Knowledge is the key to solving complex problems. Knowledge, however, is not easy to come by. And while a little knowledge goes a long way, to solve really complex problems you need a lot of knowledge. Knowledge delivers its greatest value when it is integrated with other knowledge, especially when it achieves critical mass. The traditional techniques of knowledge acquisition do not scale very well. The process of acquiring knowledge and representing it in a form needed to be useful to automated systems to solve problems is a long and expensive road. In the early days of Expert Systems, this dilemma was referred to as the Knowledge Acquisition bottleneck. This issue has been largely responsible for the limited success of Expert Systems in particular, and until recently Artificial Intelligence in general. To be truly successful, and reach its true potential, AI must achieve Knowledge at Scale.

At the core of much of our work are several technologies that we believe can start to overcome the Knowledge Acquisition bottleneck, and begin to realize the vision of Knowledge at Scale:

• Machine Learning
• Crowdsourcing
• Semantic Web

Since the early days of computer science, one of the most promising approaches toward acquiring knowledge has been the ability that most distinguishes intelligent behavior in nature: learning. From Neural Networks which are inspired by the mechanisms of nature itself, to learning of rules and statistical models, Machine Learning has been one of the most ambitious and anticipated areas of AI research. More recently, Machine Learning has been at the heart of some of the most spectacular progress in recent years and at the center of operations at places like Google and Facebook. This has been thanks in large part to the steady advances in computational power, and the vast amount of data now available to these algorithms.

Intelligent Systems has its roots in Machine Learning research for the Defense and Intelligence industry, and it has been a key part of our work, and a key weapon in our toolkit, since its founding. We believe that recent success by Google and others is only a prelude to what is in store, and that, thanks in part to the vast amount of data flowing through the economy, as well as advances in the technology itself, we are at the cusp of a revolution in Machine Learning which will transform both business and society.

To learn more about our work in this area see Machine Learning

Humanity has experienced several revolutionary transitions that completely transformed how we live and produce. Among the most significant were the transition from hunter-gatherer to agricultural society and the Industrial Revolution. People often talk about today's world as being in a similar period, with information technology similarly ushering in a new world. This is often described by the label the Information Age.

For the most part, however, our acquisition of information and knowledge often more closely resembles the period of the pre-industrial world where knowledge is mostly hand crafted and done so in isolation by so-called knowledge workers. To be sure, raw data is processed automatically, but turning this data into knowledge and creating the tools which process this knowledge, whether it be websites or software programs, is generally performed by human experts in essentially a custom one-at-a time process, requiring a significant investment in time and money. The ubiquity of the Internet and means of easy access such as mobile devices, however, offers the opportunity to achieve for knowledge work the same sort of quantum leap in scale that the assembly line and factories did for physical production during the Industrial Revolution.

Intelligent Systems has developed tools and methods for harnessing the the power of the internet to acquire knowledge in a scalable way directly from thousands of users, and to do so with high quality. This offers the promise of bringing the lessons of the factory and assembly line to the knowledge economy, allowing the acquisition and use of high quality knowledge at an industrial scale.

To learn more about our work in this area see Crowdsourcing

The World Wide Web and the Internet itself, were a dramatic demonstration of what can happen when a critical set of technologies and standards unleash the combined efforts of millions of people and organizations to organically construct the world of information we now live in. Overnight. This is information at scale. We believe that the same principles can be harnessed to achieve knowledge at scale. What made this possible in the case of the Internet and World Wide Web, were a set of standards for representing, identifying, communicating, and integrating information (HTML, HTTP, TCP/IP, URL's, hyperlinks) that allowed millions of people and organizations to easily create their own page or website and connect these together via hyperlinks because everyone was talking the same language. Suddenly, each of these individual small contributions could be woven together into a tapestry of previously unimaginable size and scope. The multiplicative power and immense network effect, resulted in an a combinatorial explosion and exponential growth that changed the world in what seemed a blink of the eye.

Every day, countless individuals and organizations are generating knowledge to run their businesses, organize their lives, and build the websites described above. Like the documents that were generated in the days before the World Wide Web, however, much of this knowledge is isolated and represented in formats that are incompatible. The information on the web is integrated, but in a format designed for human consumption rather than interpretation and computation by software.

The Semantic Web is a vision, first articulated by father of the World Wide Web, Tim Berners-Lee, for applying the same principles as the Web itself to the representation of knowledge in a structured way that is designed for sharing and automated processing by machines. At its core is a standard for representing knowledge known as Resource Description Framework (RDF) which represents knowledge as simple subject-predicate-object statements which is easily interpreted by software and used in computation and inferences. As with the web itself, the key is that the knowledge is represented in a standard format allowing it to be integrated.

Of course, this has enormous benefits for individual organizations and their business partners, allowing widely different systems to be integrated and sharing information between organizations and companies. But the opportunity is even greater than this. In the same way that the Web standards allowed the contributions of millions to be combined to produce the Web we now know, the Semantic Web holds the promise of combining the world's knowledge in the same way. This is the promise of Knowledge at Scale.

To learn more about our work in this area see Semantic Web