The Human Semantic Web - Further Reflection

The web is a useful environment for enabling people to add their knowledge in both a less structured Web 2.0 way (development of less structured but interactive web tools/programs), and a more structured Semantic Web way. The greater interaction in the Web 2.0 approach at least makes it more likely that Semantic disagreements will be spotted, but it takes the structuring in the Semantic Web approach to then show the meaning of terms more clearly and unambiguously so that agreement or disagreement about and mapping of terms can be reached. This then makes possible Naeve’s (2005) ‘Semantic Collaboration’ through and also defining the ‘Human Semantic Web’ that Naeve advocates. This then enables moving on from the Web being an environment only for simple tasks to one where sophisticated programs and models could be run that enable calculation and decision support.

This combination in approaches of enabling greater human interaction, and more definition of semantics can be illustrated by adapting the table displayed in my previous post.

Table - Language and Tool Mapping - Further Development



















Thus Increased Semantic Structuring and Collaboration from right to left, combined with Increased Human Interaction from bottom to top makes it more possible to undertake modelling and programming because the information is then well mapped and structured, and made available for visualisation and human interaction. On reflection AJAX/Web 2.0 technology spans more than one part of this diagram depending on the emphasis of whether to structure it and/or enable greater interaction. To reach the top left of the diagram requires layered use of technology as per the diagram developed by Berners-Lee (2000) and also McGuinness (2003). This layering of technology is needed in order to translate from the computer centred representations in the bottom right to the human centred representations and modelling in the top left. Human centred representations are too abstract for computers and computer centred representations are too abstract for humans. Therefore the technologies in the top left are not superior to those below and to the right as they need to be built on those technologies. Further there is more than one way to reach the top left, e.g. along the diagonal arrow from Naeve’s (2005) Semantic Isolation through Semantic Coexistence to Semantic Collaboration, or by moving up then left, or left then up. Following the diagonal arrow based on Naeve’s analysis is best for planning and building such a project from the start, but the other forms of navigation might be the best way to build on an existing project that has already been moving in a particular direction, that is not on this diagonal arrow.

References

Berners-Lee, T., (2000) Semantic Web on XML – Slide 10 - http://www.w3.org/2000/Talks/1206-xml2k-tbl/slide10-0.html

McGuinness, D. L., 2003. Ontologies Come of Age. In: Dieter Fensel, Jim Hendler, Henry Lieberman, and Wolfgang Wahlster, ed. Spinning the Semantic Web: Bringing the World Wide Web to Its Full Potential. MIT Press, 2003.

Naeve, A., 2005, The Human Semantic Web – Shifting from Knowledge Push to Knowledge Pull. International Journal of Semantic Web and Information Systems (IJSWIS), Vol 1(3) (July-September 2005) pp 1-30.

Naeve - http://kmr.nada.kth.se/wiki/Amb/HomePage

The Human Semantic Web

My thoughts on the work of Naeve http://kmr.nada.kth.se/wiki/Amb/HomePage and Enoksson on the Human Semantic Web and implementation of Concept Maps.

Enoksson (2006) explains the advantages for extensibility of an open standard language, he used RDF for Conceptual Browsing on the Semantic Web. Enoksson (2006) models things with concept maps that break an overall ontology down into concept sub-ontologies/taxonomies.

Naeve (2005) argue that “combining the human semantics of UML with the machine semantics of RDF enables more efficient and user-friendly forms of human-computer interaction.” Using UML for production of ontologies is as advocated by Baclawski et al. (2001) and Kogut et al. (2002), and Enoksson (2006). Naeve (2005) examines this strong separation between types (classes), and instances (objects) and considers this to be a weakness, which he rectifies for ULM (Unified Language Modeling) developed from UML.

Naeve (2005) gives an example of the need for “semantic mapping” between different words with the same meaning such as ‘author’ in one ontology and ‘creator’ in another ontology in order to establish interoperability and machine readability.

The Table below shows tools, technologies, and languages that can assist in this, and where they are based in a hierarchy from low level information centred interaction to high level user centred interaction (bottom to top), and computing focused to human focused representation (right to left). The Table also shows how each tool fits in with Naeve’s (2005) analysis based on “characteristics of the three different semantic stages” of “Semantic Isolation, Semantic Coexistence, and Semantic Collaboration” :-

Table - Language and Tool Mapping















Naeve (2005) describes Semantic Isolation where databases are available but hidden behind web portals, though the portals advertise their address. Semantic Coexistence is achieved by databases being structured in such a way that it is possible to search them without having to know their location. Naeve gives the example of RDF Schema - RDF(S), this standardises the structuring of the information across RDF(S) databases. RDF(S) provides standardised elements for the description of ontologies, so assisting to enable Semantic mapping. Semantic mapping enables Semantic Coexistence due to Semantic mapping enabling agreement on terms. For the table above the argument presented is that high level user centred interaction (bottom to top), and computing focused to human focused representation (right to left), enable Semantic Coexistence. The tools in the top left are built from those below and to the right of them so the Semantic Coexistence is built from Berners Lee’s (2000) Layered Architecture. Naeve (2005) argues the need for semantics that are understandable to humans as well as machines. That is an important objective of the research outlined in my thesis as without semantics that are understandable to humans, it is not possible for non programmer domain experts to undertake collaborative modelling. Naeve (2005) discusses a bottom up approach where there is a set process of deciding what can be agreed on, what cannot, and on documenting both.

