What if the internet got smarter?

Every self-respecting tech guy was there: the annual Semantics conference in Vienna. If you don’t have a degree in engineering or information technology you won’t get the most out of this conference, but you’ll profit nonetheless. Because semantics is not just about a bunch of technical stuff, it is first and foremost about the future.

What if open data changed the world?

Jeanne Holm, Chief Knowledge Architect at NASA, started the conference off with her talk about the importance of linked and open data. The open data movement, she claimed, could actually help to make the most out of a lot of lives, not only in the western civilizations, but especially in third world countries. Africa, for example, has an 80% penetration rate of mobile technology. As Holm told the amazing story of an African farmer, and how she managed to integrate open data into her farming business without losing her traditional lifestyle, she gave a face to all of the beneficiaries of the open data movement and a feeling for the many possibilities that linked data holds. “Releasing and using open data is about empowering people to make better decisions”, Holm said. The open data movement wants to tackle three major challenges in the next years: climate change, the rescue of ecosystems and sustainable development. “Linked data is at the heart of how we are going to evolve as humanity over the next few years”, Holm concluded. Listen to her talk Open data changes lives at TEDxUCLA.

What if our digital and physical realities merged?

Sam Rehman, Chief Technology Officer at Epam Systems, talked about the Internet of Things (IoT) and in a few words sketched a future were the digital and the real world will merge into one holistic experience. The Internet of Things can be summarized by the following characteristics. It is

  • connected
  • pervasive
  • active or passive, and
  • beyond screens

Imagine a world where there is no such thing as online and offline anymore, a world were you walk into an area and your device responds to thousands of sensors and suddenly you stand in a digital world. “Everything is connected, and most of the time you won’t even know.”, Rehman said. But he also didn’t leave out the challenges this new world brings with it: data privacy will become an even bigger issue because traditional opt-in and opt-out solutions just won’t work at this scale, and with large sets of data behavioral analysis (profiling) will become easier and more reliable. Another main concern of the IoT is what Rehman calls “device weaponization”: image a coffee maker responding to voice recognition and someone turning on the microphone all the time effortlessly listening in on all your private conversations… you get the idea.

What if search engines were smarter?

Peter Mika, Director of Semantic Search at Yahoo, talked about the challenges programmers face when it comes to making search engines, or machines in general, act smarter. The times when ad-hoc document retrieval was enough are over, Mika explained. With ad-hoc document retrieval the search engine decides if a document is relevant according to its textual similarity and authority. But this means that search engines are not able to manage queries outside of that paradigm and aren’t able to answer complex needs. “Search needs more intelligence”, Mika said. The solution to this problem lies in the optimization of a thing called the “Semantic Web”: “a social-technical ecosystem where the meaning of content is explicit and shared among agents (humans and machines).” Put in simple terms, this means that machines are not ordinarily able to understand texts, because they only respond to symbols. In the Semantic Web the internet should be readable and processable for both, humans and machines. Google´s Knowledge Graph, for example, is based on the Semantic Web. At the end of his talk Mika pointed out the benefits of semantic search with a simple example: If you ask Siri today if she wants to be your girlfriend she will tell you a bunch of evasive answers. Wouldn’t it be cool if she knew more about us, so that she could give an answer with reference to a certain context like: “You already have a girlfriend.” This example raised, of course, the inevitable question about data privacy. “There is a trade-off”, Mika recognized. “If you want more intelligent search engines, you have to trust the search engines with your information.

(See below some of the answers Siri gives you when you ask her to go steady.)

Siri1     Siri2          Siri4

What if we linked scientific research?

The last keynote of the event was held by Klaus Tochtermann, Director of the Leibniz Information Centre for Economics, who talked about semantics in scientific libraries, and introduced a virtual library for economics called “EconBiz“. With this project he wants to use semantics to cross-link sciences through a linked data cloud to better express the relationship between two publications. One of the main problems, Tochtermann said, were the limited rights for the texts and for data mining. Asked by the audience if “EconBiz” was meant as a competition to Google or other search engines in the scientific field, Tochtermann said that, on the contrary, “EconBiz” depends on the big search engines to get found by people interested in scientific research.

Photo credits: Verena Ehrnberger

Share this post

Leave a comment

Your email address will not be published. Required fields are marked *