Context and intent in semantic search

There’s a quiet revolution taking place that is affecting everything we do and hardly anybody has noticed it. Semantic technologies encode meaning separate from data and content files and separately from application code. They enable a meaningful moment of action to be created at what developers call execution time (a.k.a. the run time of the program). 

Across all our devices a new generation of apps analyze what we do, map the patterns of our past actions, create predictive profiles of ourselves to help us do things we usually do faster and work to broaden our horizons by helping us discover new things that might help us. 

In the semantic technology landscape these initiatives go by names such as user profile mapping, predictive analysis, engineered serendipity and semantic search. Depending on the computing environment you are in, the app you use and what you’re trying to do any of these may contain partial elements of all the others. In truth they are all part of a family with a common lineage that springs from augmented intelligence programming and machine learning. 

Those of you who have been following my semantic search posts understand that essentially what this means is that end-user usage is modelled, natural language is processed and understood, context and intent play a pivotal role in shaping the app’s output. 

This general outline holds true whether you are on Google carrying out a search, using Google Now to see what it suggest you need: 

Google Now Predictive search


It works, when you are surfing YouTube which scans your viewing history and suggest videos you may want to see (which are preloaded to your profile for a seamless viewing experience), keystrokes that you save as you use your smartphone: 


semantic technology makes apps and devices appear to be intelligent


It Gets Deeper

Take one of my favorite apps for instance: Evernote. This is one of three apps I use to help make my Google+ experience manageable

Evernote has used augmented intelligence and deep learning to add context to its every page:

What this means in more practical terms for those of us not working as journalists on The Wall Street Journal is that articles saved in Evernote are scanned, machine read and understood and Evernote can then bring up suggested articles from across the web that add additional layers of information to the subject that has been saved. 


Evernote's Context is a Classic example of semantic technology

So What?

The keywords you need to focus on here are context and intent. Depending on which side of the marketing divide you happen to find yourself on you can use both to actually make some real gains. For those of us who carry out research, the fact that the apps on our devices make them smarter allows us to leverage our usage behavior to get more done with less effort (case in point the Evernote Context feature). 

But there’s another, greater benefit here that serves all those who are looking for an audience. Context and intent allow content to surface on occasions and at times when in the past it would have been unimaginable. For a bookmarked item of information, for example, to become the reason for similar content to appear that adds extra layers of information to it, is something that in the past would have sounded like science fiction. 

This becomes the signal for those who market products or services to create the kind of content that is of sufficient depth and relative value to their audience that it actually becomes capable of surfacing across the web in areas beyond the website that contains it. This means it requires a certain degree of engagement and reshares and the website that produces the content has to work towards becoming authoritative and trustworthy. 

There are two distinct networks involved here: The one of devices and apps where information crosses over as it is pulled in from authoritative sources or popular websites. And the one of social networks where people pass information they have found useful or interesting around. Context is where these two intersect. Intent is what defines the value of the information that surfaces at the intersection. Quality is what drives everything. 

If that sounds like a long-term, marketing formula that can deliver a win, it’s because it is. 


Evernote Context Feature
Designing serendipity in semantic search
Engineering Serendipity in a Semantic Search Space