) Computational Game Creativity. 5th International Conference on Computational Creativity
[11] Norton, D., Heath, D. & Ventura, D. (2014) Autonomously Managing Competing Objectives to Improve the Creation and Curation of Artifacts. 5th International Conference on Computational Creativity
[12] O’Donoghue, D. P., Power, J., O’Briain, S., Dong, F., Mooney, A., Hurley, D., Abgaz, Y., & Markham, C. (2014) Can a Computationally Creative System Create Itself? Creative Artefacts and Creative Processes. 5th International Conference on Computational Creativity
[13] Pachet, F. & Roy, P. (2014) Non-Conformant Harmonization: The Real Book in the Style of Take 6. 5th International Conference on Computational Creativity
[14] Pérez y Pérez, R. (2014) The Three Layers Evaluation Model for Computer-Generated Plots. 5th International Conference on Computational Creativity
[15] Toivanen, J. M., Gross, O., & Toivonen, H. The Officer Is Taller Than You, Who Race Yourself! Using Document Specific Word Associations in Poetry Generation 5th International Conference on Computational Creativity
[16] Tubb, R. & Dixon, S. (2014) A Four Strategy Model of Creative Parameter Space Interaction. 5th International Conference on Computational Creativity
[17] Veale, T. (2014) Coming Good and Breaking Bad: Generating Transformative Character Arcs For Use in Compelling Stories. 5th International Conference on Computational Creativity
Comments
The CS Blog Task Force
Paba is a PhD student in Ubiquitous Interaction Group (UIx group) at
HIIT, CS Department of UH. Her research focus is on developing
interaction models to predict user interests and information-needs in
exploratory search.
Aaron is doing his PhD in the NODES group at the CS department. His research focuses on mobile computing and energy efficient design for multi-interfaced mobile devices.
Ella is a PhD student in the Nodes group. She is interested in e.g. distributed algorithms, real-life data mining, clouds and ubiquitous computing.
Giulio is a Professor at the CS department. His area is Human-Computer Interaction. For more information, please find his homepage here
Tomi is a Professor at the CS department. His area is Software Engineering. For more information, please check
Add comment