Making Information Useful

When trying to narrow down the most relevant information from an otherwise comprehensible list of options, context matters. The Free On-Line Dictionary of Computing defines context as “that which surrounds, and gives meaning to something else.”  Essentially, this means that our surroundings have an impact on what matters and what is meaningless.  What is important in one context, may not be important in another.

This point is made most clearly through examples. Let’s start with location. A user’s current location can have a huge influence on what information is most relevant. Nokia’s “Location, Context, and Mobile Services” paper states that location is one of the most important pieces of contextual information. This paper says that so many things can be inferred contextually from a user’s location, including the weather they are experiencing, the traffic they are in, the language they are speaking, the price they are paying, the nearest cafe, where they were two minutes ago and what direction they are headed, and if they happen to be at a known point of interest (theater, church, etc.) then their current activity may also be known. Using this contextual data, information can be narrowed down to that which suits the users current needs most directly. For example, if the user is known to be in a grocery store, a contextually aware search for apples might return information on what fruit is in season this time of year, price comparisons, and recipes containing apples. Information about the history of apples, the Big Apple, and Johnny Appleseed could be easily filtered out. The key is that this filtering takes place without the user having to think about it. This is what makes contextual information filtering so powerful and useful.

The more contextual information the better. Knowing that a user is looking at travel information for business purposes vs. personal travel purposes will result in very different relevant information. A business traveler may want to know the weather, local business customs, good business hotels, or a place to take clients out to eat. A personal traveler may want to know sightseeing activities, the best resorts, and where to play a round of golf. Knowing that the user is sitting on a plane or walking down a street may also be relevant. The latter will need information that is much more concise and to the point. A user who is looking for information about “queen” while listening to classic rock is probably looking for something significantly different than the person walking around Buckingham Palace.

The examples are both endless and obvious. It is so obvious because people do this naturally. When asking for directions to the nearest Burger King, you don’t need to specify what town you are in – the other person already knows. This allows humans to communicate, and provide relevant information, much more effectively than computers. For this reason, getting information from other people is generally the fastest way to get what you are looking for. This makes the power of ones social network a key factor, but that is not the topic of this post. When the information you are looking for can not be obtained through other humans in a timely manner, people generally turn to electronic devices such as computers and smartphones. This is where the power of context is largely untapped. Despite the efforts of many to bring contextual filtering to the fingertips of the masses, commercial products and services with this capability are still in their infancy.

Location services are the most widely used applications with contextual data. In these applications, a user’s position is used to provide relevant contextual information. The most common services today provide information on nearby businesses of nearby friends.  At the time of this post, over 1000 iPhone applications on the Apple App Store use location information to provide context.

[more than location context is important - add other types of contextual information]

[need to add survey of new research and products using context]