When we interact with information sources online in 2014 (note I’m not calling these things search engines, though the things we call search engines right now will be part of this segment) we will be experiencing something quite different from the web search engines of today. Gone will be the game of attempting to get the system to provide you with the information you need via a ping pong of semi-literate fragments of text.
When we interact with information sources online in 2014 (note I’m not calling these things search engines, though the things we call search engines right now will be part of this segment) we will be experiencing something quite different from the web search engines of today. Gone will be the game of attempting to get the system to provide you with the information you need via a ping pong of semi-literate fragments of text. Gone will be the mundane presentation of 10 places you might visit which might help you with whatever you are trying to do (the system has no real idea). Instead, we will be working with agents that blend a deep(er) understanding of the real world with an appreciation for how those real world entities manifest in documents (be they web pages, books or your personal content and private communications).
Fueling this change is the move from textual systems, to blended systems which marry structured data (sets of facts encoded in a language which is intended to at least be able to establish the fundamental relationships of identity and so on) with documents and other forms of web presence.
The winners (or at least the leaders) in this space will be those companies which can master three competencies.
- Understanding the Web: being able to interact with and reason about the objects that make up the web. Object like URLs, HTML documents, dynamically rendered content, javascript, redirection, sites, domain parking, spam, etc. We think this is what traditional web search engines have already achieved but they still have plenty of headroom for improvement.
- Understanding the World: being able to represent and reason about things like a song having an artist, a license to be performed in a movie, that a hairdresser doesn’t have private dining reservations and that taco trucks can’t be in two locations at the same time. This part of the equation has been the domain of local search, eCommerce and other catalogue oriented systems.
- Understanding Web Presence: this is the tricky part. How do entities appear (i.e. represent themselves, are mentioned by others and so on) on the web (or in other content available online?
There are many companies and products that are at the fringes of this transformation. Facebook is fundamentally about the relationship between entities and content; Ebay is very much about entities for which they mediate transactions; Google and Bing started in the document space but more and more include experiences which rely on the entity concept.
In recent months, the term ‘data scientist’ has been the buzzword associated with the most attractive career trajectories. I suspect that we will be inventing new terms like ‘representation science’ and ‘reasoning science’ in the not too distant future (the industry, of course, won’t entertain the job title philosopher for some time).