In October 2005 I reviewed a potentially disruptive search engine called ePrécis, from Syntactica. Unfortunately it got shut down by Google (they "nearly put us out of business", said Syntactica President Henry Neils). But there was some heavy duty linguistic theory behind ePrécis and the good news is that now Syntactica is back with a new product, which looks equally promising.

iReader is a browser extension (for both IE and Firefox, and on PC or Mac) that lets you preview the content of a link, before you click on it. It's similar to the web previews products we reviewed back in January - Browster, Cooliris, Snap, and Sphere. Our conclusion in that post was that previews are good, if implemented correctly - because previews can save us a lot of time over the long run.

iReader (full name: iReader 2.0 Web Previewer) is probably the most sophisticated previews product we've seen on the market yet, because it doesn't just preview a webpage like Snap, or relevant links like Sphere. iReader actually studies the semantics of the content behind each link, and pops up a preview of that content in the form of a short list. The best way to illustrate this is to show you an example. Here is what happens when I hovered over a link in an earlier R/WW post:

As you can see, the preview is useful because it gives me a bullet point list of (hopefully) the main content in the webpage behind the link. This is called an “intelligent summary” by the company.

However there are some rough edges to the app. Sometimes the preview content doesn't seem relevant, as in this example (also from R/WW):

The link previewed there was about Talkr, yet the pop-up displayed content about AOL (which was a whole other post). Perhaps this was pop-up lag, but I did notice some funny results elsewhere too. Another slight criticism is that the pop-ups tend to drive you crazy after a while. Every time your mouse passes over a link, up comes the preview pop-up. That's distracting when you're browsing the Web. But it is a beta product, so you'd expect some rough edges.

However when it comes down to it, there's something about iReader that is attractive. In the press release which will go out later today, it states that iReader "is based on linguistic technology the company has been developing for several years." And indeed, creating an on-the-fly pop-up summary of the content behind a link - in an easily digestable list - is a clever thing. So give it a bit of time. And for those interested in the technical explanation:

"The technology behind iReader 2.0 is Syntactica Web Services, which can be embedded into a number of computer programs that process English language text – including, for example, those for search output, search indexing, or book indexing.

The iReader 2.0 technology works by quickly digesting archived information of a web page and providing the essential meaning, or sense, of the text on that page. The technology can also be integrated into Internet search engines to produce relevant abstracts of text information in real-time. “In short,” said CEO Neils, “this technology quickly compresses massive amounts of electronic English text into meaningful short abstracts along with a reference index.”

What's more, the iReader 2.0 “macro” is being offered on an Open Source basis for developers.

I think this is promising product, albeit possibly too much of a distraction when browsing. But check it out for yourself and let us know what you think.