wrong audio version (repeat of last week)

Correct MP3 is now live.

Doesn't Sherlock do this on the Mac (the scraping and Precis generation I mean)? It did at one time.

The gesture feature (which I guess they just added) eliminates the need to have a function key to turn on and off.

You hover, notice the little box appear. If you want to see the iReader summary, just move the cursor into the little box. Not bad!

I installed the Firefox download, restarted Firefox and then noticed only the tiny boxes. I thought something was wrong, so I looked at the Syntactica site and got the word on the new 'gesture' feature.

Just move your cursor into the tiny box.

This is good info and insight into search...

BUT - how does this fit in to the greater scheme of things that you typically discuss?

I want my big picture Cringely back - please.

And now back to Apple...

That is great, don't let this Cringely go - interesting but with explanatory power and a prediction, what more could one ask for from an industry commentator? It is this sort of stuff that lets one understand the big picture, thankyou.

I'm not very impressed. The pop-up is annoying, right click options such as the select and click option in firefox are much less annoying. As for the precis concept, in most cases you have a pretty good idea of where you are going. In most cases the search engine has selected a set of similar links, and you are trying to look for the differences. What would be really useful would be a smart diff. Sort of a cross product between the search phrase and the entry. That would change everything radically!

I use Copernic Summarizer to create summaries of any electronic document. Over the years I've found it very useful to get a quick overview of any website, or pdf document ... in fact almost any document. I've rarely come across any mention of it on the web. Maybe your readers will find it interesting too.

Syntactica's previous product was a profoundly unimpressive 'automatic indexing' program which I reviewed here as follows:

Syntactica does almost nothing that an indexer would recognise as analysis. It also makes blunders which any indexer would avoid.

A brief trial of IReader does nothing to restore my confidence in the company's abilities. Here is an iReader-produced summary of my article on Linux.

• Keep track and modifying contents of computer's storage media
• Two requirements mean quite difficult and dangerous to have two or more operating systems installed on same computer
• Windows shipping as part of huge package including set of internet programs, basic graphics and word processor applications and even games

In a word, huh?

Quite apart from the mangled English, how can we take a 'summary' seriously when it doesn't even manage to pick up the main topic of the page?

Syntactica needs to write out a hundred times:'I must not make extravagant claims that I can't fulful'. And Cringely needs to write fifty times: 'I must test exciting new concepts before I write about them'. He's losing credibility fast.

Syntactica's previous product was a profoundly unimpressive 'automatic indexing' program which I reviewed here as follows:

Syntactica does almost nothing that an indexer would recognise as analysis. It also makes blunders which any indexer would avoid.

A brief trial of IReader does nothing to restore my confidence in the company's abilities. Here is an iReader-produced summary of my article on Linux.

• Keep track and modifying contents of computer's storage media
• Two requirements mean quite difficult and dangerous to have two or more operating systems installed on same computer
• Windows shipping as part of huge package including set of internet programs, basic graphics and word processor applications and even games

In a word, huh?

Quite apart from the mangled English, how can we take a 'summary' seriously when it doesn't even manage to pick up the main topic of the page?

Syntactica needs to write out a hundred times:'I must not make extravagant claims that I can't fulful'. And Cringely needs to write fifty times: 'I must test exciting new concepts before I write about them'. He's losing credibility fast.

Sounds like it would be good for creating synopses in an RSS newsreader. As a browser plugin, it sounds less useful.

One Test Case Does Not a Review Make.
Syntactica's iReader works well some summarizing some pages, not so well on others. Just because it may not have worked on one article on Linux as the author may have wished is no reason to take affront. How many useless pages of search results has he ever gotten from Google (and any other search engine, for that matter)? Never having seen the article the author cites, bullet two of the summary seems to get at a major point on the topic of Linux, running multiple operating systems. And the OS it's often run with is mentioned in the next bullet. That tells me something quickly about the article's content, and whether or not I want to read it. Sure, the word "Linux" doesn't appear in the summary, but it very likely appeared in the link itself, the article title, that the user hovered over. Granted, I wouldn't want to see a summary of one of my hard-written articles come out so sketchy either, but that doesn't mean it can't save time for the reader as a scanning tool.

