BIG Data and Journalism

An interesting post, over at GigaOm, on When big data meets journalism, talks about how companies are using the power of tools that allow journalists to analyze information. At the least, through simply analyzing content for times, dates, places, phone numbers, data (structured and unstructured) and people references, a lot of connections to a resource can be uncovered (remember Zoetrope?). This becomes the basis for a whole lot of possible collaboration and contribution to a topic.

I think this is a format that is extremely well suited for collaborative educational research. I am sure people will be worried about the quality of connections that are generated by a machine algorithm, but this can get better over time and actually allow curation as well. But the idea that structured and unstructured sources of information can come together in a Powerset, the erstwhile Twine and Zoetrope manner, is brilliant for the learning process.

The ability for a learner to be able to get such views of information in a curated manner is going to be really important. The information largely exists on the web, but really in as many bits and parts, rendering it unusable or very inefficient from a learning perspective. Our current mode is to do the search and use intelligence to sift through many dead-ends and irrelevant information to actually get what we need in the way we need it. There simply has to be a way to accomplish the latter, at least to a large extent. It is only then that the online learning process will become efficient enough for more people to use.

 

Read Full Post »

The temporal internet – Zoetrope

The vision of the web as a site of history may not be old or far-fetched. Check out the Wayback machine at www.archive.org. They even have a K-12 Web Archiving Program!

We all know that content on the web changes constantly. How do we maintain a track of that content change? How do we play back history? Zoetrope points the way.

The Web is ephemeral. Pages change frequently, and it is nearly impossible to find data or follow a link after the underlying page evolves. We present Zoetrope, a system that enables interaction with the historical Web (pages, links, and embedded data) that would otherwise be lost to time.

Zoetrope uses technology to search and analyze data in an Internet archive. The search can be on a specific section of a web page (say, for example, a news headline section on a web page).

A zoetrope is a device that produces an illusion of action from a rapid succession of static pictures. The term zoetrope is from the Greek words zoe, “life” and trope, “turn”. It may be taken to mean “wheel of life” or “living wheel.” (Wikipedia)

The search can be then correlated with other contemporary (to the original search) searches and a picture of the search domain at that point of time becomes available.

Of course, this picture is as fallible or accurate as the web content can be.

What does this kind of technology imply for education? There could be multiple applications. I could trace the way the world political map has evolved or how a city’s traffic system has evolved by simply playing back such images from the Zoetrope system. I could investigate multiple factors affecting a certain economic situation which occurred 20 years ago. The possibilities are enormous. The technology then becomes a guide through digital memory and the playback a conversation that the “guide” could have with a student.

Interestingly, this is the second time in a few weeks that the term “playback” has crossed boundaries for me. First it was Google Wave that played back email conversations temporally, now Zoetrope.

What would be lovely is to have digital visual immersion into the past as well (or sensory at some point!). The context for the learner would be considerably enhanced by virtualizing the past rather than reading it from a book/online or listening about it from a teacher. This in turn would create vastly more effective learning environments. (Add Project Natal and we could even “be” in that world!)

The power of this technology may also manifest itself in forecasting. When engines become powerful enough to traverse this huge database and generate many more relationships/variables hitherto unavailable in research, temporal data could be modelled to generate predictions of future behavior using time series or other advanced analyses.

Read Full Post »