Evolution of search: Relational navigation

Jack Berkowitz: Hi. My name is Jack Berkowitz, I'm the VP of Engineering for Siderean Software. Today I'm going to be talking about the evolution of search. In particular, I'm going to be talking about relation with navigation.

Today businesses have a big challenge, there's an explosion of information. They're trying to provide their information to their end users. They're also trying to combine it with user generated content or access feeds, and new data sources coming online. The biggest challenge for them right now is maintaining context. Context for how their information plays to end users, and the meaning of that information.

So, over the past 10 years people have been using search to try to find information. Let's give an example, suppose somebody was trying to hire a new employee. First thing they might do is sit down at a search terminal, type in the employee's name. This would be the key word that would help them find a series of documents about that person. Now they're going to get a collection of documents that they can take a look at, and over time they might end up refining by using additional keywords.

The important aspect here is, that to find context the person's going to have to read every one of these documents, understand the inner relationships. Probably over time, they're going to end up going back to the beginning again to try to refine the search, find the right person, find the relationships that person has.

Essentially, over time we see that search's unstructured finding of information effective, but it takes some tie. Now over the past three years the reaction to the time it takes for search, people have been moving to faster than navigation, some people call it, "Guided."

The essential ingredient about this is providing a structure up front. This structure allows people to organize information, and follow that information through a structured taxonomy. Now, this works great except for the cases where for example, the job you're trying to place didn't exist before. You don't have a taxonomy node that you can locate that person's information about, or for example if the end user having moved down a certain part of taxonomy wants to leap across the taxonomy to find different types of information about that person.

So, again, the biggest issue with facet navigation today is it's rigid. In reaction to that rigidness people are now moving to relational navigation. Relational navigation allows people to explore the concepts, the concepts between and across information, as well as the relationships.

So, let's go back to our example. It's important to find out who that person is. But it's also very interesting to find out where that person's worked, projects that person's worked on, and people that they've worked with. More importantly to make that hiring decision what you want to know is, what is the combination of that information to provide the end context?

Relational navigation inherently is web centric. People use it exactly the same way they would use the web today. It's flexible, and it really does provide that end user context that people are finding so important in businesses today.

So, if we look at the evolution of search, and we started with search which we've seen as inherently unstructured but accurate people reacted to it through a facet of navigation. But, it had a problem of being too rigid.

Today, we're looking at relational navigation. Really the advantage here is the benefits of being flexible, and providing that context people need to find accurate information.