Can Apple Capture User Intent With Siri?

By Azam Khan 

It may be easy to pass Siri off as another voice recognition application but its domain stretches beyond AI algorithms, continual learning and contextually aware systems. The new Siri stands tall as Apple’s flagship feature to its iPhone 4S and for good reasons. More after the jump.

Siri’s attempt to mimic human conversation is being received well by users and spectators alike. Researchers for decades have longed to come close to creating a technology that surpasses the traditional computer generated voice responders. Most voice recognition systems have been around for ages while even the government was conducting research in parallel via DARPA. According to a discussion on Quora, Siri has been built upon 40 years of research funded by DARPA via SRI Internationa’s Artificial Intelligence Center.

Research from various universities including Carnegie Mellon, Univ. of Massachusetts, Univ. of Rochester, Stanford and others has come a long way from various fronts such as natural language understanding, machine learning, evidential and probabilistic reasoning, ontology and knowledge representation, planning, reasoning and service delegation. Siri comes from SRI International which was tasked by the Department of Defense in 1966 to develop “computer capabilities for intelligent behavior in complex situations”. Fast forward today, Siri touches on 3 important aspects of such technologies:

  • Conversational Interface
  • Personal Context Awareness
  • Service Delegation

Hands free communication with the device is the natural evolution in the way humans connect with their mobile devices. Although touch keyboards and even gestures, currently being explored alongside holographic display technology, will eventually come to dominate interactions, voice control is set to be the dominant ethnology in this next phase of device evolution. As envisioned by the couriers of the semantic web, Siri will eventually evolve to become a personal assistant that interprets language and requests and spits out the right information based on location, task, time and dialog context.

Search can be rather limiting on mobile devices with consumers reluctant to go on extensive information excursions on their limited screen size. Siri’s aim, to become the intelligent assistant using the aforementioned form factors like place and time, coupled with elegant design is just what’s needed in this next phase of device evolution. Visiting back to the 3 aspects of technology, Siri is strong on all three fronts:

1.  Conversational Interface: Siri’s interface makes it easy to say the right things and get the right answers. Whether its task completion or hunt for an answer, Siri is armed with language understanding components that are quite impressive

2. Personal Context Awareness: Siri understands conversational intent which is comprised of location, time, task and dialog contexts. Siri also knows who your friends are and additional information that is accessible to it

3. Service Delegation: Siri’s ability to combine vertical and horizontal searches across multiple information sources and then editing them based on a dynamic criteria allow it to get tasks done

There is little doubt Siri will be successful. From a user perspective, the technology will go beyond iPhone 4S to iPad 3 and in the home on Apple TV. From a corporate perspective, it could be an enormous revenue generator if Apple manages to make back-channel partnerships with vendors and earns commissions. This can be done if Siri can be the pathway for Apple to capture a user’s intent.