Bringing data alive - democratising information
Computation has come of age, says Conrad Wolfram, changing how we represent knowledge, and how authors and readers interact, quite dramatically - changing what 'publishing' constitutes, and what comprises both human and computer 'knowledge'.
Linguistic analysis and computational knowledge
Conrad starts with a quick initial demonstration of Wolfram Alpha (WA), the computational knowledge engine launched last year. He describes how WA picks apart a question ("what was the weather like when Gordon Brown was born?"), interpreting it mathematically to find and collate answers. He demonstrates other searches with ambiguous terms, to show how it uses context to interpret terms in different ways (Brown as a person or a colour; Wolfram as the name of a person or an element). WA linguistically processes and pattern-matches input to create symbolic expressions that represent our query computationally. (Lots of librarians at Wolfram apparently!) The query is matched by scanners to the available data, and the matched elements are combined and rendered together as your answer. (Or, um, something broadly like that.)
Changing role of human experts
Lots of organisations are involved in making knowledge computable - content providers, curation services and so on. These won't become unnecessary - we will need them *more*, to add the automated layer - but their expertise will become broader. (He's not specifically talking about librarians, or publishers, but to me this reflects exactly how all our roles are changing).
Research apps: visualising, not describing, science
WA is one window onto a much bigger project. It is built on four pillars: linguistic analysis (algorithms), curated data (30 trillion pieces, mixture of manual and automated curation), dynamic computation, computed presentation (a bigger deal than you'd think - how to make a document live - issues such as layout, when you don't know what kind of data you'll be rendering). The team is currently working to mesh documents and applications - very separate concepts for most people, but apps are "one of the best ways to communicate technical ideas - the roles of authors and readers need to change to a bidirectional flow, a much more interactive process than reading a journal article." Conrad shows a document with an in-built app that is rendering the data for user manipulation. You can describe science, but building an app into a journal article makes it much more powerful - easier to engage and understand. "The author has set up a workspace, and I'm driving through that workspace" - improving what Conrad calls the "bandwidth of communication" between authors and readers. (Having played with Conrad's iPad briefly this morning, I can suddenly see the wonderful potential of research apps).
Live textbooks
The Wolfram Demonstrations Project is a 'knowledge apps' site where people post apps designed for communication. "What's exciting is the idea of what's here for other publishing ventures:" Later this year, Wolfram will be pulling together several of its technologies to create a "computable document format" - open, interactive, supported by solid publisher workflow, making development and deployment easy - "to finally have active stuff happening in textbooks, scientific journals". This same approach will see apps rendered as part of WA results, so that you can manipulate and engage with the data it returns - "not a dead document but an application that is live, unique".
Democratising information application
"The web has democratised information retrieval outstandingly successfully; the challenge now is democratising its application so that people can actually use it, with the help of automation." Examples include government assets - not just putting the knowledge out there, but making it open and useable. "Useable increasingly means computational" - governments are spending $3bn annually on research, with just flat papers as output ("low bandwidth"). "There are much richer ways to do it - we don't have to be stuck in the Newtonian age."
Conclusions
9th June - one day conference - London Computational Knowledge Summit.
Linguistic analysis and computational knowledge
Conrad starts with a quick initial demonstration of Wolfram Alpha (WA), the computational knowledge engine launched last year. He describes how WA picks apart a question ("what was the weather like when Gordon Brown was born?"), interpreting it mathematically to find and collate answers. He demonstrates other searches with ambiguous terms, to show how it uses context to interpret terms in different ways (Brown as a person or a colour; Wolfram as the name of a person or an element). WA linguistically processes and pattern-matches input to create symbolic expressions that represent our query computationally. (Lots of librarians at Wolfram apparently!) The query is matched by scanners to the available data, and the matched elements are combined and rendered together as your answer. (Or, um, something broadly like that.)
Changing role of human experts
Lots of organisations are involved in making knowledge computable - content providers, curation services and so on. These won't become unnecessary - we will need them *more*, to add the automated layer - but their expertise will become broader. (He's not specifically talking about librarians, or publishers, but to me this reflects exactly how all our roles are changing).
Research apps: visualising, not describing, science
WA is one window onto a much bigger project. It is built on four pillars: linguistic analysis (algorithms), curated data (30 trillion pieces, mixture of manual and automated curation), dynamic computation, computed presentation (a bigger deal than you'd think - how to make a document live - issues such as layout, when you don't know what kind of data you'll be rendering). The team is currently working to mesh documents and applications - very separate concepts for most people, but apps are "one of the best ways to communicate technical ideas - the roles of authors and readers need to change to a bidirectional flow, a much more interactive process than reading a journal article." Conrad shows a document with an in-built app that is rendering the data for user manipulation. You can describe science, but building an app into a journal article makes it much more powerful - easier to engage and understand. "The author has set up a workspace, and I'm driving through that workspace" - improving what Conrad calls the "bandwidth of communication" between authors and readers. (Having played with Conrad's iPad briefly this morning, I can suddenly see the wonderful potential of research apps).
Live textbooks
The Wolfram Demonstrations Project is a 'knowledge apps' site where people post apps designed for communication. "What's exciting is the idea of what's here for other publishing ventures:" Later this year, Wolfram will be pulling together several of its technologies to create a "computable document format" - open, interactive, supported by solid publisher workflow, making development and deployment easy - "to finally have active stuff happening in textbooks, scientific journals". This same approach will see apps rendered as part of WA results, so that you can manipulate and engage with the data it returns - "not a dead document but an application that is live, unique".
Democratising information application
"The web has democratised information retrieval outstandingly successfully; the challenge now is democratising its application so that people can actually use it, with the help of automation." Examples include government assets - not just putting the knowledge out there, but making it open and useable. "Useable increasingly means computational" - governments are spending $3bn annually on research, with just flat papers as output ("low bandwidth"). "There are much richer ways to do it - we don't have to be stuck in the Newtonian age."
Conclusions
- Knowledge is now available - the question is to make it accessible.
- The semantic web isn't enough - we need to make it computable.
- Value used to be in based knowledge, base content. Now it's in active workspaces, workflows for authors distribution models, compelling interfaces for knowledge.
9th June - one day conference - London Computational Knowledge Summit.
Labels: "active workspaces", "computational knowledge", "knowledge applications", "linguistic analysis", "live books"
posted by Charlie Rapple
at 4/12/2010 12:59:00 pm
Subscribe to this blog
1 Comments:
Good post, here's a couple posts that link digital content to the library, also discusses the importance of curating in a digital age. The librarians need to somehow get control of the important content otherwise its the wild west! Here are the links:
Curating content
and
Information and Content topics
Post a Comment
<< Home