Semantic Search

If you were to walk into a café and say you were thirsty, the staff would understand that you were asking for coffee, tea, water, or another beverage. They wouldn’t dial 999 and think you were suffering from dehydration. In essence, they would intuitively understand the ‘context’ of your request.

Semantic search is the web’s attempt to be as intuitive as the café staff when it comes to searches; to understand the relationships between content as we humans do intuitively, and produce more relevant results.

The primary obstacle for semantic search is that most web pages are predominantly written in HTML, which structures rather than understands page content. But things are slowly changing.

Companies that are recognising the potential of semantic technology are beginning to invest time, money and resources in it. Brands that are already using semantic search successfully include IBM, Facebook, and Google. For example, IBM’s Sentiment Analysis uses natural language processing to differentiate between sarcasm and sincerity in social platforms in order to define and evaluate social trends. The potential this has for marketing is huge.

THE FACEBOOK OPEN GRAPH AGGREGATES MEMBER’S ACTIVITIES (SUCH AS COMMENTS AND LIKES) TO OFFER TARGETED ADVERTISING SPECIFIC TO THAT PERSON. IT’S THEREFORE MAKING INTUITIVE GUESSES ON WHAT WE MIGHT WANT AS CONSUMERS.

With such a mercurial subject, it’s hard to pinpoint exactly how semantic content might evolve, but we can be sure that data on the web will become more meaningful and useful. Not just for the ‘Average Joe’ web user, but also for us online marketeers looking to build a more intuitive and intelligent relationship with their consumers.

As Charles Silver says, semantic technology is introducing a new phase of the web where we stop searching and start finding. The challenge for brands will be to make sure that when customers go online, they are the first to be found. We can’t wait to partner brands through this process.

Original post here: http://nxt-book.mzl.com/semantic-search/