By Staff Writer – Boonsri Dickinson (@boonspoon)

We waste so much time searching for what we are looking for. We already expect our searches to be conducted in real-time. Thanks Twitter for making us realize that. Still we expect more from search. We want the search engines to respond intelligently and cater to our needs. And we want them to be our personal translator.

The future of the Internet is coming soon. For one, researchers at Cornell University are designing search engine software that really understands you.

Computer scientists, Thorsten Joachims and Robert Kleinberg, want the search engine to learn by looking at the links people click on and see how they change their search behavior when it doesn’t pay off the first time around.

Good ranking function works for mega search giants like Google. But small collections don’t have that luxury, so they depend on machine learning to do the heavy lifting. The scientists are starting with a number of physics, math, and computer science papers to create this smart search engine for Cornell’s physics arXiv website.

With our chronic problem of information overload, there’s a need for high quality and focused search software. Words have different meanings. In the world of physics, the word “uncertainty” could suggest subatomic particles. But when an unemployed person searches for “uncertainty” for career insight, uncertainly takes on a completely different meaning.

In the future, search needs to become more niche, so people can use specialized search software to fit their specific industry needs. For example, IBM is working a project called n.Fluent that would “break language divides on the net.” Like magic, the tool would translate a foreign URL into your mother tongue, so it can be understood and read instantly. IBM also has an app that you can put on a website to translate the text to your desired language. And they plan on creating the same thing for mobile phones and instant messages.

Google is also interested in creating a translation tool and software that goes beyond translating websites. It would also be able to translate web searches. Soon, the whole world really could be yours. You’ll just have to search for it.

Image: flickr/ curiouslee (via OPTE)

Related articles by Zemanta

• Start-Ups Shouldn’t Be Afraid To ‘Fail Fast,’ ChaCha Says (blogs.wsj.com)
• How does meta-search engine compare the results. (startups.com)

Search Engines That Can Learn And Translate is a post from: TechStartups.com

Linking words together online is an hobby. Creating context from those words is an art. Making them work in harmony is a gift. The team at Thomson Reuters have done this for some time with their Calais project and the first consumer product from it OpenCalais.

OpenCalais is a monster database filled with linked data that can be accessed by anyone via a special API. Currently it works on an individual application level and as plugins with other popular CMS and blogging platforms. In tandem they create a more powerful experience for the publishers of new content by allowing the integration of linked data for linking, tagging and finding optional third-party resources for inspiration.

Word on the street and from a trusted source is that the Calais team has something new on the horizon – their own CMS. A release of a CMS that has semantic technologies built right into is a powerful addition to the marketplace. Many blogging platforms and CMS products available today have stagnated under the weight of their own popularity.

It is hard to innovate when you are leading the pack or trying to figure out how to make money from your product that is free. Time and resources become devoted to maintaining the status quo, patching holes and community relations and support. Not to mention that many of these tools have crept out from the minds of developers and not publishers.

The same can’t be said for Thomson Reuters Calais Initiative. With a business model built in from the inception of the project they have the ability to create value from the wealth of content that is entering their system. It is treated to some natural language processing, machine learning and prepared to link with other content. As the initiative currently exists it is the side car to better publishing as there is no status quo or hole patching. They realized this.

Throwing a CMS on top of this type of foundation is a logical move for OpenCalais to extend development and create publisher relations. There has been so little innovation in core CMS offerings that most rely on third-party developers to create additional functionality or implement API’s like OpenCalais, Zemanta and other discovery products.

An OpenCalais CMS, blogging platform or whatever comes from this group is going to be a welcome addition to the choices that publishers have available today. The closest opportunity today to creating a true semantic experience in a publishing platform that is readily available would be to install the Zemanta plugin in WordPress. It transforms the writing experience. An offering from Calais could transform the writing and CMS experience.

Update: Krista Thomas (@kristathomas) from OpenCalais got in touch to say that there is a fully integrated CMS based on Drupal called OpenPublish available now.

DISCLOSURE OF MATERIAL CONNECTION: http://cmp.ly/0

CMS Innovations from Thomson Reuters? is a post from: TechStartups.com