Apple’s OS X has had this feature for some time. It is pretty neat, and largely unadvertised.
http://docs.info.apple.com/article.html?artnum=151890

I don't think either one of you really read the article fully.

He did not have problems with the product because he used Linux; he found it moderately useful but not useful enough to be worth keeping active all the time.

OS X can summarize, but it doesn't, as far as I know, summarize the content of web pages as you put the mouse over the URL; nor does it probably summarize as effectively or as accurately, without that special dictionary.

So you do or don't like inflated women?

"Sea mammal obsessions are bad" is the meaning of Moby Dick? I guess the meaning of the Bible is that the Jews lived in the Middle East. Or the meaning of Lord of the Rings is that there are bad people around.

Just gotten hooked over the last month or two. Bit disappointed with this article, which is not to say that it's bad, but just that the others have been so GOOD. This is just a product review. It seems out of place, unless you're setting the stage for some future discussion...

Good stuff. Not the usual Google or Microsoft vs. the world talk. I'll cehck out Ireader and Sintactica.
Keep it coming.

Your article apparently can be summarized as:

• Following by true graphical interface and last decade lot of people viewing that graphical interface least in part through browser
  
• Large part for legal reasons intended to keep Googlers at bay
  
• Throw wording that carry no real information words like
  
• Adding vectors in column yielding single vector representing column in multidimensional space defined words articles in database

Would anybody recognize that as having anything to do with your article? The "summary" doesn't even mention iReader.

Funny, after installing iReader, I moused over your link for "Syntactica, iReader White Paper" and got no little pop-up balloon. I guess it's not capable of introspection.

I just downloaded and tested iReader and agree with everything you said. I really like the enhanced browsing functionality this offers.

Based on their blog, they have also added a new "gesture" feature which pops a tiny talk balloon prior to showing the entire abstract. This way, users can mouse over the balloon to see the full abstract, or just pass and move on to the next link.

I also wouldn't be surpsrised if one of the big guys (Google, MSN, Yahoo) bought Syntactica out and started running sponsor links within the abstract text. Now they could effectively generate revenue from any site on the web that has a text or banner link...including their competitors.

iReader added the "gesture" feature a few days ago, which works only with the Firefox plug-in right now, as their blog states. This product is still very much in beta, and they're updating the plug-ins (both IE and Firefox) to new versions quite often. I know some of the folks there, so they let me test a version about to go live that now adds a translation feature for your summaries. In Firefox for the Mac, you just go to "Preferences" and use the drop-down to select the language. Prettty cool, but I don't know any of the languages well enough to comment. Anyone out there care to weigh in on what the translations are like?

Trying to mouse over the white paper link won't I think the new feature of only displaying a gesture on the link, and then waiting to get a preview until after you mouse over the gesture will make iReader much more usable. This improvement was based on comments by blogger's like you.

There is also a new feature that you can set using Tools/Options to select a different language to translate the preview into. Only in the Firefox version right now.

The best reviews we've received on its use so far is from people who are doing a lot of research using web searches. They can easily preview content quickly when looking for things.

It really is like speed reading the linked web page. And there is an item in Tools/Options to set how long of an abstract you want: the default is fairly short, so you might want to set it higher.

The first paragraph of my last comment was mangled.

A mouse over on a link to a PDF doesn't work. iReader can't currently analyze PDF documents.

I use a Firefox plug-in called Browster that is also a mouse-over tool (well, actually you have to mouse over an icon near the link so it's not as annoying as what you describe). But rather than summarizing the contents of that link, it just shows the webpage on a popup screen.

When I first heard about it, I didn't understand the usefulness but soon, it became an indispensable tool for me.

By the way, can we just get back to Apple, Google and Microsoft rumors? I miss the good old days!

I downloaded it and checked it out. The gesture feature is really nice, but when I tried to use it get an idea of the pages on Syntactica's site I didn't get any useful information (most pages didn't have enough information to generate a good precis).

What would a page have to read like to humans to generate a summary that makes sense?

As it is, there is some limited applicability to this tool, but a couple of simple features (like grabbing the page description from the header metadata) might be helpful in browsing (and encourage people to fill in their information in the description)

We have a tropical cyclone watch in place at the moment, I tried using this tool on the BOM URL, and this is what I got;

* Seeing Glossary to find out weather terming mean
* Please note Acknowledgement notice relating to use of information on site.

