A few weeks back I was having dinner with a friend and his family. While we were chatting at the dinner table, his 9-year-old son came up with a demand to download a new game. His logic was simple – “the game is free (So dad you should not have any objections).”  My friend offered a sage advice – If the app is free, then perhaps you are the product! While the kid was not much convinced with that sentence, it made me wonder how humans are progressively transforming from being the beneficiary of technology to becoming a target (or object) of technology. Has the era of “Internet of humans” arrived? Let us start by exploring how user data is collected and used commercially.

User data profilers

Internet carried 1% of information flowing through two-way telecommunications networks in the year 1993, which rose to 97% by 2007. Lou Montulli introduced cookies in Netscape browser as early as 1994, and even today they are used for collecting user information – although with explicit approval by users after federal trade commission’s intervention. This was the first significant wave.

The second wave of user data capture came with mobile apps ignited by the release of first smartphone and Appstore by Apple.  Over the years, most apps have been marketed as free apps. These apps ask for blanket approval to access resources on the phone. Lured by free apps, the unsuspecting user gladly provides permission to access resources. The apps silently capture user data which is used for Ad revenue.

The third major wave came with social media platforms like Facebook, Twitter, LinkedIn and numerous others. These sites gather enormous information about the user based on user browsing habits. Similarly, online retailers started profiling users based on products searched and purchased. These sites have enough information to profile users, forecast their need and run successful targeted campaigns.

The fourth wave came with Bots, wearables and personal assistants. Whether you consider Siri, Cortana, Alexa or a number of other alternatives, the Interactive voice bots capture large amount of user data based on user interaction. One could argue, these Bots are silently listening the ambient noise, detecting patterns and profiling the household. Similarly, wearables capture myriad set of parameters like activity levels, sleep patterns, heart rate, blood pressure location information etc.

The Missing Link

While massive data sets of user information are captured by cookies, Apps, social media, retailing sites and Bots/wearables, the information is still in a silo and resides with an individual player. Even if every player is willing to share information based on a commercial arrangement, the data cannot be consumed for any meaningful interpretation. The reason for this is world wide web organizes information as a structure of data and not the meaning of the data. So while the data is captured, the semantic link is missing for realizing “Internet of humans.”

“Semantic web” – a term coined by Tim Berners-Lee in his paper in 2001, is an extension of world wide web and has now been standardized in W3C. The model extends the network of hyperlinked human-readable web pages by inserting machine-readable metadata about pages and how they are related to each other.  Resource Data Framework (RDF) provides the foundation for publishing and linking data. Web Ontology Language (OWL)  and  SKOS (Simple Organization System Namespace) are used to organize the data and provide meaning to the data.  Once data is arranged in this fashion across enterprises and domains, one should be able to query this data – a query language called SPARQL does that job. And finally, you want to draw inference data based on rules which are supported by RIF (Rule Interchange format).

Equipped with this, you can potentially consume data from any structures or unstructured data source, create RDF, ontologies and RIF’s and then search for an answer to any random question. For example, “Find a target who has liked someone’s picture of Miami Beach (on Facebook), has recently searched flights to Miami (from Cookies), asked for the weather in Miami on Alexa and may have looked for surfboards on Amazon.  When the meaning of data is linked using semantic technology instead of web pages, any ad-hoc information can be queried in using graph theory rules.

What applies to web-scale also applies to enterprise scale. The enterprise would want to use this semantic technology to organize information in data lake as ontologies so that they can query and infer along any dimension unlike fixed dimensions in today’s data warehousing systems.  One product worth mentioning here is Anzo, A data lake platform by Cambridge Semantics.


It’s a scary thought that as a technology consumer that our every action creates a data point, which gets recorded, analyzed, organized and inferred for potential exploitation. It is very likely that world will move to the Semantic web and with every piece of data interpreted by machines, predictive buyer behavior models would evolve as never before. At the same time, privacy laws will like to become stricter to protect the exploitation of the buyer.

Fundamentally humans are more complicated than things. So unlike sensors, humans produce multi-dimensional information which will need massive infrastructure to build ontologies to cover every relation.  Only time will tell if “Internet of humans” becomes a reality, bit if it does Semantic technologies would most likely play a central role.