I last discussed Wolfram|Alpha here in Network World Web Applications Alert newsletter in two issues at the end of May and the beginning of June.
What attracted my attention at that time was the first birthday of the service and its huge expansion in terms of what it “knows”, which has been a huge step forward.
Today, there’s a new development in Wolfram|Alpha that is just as intriguing: Wolfram|Alpha Widgets.
WAWs are essentially snippets of dynamically extracted Wolfram|Alpha data. You start by formulating an inquiry from the “New Widget” link in your Wolfram|Alpha account (you’ll need to sign up for a free account if you don’t already have one).
First you set up a query. For example, “$1 in 1900” results in the following:
Input interpretation: $ 1 (current US dollars) in 1900
Result: $ 0.04(1900 US dollars)
Average rate of inflation: 2.94% per year
Clicking on the “Use it” button redisplays the query in the “Query Template Editor” which allows you to select the terms in the query that are to be considered as variables. You can select the variable name, the label to be displayed, the default value, the type of variable (the service calls them “Component Types”), and the Interpretation or data type (number, date, etc.)
Once you are happy with these definitions clicking on next takes you to the layout section where you can use drag-and-drop to move the fields you defined around in the widget layout.
Clicking “Next” takes you to the output section where you can choose which of the output formats you want the widget to display. You can include the assumptions as deduced by Wolfram|Alpha, the warnings about the results, and then select any of the various outputs.
The next section allows you to choose the format of the display of the widget: These are Lightbox (pop up on top of the page with the page grayed out), popup (as the last but in a new browser window), or iframe (suitable for embedding in a Web site). You can also define the height and width of the final widget.
The final step is publishing your widget and Wolfram|Alpha provides a pretty complete range of social networking tools to do this as well as providing embed codes.
I tried the service on a number of queries and only the “$1 in 1900” query was a problem (while the results shown for the initial query were right, the results by the time I got to the output step were “(no exchange data available)” …
So, the bottom line is that this service is brilliant. Not only is it free (for the time being) but it leverages the huge data and algorithmic resources of Wolfram|Alpha to make available widgets that analyze specific data sets in a compact and relevant manner and that can be integrated into Web sites and blogs.
That said, it is early days and when the service produces results that are not what you expect then, well, then it’s a long slog to figure out what’s going on.
I expect Wolfram|Alpha Widgets to become a huge market force and the only limiting factor will be how much they can improve the service to make the results understandable when they aren’t what you expect.
