I am super excited to announce that my paper-“Applying Chatbots to the Internet of Things: Opportunities and Architectural Elements” has been published in the December ’16 edition of the International Journal of Advanced Computer Science and Applications.

You can read the complete paper here: http://dx.doi.org/10.14569/IJACA.2016.071119

Or at its ArXiv location: https://arxiv.org/abs/1611.03799

But today I wanted to give you an idea of why this paper is important and the reasons behind writing it.

First, we must understand the context. Chatbots are definitely not a new idea, they’ve been around for decades, since the days of the ELIZA bot built at MIT in 1966 by Joseph Weizenbaum. In that time, AI has advanced and so has the capabilities of these bots. They have started acquiring features that could be associated with Intelligent Agents. Become more and more autonomous. Better at planning, reasoning, and learning. Turning into the more natural conversational interfaces you see today. In fact, you may already be using it on a day to day basis.

Like this one:

Ok, maybe not Monkey pets. But you get the idea.

So as more new use cases were being developed, the bot community of developers, as well as users, started to grow at a dizzying pace. Right now, all of the Big 4 Tech companies- Amazon, Google, Microsoft, and Apple have either launched their own bot frameworks or acquired them. This shows you how serious they are about this space.

But let me give you an idea of how fast and recent the development has been: 6 Months ago, None of these Big 4 companies had frameworks for Chatbots.

So while we have known how Natural Language Processing, Machine Learning, etc. work. The opportunity lies in where we implement it.

Internet of Things is also growing rapidly, from the 6.4 Billion (approx) connected things this year to an estimated 30 billion by 2020. So it’s been a relatively recent phenomenon. (There are a lot of reasons for this, a number of which involve the economics of making chips. You can get a better understanding of it by reading my earlier blog post Here.)

So far, Chatbots were being focused on the digital space: for conversational commerce, news, weather, etc. which was in accordance with the typical use cases of chatbots-Conversing and retrieving information.

Again, the research and knowledge behind this from Information Retrieval to Information Extraction has been well known and well documented.

However, there was an evident gap between the research on Chatbots and research on Internet of Things. There was no mention of the combination of the two concepts together. Not in the books, papers, nowhere.

As a developer myself, I wanted to understand the what it meant to connect Conversational Interfaces with the physical space. I was only able to find DIY projects and simple code snippets of how these two ideas could be connected.

The paper:

• Presents a modern Chatbot architecture and its relation to the IoT and Web Of Things architectures.
• Identifies specific use cases in Chatbots for IoT. Particularly in Home Automation.
• Summarizes and captures the modern Chatbot landscape and performs a background study on WoT.
• Attempts to solve important and prevalent challenges in IoT through Chatbots.
• Identifies future directions in Wisdom of Things, Semantic Web etc.

I cast a wide net of future directions from using Human in the Loop data to identifying Cyber-Physical Systems as a key research area for this technology. This was the goal from the start.

Where do we go from here? This experience of doing my own research, working on something that hasn’t been done before has been challenging and fulfilling. I will continue to push forward new ideas in the coming days as I now go in-depth into the field of AI.