Powerpoint Deck: Making Sense of the Semantic Web, and Twine

Now that I have been asked by several dozen people for the slides from my talk on "Making Sense of the Semantic Web," I guess it's time to put them online. So here they are, under the Creative Commons Attribution License (you can share it with attribution this site).

You can download the Powerpoint file at the link below:

Download nova_spivack_semantic_web_talk.ppt

Or you can view it right here:

| View | Upload your own

Enjoy! And I look forward to your thoughts and comments.

Comments

[Lars Ludwig]

That is a knowledgeable overview, thank you. - However, often people in the Semantic Web community tend to overlook what is changing in the human mind when introducing triples. Yes, it enables machines to better use data. But, at the same time, it is a fundamental shift in how we deal with information as well. Let me explain: Say I am reading a text. When I note something down, traditionally I do so in a document, in a sequential manner. If I, instead of writing a text, annotate the information using triples, I will do so in the same sequential order. There seems to be no fundamental difference. Later on the triple information can be searched and queried. Writing down triples seems to be a lot of work, people think. Why not use the information given and annotate it automatically by transforming databases or micro formats etc. What is forgotten here is that we think in semantic associations. A triple is a simple form of semantic association. Empowering us semantically means to make these associations available in a non-sequential order to create and extend our semantic association net of thoughts and words. One could call this creating a personal artificial memory. Every annotation is part of a personal artificial memory. It is knowledge, not information. The Semantic Web thus becomes a Web of personal artificial memories. If we do not create personal Semantic Webs, the knowledge triples tend to lose some of their value, as triples that are out of our knowledge realm will mix with information that is not relevant or simply not known to us. That seems ok for the purpose of enabling search and answering queries, but it prevents us from including new knowledge to a personal reference net. We cannot see anymore what is relevant to us. The advantage of finding new information will be outweighed by the cost of filtering irrelevant information. To give an example: Would it be useful to note down a word in a foreign language, and later, when you want to repeat it, instead to open a huge dictionary, just because the foreign word is part of a foreign language that you learn. The triple ‘bonjour is a French word’ is useful if it is available to me as the French words I know/learned before. One might say, well, where is the problem: you can search your own annotations. Well, not really, because personal information will probably be scattered across different semantic webs in the same way my social graph is scattered across different platforms today. Another value of a personal Semantic Web would be that instead of writing non-normalized documents, each of us could create a mirror image of our mind that would be an ideal source for communication and knowledge exchange. Well, there is so much more to say …

Lars Ludwig, www.artificialmemory.net

[Samuel Driessen]

Thanks for sharing this presentation. I've been looking at Twine and following your blog (and waiting for an invite to use Twine!).
I'm really curious where this will bring us.
A couple of questions based on your presentation:
- I'm surprised you position the semantic web in 2010-2020 (web 3.0) - sheet 6. We already were promised a semantic web in the nineties right? Shouldn't you call this new web something different?
- This brings me to the next question: what has really changed since then? Is it the 'Hybrid approach' (sheet 16): AI, NLP and us? So we had the first two (and of course that improved in the last decade), and now making that social. I do think this will take us a step further. But do you think it's enough? Will we get people to help build this web?
- Sheet 27: you mention that technologies and tools were "insufficient". What do you mean by that? And why is it now "sufficient"?
- Sheet 32: Twine is focused on digital data/information. But our lives also consist of physical (knowledge) objects, like books and paper. How do we add them to Twine? Is that possible or will it be?

[Samuel Driessen]

Thanks for sharing! I'll be posting about Twine and other 'semantic web tools' these days.

[peter royal]

you should post 'em at http://slideshare.net .. you can then get a little widget and embed the slides in your blog :)