How Does Gnowsis Relate to NEPOMUK and Others?

We are close to releasing a Beta-Version of our software Cluug, about which you will read more in future blogposts. While we at Gnowsis are busy making this happen, there is time to compare Gnowsis with others and explain what we do differently and why.


To sum it up: Cluug is a semantic personal information management system developed by Gnowsis which runs as a cloud service. The ancestors and siblings differ in one or another point. This blog posts compares Cluug with the stuff done in the past.

Gnowsis as a name exists since 2003. My church minister Hannelore Rus helped finding a greek-sounding name for knowledge. A "w" was added to the original Greek "Gnosis" to refer to the English "knowlege" and to be distinctive in Google. You see - "Greek" is only one edit away from "Geek". The name Gnowsis was used for:

  • Gnowsis, a diploma thesis and software done by myself in 2003.
  •, an open source project with two major versions: 0.8.* was based on the diploma thesis and developed within the EPOS research project and 0.9.* developed from scratch within the NEPOMUK research project. This was done at DFKI from 2004 to 2007.
  •, the company started by Leo Sauermann (me) and Bernhard Schandl in November 2008. This is alive and kicking, you are reading our company blog right now.

The past Gnowsis prototypes worked very well in production use and showed how information can be organized. Technically, this was a desktop system featuring a "personal semantic web server" and a semantic wiki. Users preferred the "Miniquire" sidebar view and were very quick to organize their appointments in Outlook and taking semantic notes.
The difference to Cluug is that a typical "desktop" of a knowledge worker also lives in the web, so we moved the "personal semantic web server" from the desktop to the cloud. This also helps a lot to integrate it with other web applications and to access your personal data from your mobile (or iPad).

NEPOMUK started as the name of an integrated EU project NEPOMUK running from 2006 to 2008 (size: around 17 million EUR total). I helped writing the proposal and gathering the consortium from 2004 to 2005 and stayed within the project from beginning to end. Within the project, we had a "zoo" of prototypes to prove and test each idea:

  • NEPOMUK-PSEW in Java, a fully integrated Semantic Desktop developed since 2006 by the whole consortium and continued by DFKI and as testbed. It features graphical browsing (mindmapping), a task-tree to navigate tasks, a semantic wiki, a facetted search engine, ... . Basically, it merges many prototypes into one and is an excellent way to prototype and test ideas, and is used in production also in a handful of cases.
  • Beagle++, a semantic search engine done by L3S. It features links between files and much additional metadata.
  • iPad Electronic Lab Notebook done by Cognium Systems (funnily a company from Cupertino named another product iPad sometimes later without asking, damn). iPad ELN concentrates on taking laboratory notes in a legally compliant way.

  • Sponge done by IMU which reuses parts of the Java backend of NEPOMUK-PSEW and adds a different GUI.
  • iMapping by FZI which was compatible with the NEPOMUK-PSEW backend through a concept called "Conceptual Data Structures" and allowed users to graphically and hierarchically map their knowledge in a big "map".
  • ... some I forgot now, please comment!
  • NEPOMUK-KDE done by Mandriva Linux which takes the minimum viable and usable concepts of all the above and brings them to the KDE window manager. Nokia shows interest in it and started porting parts to Gnome.

From all of the above, the last one is the most popular one because KDE is used by millions of people. It has also the most active open source developers. Cluug is different from NEPOMUK-KDE as it focuses on linking things together, not just tagging them as we first did on KDE. NEPOMUK-KDE is also running as a desktop daemon and it still misses group functionalities (both and KDE are working to get group-functionalities into their systems).

To keep the active projects compatible after the NEPOMUK EU project ended in 2008, we founded OSCAF in 2009. Within this non-profit-organisation we (= mostly Sebastian Trüg from KDE/Mandriva, Nokia contractors working on Gnome, and me from Gnowsis) are discussing how to keep things simple and compatible. Mostly, the standardization work just happens within the open source projects.  More companies and research institutions are backing OSCAF and we are working to streamline OSCAF into a more focussed group. Cluug is compatible because the underlying data formats are standardized, this is very important to us.

Besides "our own European research", there is an important parallel development funded by DARPA in the USA with the CALO project. As far as I remember, they had a budget of around 200 million dollars - which definitly WOWs the tiny 10 million EUR public funding we spent on NEPOMUK. The results of CALO are many, for Gnowsis most interesting are:

  • OpenIris is a research prototype for an embedded workplace, featuring a semantic assistance. It is a complete new desktop application (reusing the Mozilla Suite of Calendard, Email, and Browser) and does not focus on integrating running systems such as Microsoft Outlook.
  • was started in the wake of CALO by Nova Spivack, it was a personal bookmark managing system that also features semantic analysis of text content. On 11th March 2010 it was merged with Evri, another company backed by Vulcan Capital. Nova wrote an excellent blog post explaining the story of Twine and the following merge.
  • was started by many researchers from CALO and backed by venture. In the beginning they were very "stealth mode" about their secret goals, but it somehow leaked that they were working on an audio-based personal assistant based on CALO and Semantic Web ideas. Their business model is focussed on providing consumers answers to good questions like "where is the next pizza?" or "which movie starts in 60 minutes in cinemas near that pizza restaurant?" in an intuitive iPhone interface. They launched February 2010 and were bought by that Cupertino company in April, I guess the investors are more than happy about the return their 8.5 Million USD (or more?) made in this outstanding company launch.

Alongside these successes, there are numerous prototypes like Haystack that worked on the Semantic Desktop idea. The difference of Cluug and Gnowsis from the USA-based CALO spin-offs is that our NEPOMUK lessons were learned and financed within the EU under the Lisbon Strategy. In 2000, the EU stated the goal to create the most competitive economy in the world, and a knowledge-based economy is a key prerequisite. So we at Gnowsis focus on knowledge workers in SME and enterprise settings, which drills down to the daily information-hunt of "find the stuff you need for that next meeting".

Hey, it's 2010 and we release Cluug Beta. Lisbon Strategy: accomplished.

Besides these fascinating stories from the recent years, there is a nebular history behind Cluug that got rolling with a printing press, the Buffalo library of 1883, a Memex desk in 1945, a Xanadu of links, Enquire in 1980 to link documentation, and the Web in 1989. We grew up with a tendency to built systems to link information together. I see that Cluug builds on this idea and is now the first product to give bidirectional hyperlinks to normal people. Let's see what you do with these links.

To sum up, Cluug is a semantic personal information management system developed by Gnowsis which runs as a cloud service, that differs it from what was done before:

  • Cluug is easy to integrate into other applications because there is a web API running on a cloud platform. This is why we moved away from the desktop - to integrate Cluug with your Google accounts, Salesforce, etc..
  • Cluug is available 24/7 in the cloud whereas your desktop is not. This we found out in our many prototypes - to make a semantic personal information management tool useful, you must be able to reach it using any application or device you have at hand. Be it your laptop at office or your mobile phone you use to lookup that crucial name in the bathroom during coffee break of that meeting - you need to re-find your information 24/7.
  • Cluug is geared towards professional knowledge workers, be it busy project managers or social marketing geeks. We need to cater the crowd of Microsoft Office users first, who really need a system to find what's related.
    (Apple users: yes, we also want to provide you with Cluug, buy the online account as soon as you can to help us fund the development! Most of our developers use Apples, they are also eager to get going on the MacOsX version  :-)
  • Cluug is compatible with web standards and we care that our users can keep their Cluug data for a lifetime. This is why we are part of the OSCAF standardization organization.