“I have a problem I’m sure many other bloggers face; I am perfectly comfortable sharing intimate details about my emotions with complete strangers I meet online but shy away from expressing my true feelings to anyone I know in real life.”
– a woman in Maine
We Feel Fine is a data mining project that collates human feelings from the World Wide Web. Described by its creators as “an exploration of human emotion on a global scale”, the site provides its users with an insight into the human psyche that is so far unseen in the digital world.In short, a data collection engine conducts searches of the web’s latest blogs for incidences of the phrases “I feel” and “I am feeling”. It does this approximately every ten minutes and when it identifies a phrase, the entire sentence is ‘harvested’. The “feelings” can be explored, sorted and analysed by the viewer using the We Feel Fine applet, accessed via clicking on the “Open We Feel Fine” button clearly and conveniently located to the right-hand-side of the home page, seen above.
The engine is able to add up to 20,000 new feelings every day, and as each sentence is stored the system is able to detect the particular “feelings” they express. Any photos that accompany the blog posts are also saved and can be viewed by the user. The intelligence of the system is so developed that it can extract the age, gender and location of the author. It can even identify local weather conditions.
Once the applet is opened, the user is invited to explore “six movements”. These are essentially six different interfaces that allow visitors of the site to uncover stored “feelings” in various ways. The user can quite simply navigate themselves between movements by scrolling over the heart in the bottom left-hand corner of the applet at which point a simple list of the interfaces appears. As well as this, the “feelings” are made easy to distinguish with use of different colours for each.
Madness – mass of particles bouncing around, each representing a single feeling. Designed to mimic the madness of the human world, particles can be clicked at any time to reveal the sentence or photograph inside.
Murmurs – sentences appear letter-by-letter like they are being typed by the author and fade to black as new sentences appear.
Montage - what do feelings look like? By clicking on the thumbnails the user can view photographs from blogs with "I feel" phrases.
Mobs – shows the most common feelings of the sample population.
Metrics – shows the most salient or representative traits of the sample population.
Mounds – measures the frequency of feelings found in the database (the bigger the mound, the higher the frequency)
At the crux of this project is the ability to reveal fascinating information and statistics about how people feel from certain countries, cities, age groups on particular days of the year. This is enabled by The Panel located at the top of the applet.
The Panel - this is how the user controls the sample population that is shown on the screen.
When clicked, drop-down type menus appear and the user can constrain the population shown by the following axes:
Hit the “Find Feelings” button in the bottom right corner and you’re on your way!
By utilizing these axes, the project can answer questions such as:
- Do women feel fat more often than men?
- Does rainy weather affect how we feel?
- Which are the happiest cities in the world?
Aside from the trivial, the applet acts as a useful method of recording the human experience. It is, fundamentally, a historical collection unlike any other. Imagine being able to analyse how people in West Germany felt when the Berlin Wall fell? Or how people in New York felt on the day of the 9/11 attacks? The project has only been operating since 2005 but, no doubt, will be a compelling source of information for the momentous historical events of the future.
Although the project is primarily online-based, Harris & Kamvar have also transferred their ‘favourites’ from the collection into a book that is sold on Amazon. The data is, effectively, a story of what it’s like to be human so I can understand how the concept could work as a flip-through, coffee table book.
Created prior to the rise of the iPhone, iPad and the mobile ‘app’ platform, I would be interested to see if project can be transferred to into an application format. With the use of satellite technology, one could receive information about how people in certain areas are feeling. If I am house hunting in an area where people predominantly feel violent, I think I would like to know! Accessibility to data collection along these lines would be extremely useful.
We Feel Fine also has a Facebook and Twitter account, both of which are currently inactive. I believe this is mostly because these were used to promote the launch of the book, not so much the applet itself as there are no links to these pages from the project site. I can envisage, however, that Twitter followers would engage or appreciate quirky statues/tweets about “what most people are feeling today”, as would those on Facebook. Updating the project site with links to such pages, and maintaining statuses/tweets would regenerate visitation to the site and perhaps increase book sales.
Undoubtedly, the web has expanded dramatically over the past decade; there is significantly more content available and the nature of that content is becoming increasingly diverse. It is a larger place than it was in the 90s, it is more immediate and it is more social. There are people on Facebook, people on live messaging, bloggers, professional writers, buyers, sellers, groups, pages, and videos. The list is endless. The web is endless. And this makes it a relatively untouched resource of human interaction. We Feel Fine attempts to capture this and does so in a very successful manner.
Traditionally, data mining and information retrieval has been primarily focused on queries and documents. Alternatively, Jonathan Harris and computer engineer Sep Kamvar have sought to create a program that focuses on the people behind the queries and documents. By merging elements of computer science, anthropology, visual art and storytelling, they have produced a social experiment that entirely re-imagines how humans relate to technology and also to each other.
Inspired by techniques used by Ben Rubin and Mark Hansen in their installation project Listening Post, We Feel Fine scans blog posts from a variety of online sources, such as LiveJournal, MySpace and Google amongst other social media outlets. It is because of this that, in essence, We Feel Fine is an artwork made by everyone, for everyone.
According to Harris, We Feel Fine “will grow and change as we grow and change, reflecting what’s on our blogs, what’s in our hearts, what’s in our minds.” As the user views and interacts with the “feelings”, they undergo their own experience of reflection on the world around them, and, consequently, are provoked to reflect on themselves. The project, it seems, is just as much about the user as it is about the data collected. As they explore the site’s interfaces, the user is reassured of the shared human experience; no matter what they feel, it’s likely that there are a number of people around the world that feel the same.
“We hope it makes the world seem a little smaller, and we hope it helps people see beauty in the everyday ups and downs of life.”
– Jonathan Harris