Naeve (2005) argues that where knowledge is tacit it is vital to keep track of the individuals or groups who have this tacit knowledge, and that also the ‘Human Semantic Web’ can help elevate tacit knowledge to explicit.

References

Baclawski, K., Mieczyslaw, K., Kogut, P., Hart, L., Smith, J., Holmes, W., Letkowski, J., Aronson, M., 2001. Extending UML to Support Ontology Engineering for the Semantic Web. In: Proceedings of the 4th International Conference on The Unified Modeling Language, Modeling Languages, Concepts, and Tools, pp 342-360.

Berners-Lee, T., (2000) Semantic Web on XML – Slide 10
http://www.w3.org/2000/Talks/1206-xml2k-tbl/slide1-0.html

Enoksson, N. (2006) Serverside Solution for Conceptual Browsing on the Semantic Web. MSc. Dissertation, Stockholm University.

Kogut, P., Cranefield, S., Hart, L., Dutra, M., Baclawski, K., Kokar, M., Smith, J., 2002. UML for Ontology Development. The Knowledge Engineering Review Vol 17(1) pp 61-64.

Naeve, A., 2005, The Human Semantic Web – Shifting from Knowledge Push to Knowledge Pull. International Journal of Semantic Web and Information Systems (IJSWIS), Vol 1(3) (July-September 2005) pp 1-30.

How Ontology and Modelling Tools were chosen for my PhD

The modelling tool Vanguard System was chosen because it handles Units and uncertainty well something that was necessary for the DATUM project with Rolls-Royce as partners (Scanlan et. al, 2006). The same advantage was the case for the PhD because of the need for extensibility such as risk and uncertainty modelling. Other advantages of Vanguard System were ease of installation and use, ease of linking to spreadsheets and databases, facilities for web enabling of models, and the ease with which formulae can be entered and linked with a high level programming language where necessary (similar to spreadsheets but with a tree based user interface). Experimentation with Protégé showed it was possible to translate the Protégé tree into a Vanguard System tree. This fit in well with the stepped translation to be developed. The open standard nature of Protégé made it possible to use this software without being locked in to it. Tools such as TopBraid Composer can provide additional higher level functionality such as an improved user interface and more tools for user interaction and modelling by end-users so is future work.
There was a need for the DATUM project and for the PhD to minimise programming, so Jena was not used, but Leavers' (2008) MSc project used this effectively, and there was regular contact with the developers of ACUITy (Aragones et. al., 2006), to examine how that more software centred approach was used. Given more time that approach could have been used. Metatomix M3t4 was also used effectively as a high level tool to interact with Jena. So results of research with Jena and/or ACUITy and m3t4 would have shown similar results to the approach of using Protégé and Vanguard System.
These types of tools improve with research and development each year so reproducing this research is becoming easier. New ways of modelling at high level with involvement of end-users is thus practical.

References

Jena - A Semantic Web Framework for Java [online]. Available from: http://jena.sourceforge.net/ [Accessed 21 April 2011].

Leaver, N. (2008) Using RDF as an Enabling Technology. MSc. Dissertation, University of the West of England, Bristol.

Metatomix M3t4 Dashboard [online]. Available from: http://wiki.m3t4.com/homepage.action [Accessed 21 April 2011].

Protégé Welcome to Protégé [online]. Available from: http://protege.stanford.edu/
[Accessed 21 April 2011].

Scanlan, J., Rao, A., Bru, C., Hale, P., Marsh, R., 2006. DATUM Project: Cost Estimating Environment for Support of Aerospace Design Decision Making. Journal of Aircraft, 43(4).

TopBraid Composer, The Complete Semantic Modeling Toolset [online]. Available from: http://www.topquadrant.com/products/TB_Composer.html [Accessed 21 April 2011].

Vanguard System [online]. Available from: http://www.vanguardsw.com/solutions/application/modeling-and-simulation/
[Accessed 21 April 2011].

What's up doc? Join the blog for postgraduate researchers - My case for an invite

This relates to my previous post of April 7th about the research blogging event in Birmingham next month. I've posted to their blog and emailed them to make my case for inclusion. This is my post to their 'What's up doc?' blog - Postgraduate Conference for Computing: Applications and Theory (PCCAT 2011). The post itself is about a different event, a computing Conference at Exeter University and the abstract of my paper for that.

What's up doc? Join the blog for postgraduate researchers

Below are details of an interesting event about blogs for and by researchers :-

'What's up doc?' is a new blog which provides a forum for postgraduate researchers to share experience, give information and tips on matters relevant to them in a fun, yet informed way.

It replaces GRADBritain, which has been a popular online magazine written by and for postgraduate researchers studying in the UK.

As part of the launch we will run a free blog training day in Birmingham on 12 May. Apart from learning what makes a good blog post, we will also explore how to increase your online profile. To apply:

* post an example of an article on the 'What's up doc?' blog

* send an email outlining why you want to be involved, what discipline your research is in and what you would do to champion 'What's up doc?' to tennie.videler@vitae.ac.uk before 25 April.

We will select applicants based on this and reimburse reasonable travel expenses. Places are limited.