Archive for June, 2008

Don’t make me think?

The article Is Google making us stupid? published in The Atlantic examines the way the Web influences the way we read – and think. Many users have the impression their attention span for reading longer articles and books has decreased with the use of the Web, and the author supports this impression with results from neurological and psychological experiments. The way we skim information on the Web has an influence on the way our brain works, i.e.

We can expect […] that the circuits [inside our brain] woven by our use of the Net will be different from those woven by our reading of books and other printed works.

As a usability consultant, I feel I am not unaccountable for this development. We support the hectic, unconcentrated behavior of our users and proclaim rules for writing online texts quite contrary to what we learnt at school: Not to substitute nouns with synonyms to make a text more varied, not to use figures of speech or word-play (not to mention irony or allusions) because they are imprecise or ambiguous. Not only do we support the behavior the author of the article bemoans, but trends in artificial intelligence which treat terms open to interpretaton as bugs to be fixed. What is at stake is that

In the quiet spaces opened up by the sustained, undistracted reading of a book, or by any other act of contemplation, for that matter, we make our own associations, draw our own inferences and analogies, foster our own ideas.

So see for yourself if you are still able to concentrate on an article of over 4000 words (more than 5 A4 pages on my printer) peppered with historical examples in the best scholarly tradition. If you can, I’m sure you will enjoy – that it makes you think.

Zotero research tool

I admit I’m stretching the category of «information literacy». But I’ve just discovered Zotero, and I’m enthusiastic about it. It’s a Firefox extension to import, manage, annotate, tag, search for, export, … research sources. With one click, it imports bibliographic data from library catalogs as well as Amazon, Google Books and others, and the same functions can be used for capturing websites.

Is there a future for books?

A study of the reading habits of German produced some interesting results, e.g.

  • people spending much time on the Web read more books than others and vice-versa
  • people who read a lot spend more time reading than they did in the past (for which I couldn’t find a definition), while people who read little indicated they read even less, i.e. the gap between the «information rich» and the «information poor» is widening.
  • Harry Potter – not surprisingly – has a great effect on reading. Kids who read Harry Potter read significantly more books than those who don’t read this series.

The study (in German): PriceWaterhouseCoopers, Haben Bücher eine Zukunft? found on Infobib.

Metadata revival?

Metadata is a big thing with archivists and other people concerned with context, but I must admit that in all my professional years, I have never worked on a web project which actually used the Dublin Core Metadata set. The most probable reason that people don’t seem to bother much about metadata – at least in a standardized form – is that popular search engines don’t seem to take them into account. Or at least they didn’t until recently.

Let’s have a look at a (very!) brief history of search result designs:

Search engines I rememer from my early web experience returned results looking somewhat like this (our government portal hasn’t made much progress since):

Search result for passport on www.ch.ch

Then up comes Google and introduces a design which has become pretty much standard:

Google Search for \

More recently, both Google and Yahoo have started introducing structured search results:

Yahoo Search Gallery, Country Profile Armenia

Search results are also increasingly shown in clusters based mainly on format (text in general, news, entries from encyclopedias; images, video etc.):

Yahoo India \

So adding metadata in some kind of standardized form does seem to be a recent trend for

  • clustering search results and
  • displaying search results.

Metadata provided by the creators of web sites are used for these displays. However, these metadata are explicitly not used for search algorithms, as an article on Yahoo and the Future of Search reports. Metadata provided by the creators tends to bias the outcome, and the analysis of broader text corpus by powerful search engines provides more signifcant results than metadata out of context.

Still, the increased use of metadata are pointing to interesting directions:

  • Search results are becoming more context-sensitive. Metadata help the user to choose the appropriate context, e.g. for disambiguation or clarification of a query. Search interfaces are taking the iterative nature of search into account and getting closer to the process of questions and answers users require to clarify their needs.
  • Possible actions after having found the desired content are beginning to be transferred to the search sites (search engines becoming portals may – or may not – be part of the development). Users can view details, maps or reviews, check opening hours, buy tickets or conduct site-search without having to leave the search results page. This is enabled by a deeper integration of applications into results.

Google Search for NASA

Site-search from search results page