By Lance Weiler – In the upcoming issue of Filmmaker Magazine I write about the value of data to filmmakers. In my column I look at a number of projects and then tie them back into how they could be used by filmmakers to aid the curation, disovery and creation of films. One of the projects that I focus on in the piece is a data harvest project entitled “We Feel Fine.” Started in 2005 by Sep Kamvar and Jonathan Harris the project crawls blogs and twitter for the phrase “I feel” or “I’m feeling” and captures the results in an extensive database. This past November “We Feel Fine” was released in book form – highly recommend checking it out. Due to word count limitations found in print magazines I wanted to share the interview with Kamvar. The following is the extended version.
WBP: “We Feel Fine” is an amazing project that puts a face on various data. What types of things surprised you most about the project? In the sense that you were able to peer into a sea of what some would consider noise and in the process you created a project that has a strong emotional core.
Kamvar: One of the things that surprised us the most is when you strip away thoughts and opinions and focus on emotions, people are much more similar than they are different. The top 10 emotions are the same for men as for women, for people in London and in Bangcock, for blacks and whites. So this project for us has been about self-exploration as much as it has been about voyeurism.
That being said, there are also some real emotional differences between people. As people grow older, they tend to get happier, and further, they define happiness differently. Younger people tend to associate happiness with excitement, while older people tend to associate happiness with calm. And women express their emotions far more often than men, and have a far more nuanced vocabulary than men to describe their emotions.
WBP: When I look at “We Feel Fine” I see the potential for collaborative storytelling that makes use of various data. Have you ever considered this? And if so how do you think you’d approach it?
Kamvar: Absolutely. We Feel Fine is a story authored by millions of people who don’t know each other. The result is a coherent, authentic story. And this is not the only story that can be told this way — the story of love, the story of hurt, the story of helplessness. There are thousands of stories waiting to be told collaboratively by millions of people who don’t know each other.
When we talk about this kind of scale, the most appropriate thing to do to tell these stories is to build tools — tools that allow individuals to tell their personal stories in a meaningful way, and tools that collect, curate, recombine, and edit these stories to form the stories of the collective.
WBP: Can you also talk to the concept of data and it’s value to not only helping to discover but to also aid emotional and social connections?
Kamvar:: Most data analysis has focused on the macro level — statistics, trends, clusters, etc. These give important contextual information and meaningful insights, but rarely do they provoke a visceral, emotional reaction. On the other hand, many individual stories provoke an emotional reaction or social connection, but lack the context that data analysis brings.
For us, it’s important not only to present the high-level data analysis, but also to present the individual stories behind the statistics, and allow for the user to seamlessly shift between the two.
WBP: Do you have any opinions around DataPortability? The open accessibility to blog posts and comments makes a project like “We Feel Fine” possible. Do you have any opinions around DataPortability and the role that open data could play in the emergence of the real-time web especially related to new forms of art and storytelling?
Kamvar:: We are big believers in Data Portability. We Feel Fine would not have been possible without the phenomenon of blogging, and we have made an open API into We Feel Fine that allow people to make artwork and do data analysis with the We Feel Fine data. People have made beautiful work with the API that we would never have thought of ourselves.
In visualizing the data around “We Feel Fine” where there any considerations in terms of the way you shaped the project? Meaning did you discover and modify the project as it has progressed and if so how?
We agonized over every detail in both the website and the book. For example, the opening movement of the website, which we call “Madness”, is meant to convey the feeling of living in a large, anonymous city, like New York, where every day, we see hundreds of people who we will never see again, just for an instant. The overall energy is exciting and beautiful, but if one person were to be removed our substituted, it wouldn’t make a difference to the landscape. The swarming colored dots are meant to reflect that energy.
However, when you develop a relationship with one of the people in the city, that person becomes important, individual, and irreplaceable. The analogy here is clicking on one of the dots on the Madness movement and seeing the emotion of the person behind it.
Another element that is central to both the book and the website are what we call Montages. When there is a photo in the same blog post as a feeling sentence, our program automatically crops the photo and overlays the feeling sentence onto the photo. The resulting composition is often moving, often funny, often a nicely told sentence about ordinary emotion. We pay as much attention to what we leave out as what we leave in. By cropping the photo and not including context to the feeling, we allow space for the viewer. The viewer can fill in that space with memory or imagination, both of which are powerful allies.
WBP: Any thoughts on the future of the real-time web and where you’d like to see it go especially in relation to art, storytelling and / or discovery?
Kamvar: One thing I’d like to see is more depth in the real-time web. People’s behavior reflects the tools that they have available to them. In places where there are more McDonald’s, people get fatter. On the web, as tools make it easy to communicate via status messages, that communication has less depth. I’d like to see more web tools that are designed for deeper communication.
On the search side, I’d like to see a broader diversity of paradigm. A list of 10 ordered results work well for navigational and informational queries, but are not as good for learning more about people or communities.
WBP: What projects or technology excites you and do you have any predications towards the way people will discover stories, content and each other?
Kamvar:: I’m excited about a lot of things. One is the trend towards open source and open data. With mobile phones, there is a very real possibility that the dominant operating system will be an open source operating system (Android). Given how important mobile computing has become (and will continue to become), this will lead to more opportunities for developers and far better products for users.
I’m also excited about how little technology entrepreneurship costs. Technology that used to cost half a million dollars to develop now costs $15,000. This will lead to more unlikely entrepreneurs, more risk-taking, and more potential for highly impactful technologies.
And finally, I’m excited about the cultural shift that has led people to be comfortable with posting lots of information online. That availability of information is useful not just for storytelling, but across all the sciences. 10 years ago, a book like We Feel Fine could not be imagined. As more information flows to the web, it will be used as a database for many other things that are unimaginable today.
Posted in design