Five Things I've Learned About Context
When I was first notified that I was going to be interning at Argus Labs, I jumped on the web and started reading everything I could about their mission as well as their market because both were somewhat unfamiliar to me. I knew Argus Labs dealt with context and at first thought, it seemed simple enough. It’s only context, right? Plus, as a student studying literature, I thought I had context in the bag, so to speak. As soon as I started searching, I realized I was dead wrong. Let’s just say it was one of those moments where the more research I did, the more foreign it seemed to me. I came to the office on a Monday knowing little about what they were working on, but when you live in a foreign country at a new job, you have no choice but to learn quickly. A month later, there’s plenty I’ve learned and I’ll share my top five insights as my time here comes to an end.
1. Context is (relatively) simple.
First and foremost, you’re probably wondering what in the world context actually is. I’ll explain it through an example. Imagine you’re in a foreign country, you don’t speak the language and you have to find your way to your internship (sounds a lot like what I had to do). You have no map and are aimlessly walking around the city. You’re lost. You ask for some directions and you’re told to take a left at some big landmark. Remember: you’re not from here so to you, the landmark is completely out of context. You don’t know where you’re supposed to go. You need to understand the city, the context of the situation you’re in to make sense of the direction. In other words, context provides a foundation for information to sit on in order to make it meaningful. Now, let’s apply that to computing. Contextualization is the process of taking personal data – interests, habits, moods, and environments – and layering it on top of itself to create rich knowledge about a specific person. I’ll give some examples in a minute, but for now, think of it like this: context is all the background information necessary to understand where to go in a big city. In computing, contextualization provides the reason why a person does certain things at certain times. It’s great for individuals to understand their daily routine with an acute level of specificity as well as for companies to push the right content at the right time to the users who need it.
2. ‘Big Data’ calls for more than ‘big analytics’
You may have heard by now that humans produce more data in one year than everything produced before the year 2003. In fact, IBM estimates we double our data production every 1.2 years. That’s a lot of data…and I mean A LOT. For the past decade or so, we’ve really relied on web analytics used to deduce a number of insights about raw data. The process is fairly intelligent at this point (except when this happens), but systems should be able to do more with the data available today. Traditional analytics give insights to users using their location, maybe their interests and music history (pulled from Facebook and Last.fm), but that’s about the extent of it. With the wealth of knowledge available (which is seemingly endless and entirely accessible), it’s a huge opportunity for companies and individuals to have the tools to capture insights about the data they generate. For that reason, analytics just aren’t enough anymore. Contextualization doesn’t simply provide information about an individual, but produces actionable knowledge. This knowledge is what separates analytics from contextualization and is extremely exciting as a new frontier.
3. The very cool future of contextualization
I said it once and I’ll say it again: contextualization has a very cool future. In fact, it is so compelling to me, I’ve started to code myself so I can integrate some of the platform into my own applications. So what do I envision being built with context-aware tools? A couple things, but I’ll start from the top.
A world where music follows around an individual almost seems like a perfect movie. Imagine you go on a first date with a girl and you get back into the car from dinner and a fitting, relaxing playlist comes on. Or you just got on the road for a big trip with some friends and some exciting music starts playing that gets everyone in the car singing along. Even a time where you go into a coffee shop to study and the playlist already playing automatically adapts to your preferences. It’s what we call a ‘living playlist’ and can adjust to the situation without any disruption and even play a specific group of songs for a special time or place. It knows about you and therefore, has the ability to know what you want to listen to. That’s definitely at the top of my list for awesome applications for this technology.
Another scenario I’ve thought about over and over is the potential for a travel app to learn about how I take vacations – where, when and with whom – and propose an itinerary that meets all my standards – activities I enjoy, how much I am willing to spend, how far a typically travel, for how long, etc. The app could even plan a trip at the same time as my favorite band’s concert so I could see my favorite band at my favorite place. There’s a really interesting market for applications like this.
The third scenario (and a personal favorite) is using contextualization platforms to build dating websites. Everyone always talks about how much they enjoy long walks on the beach, but how accurate is that? Using context-aware features, it’s possible to geo-locate and track if individuals actually spend significant time walking along the water. The idea is that dating services can tap into these data silos and see what individuals are doing, as opposed to asking them what they like to do. This allows insights to be richer and more intelligent, and in the case of dating, pairings to be more likely to succeed.
4. How is data collected?
All the personal data that is collected comes from one word: sensors. But where are they? To start out, you have all sorts of sensors built into your smart phone that measure different metrics: accelerometers; GPS; light and sound sensors; gyroscopes and more, which tap into your location, atmosphere, and environment. Additional sensors contribute to this wealth of data with physical measurements. Nike has a FuelBand, Garmin has watches that measure workouts, and there are even headsets made by Pip that can dictate when a user is attentive or not. There are also sensors that are built into houses, such as Nest’s home automation system, and cars, especially with computers becoming more and more common on the road. It’s amazing to see the expansion of personal sensors over the last half-decade or so and with the new sources of data, it’s possible for us to do our job better than ever. In fact, it is these sensors that enable contextualization to go beyond simple analytics.
5. Know a little bit of code
First thing I noticed when I came to the office is that everyone spoke Dutch. As a foreigner, I understood nothing, but even if they were speaking English, the engineers were talking at such a high technical level that it went over my head. Cue HTML and CSS knowledge.
I decided to enroll in some online coding courses for the past month and have actually made quite a bit of progress. I can now piece together the basics of web development, which has aided me in my understanding of the product. When things are extremely technical and difficult to understand, it’s always beneficial to know how to communicate with the people who are building the product directly, rather than through a third-party. Although it’s not specifically tied to contextualization, it’s a lesson I learned that will stick with me forever and something I can pull from in a number of situations.
To wrap this thought up, it never hurts to have an additional skill set because you never know when it will come in handy.