Compare for yourself.
http://www.bom.gov.au/weather/nt/cyclone/

IReader needs work, a lot more work.

Services/Summarize does the equivalent in Mac OSX. Its summary this column is:

I wish the Syntactica people well (I do not own stock). They've done a good and difficult job, published parts of iReader as open source, created some useful APIs and built the whole darned thing as a web service that can be built into all sorts of gizmos. They also seem to think that iReader will become a popular alternative to browsing, especially since it comes minus the ads. I just don't see that.

"Personally, I prefer my web content -- like my women -- real, not inflated."

Sorry about this, but while we are talking about deducing meaning I can't help but point out that "Women" has two pronunciations, and that one is plural and the other singular. As a married man I guess you will take the 5th if I ask which was intended in the sentence above:-)

Unfortunately, the "BOM URL" has a bunch of links on it, but not much text content. So hard to come up with a good text preview.

Also, when using iReader on the syntactica site itself, some of the developer links on lower left have content on the pages. Several others are "under construction", and not much content.

Pretty unimpressive. The time it takes to wait on the little box to fill I could click and load on broadband. With what is in the URL one can usually know about as much as fast as this. Just another gimmick feature not worth having.

I also found its summary pretty kindergarden-ish

If you are on Mac OS X try this...

1) Go to any web page (this page is good for this excerise)
2) Perform a "Select All" to highlight all the text
3) Click on the Safari menu in the menubar and navigate your way down to the "Services" menu, from there select "Summerize" from the sub-menu that appeared from the "Services" menu.

What you should have is a full summary of the current page in a small clean UI with a slider to control the size and depth of the summary for the current page (by line and paragraph).

And it's free.

Ted to agree about OS X. Looking at the iReader examples above and the OSX examples, I just tested it on a forum page, Mac OS Summarize feels light years ahead. Now if someone could turn that into a live widget...

Articles and conjunctions ARE important, and iReader ignores them. Take for example your "submit a scoop" link, wherein the first sentence "Bob is always looking for a hot lead" is shortened to "Bob looking for hot lead." You don't want hot lead, do you Bob?

It occurs to me that the reason iReader uses news stories in their example is because these are already written in a 'pyramid' style which is specifically designed to be summarised by truncation. Summarising a story written in this way should be a no-brainer anyway.

I was also disappointed to find that iReader summaries can't be copied and pasted - so even if they were useful you can't go on to do anything further with them without getting a screen shot.

A bit longer version of the OSX Summary tool:

Though a thousand readers will correct me with their superiorly nuanced views of the past, let's say we generally began with punched cards, went to command lines, then to text-based graphical interfaces, followed by a true graphical interface and for the last decade a lot of people have viewed that graphical interface a least in part through a browser.

...IReader (which was called Speed Reader until a moment ago) started life a year ago as a search product called ePrecis, an application that could look at an article, book, or a web site, and give you its meaning in a prioritized list of short sentences.

...Searching pretty much requires scraping the Internet for content that is then indexed, while iReader's new browsing metaphor doesn't kick into action until the user mouses over a URL (no clicking required, hence no stepping on the toes of Google or any Google competitors).

...This compilation considers the many different meanings and contexts of each individual word in the lexicon, and assigns a set of values to each word, which is a heck of a lot of work and explains why most competing products (there turn out to be a bunch) don't have it.

...And if not, then I predict the product will be used to create a whole new class of annoying web denizens -- web pages made up only of such abstracts as a kind of meta-search in the same way that so much of the web is now clogged with meta-ads meant to sucker us into clicking on them.

Google's "Quest to organize the world's information and make it universally accessible and useful"
Not happening!
One of the big problems with search is that there is no attempt to organize (catagorize) the searched content as has always been done in traditional libraries.
When you search the Web most of the resulting pages contain less than 500 words. This is one of the reasons you get back so many useless hits.
Research shows that the known Internet (approximately 20 billion pages) is growing by more than 10,000,000 new, static pages each day. In contrast, the fastest growing search engine database is increasing at about 10% of this pace and so far has indexed less than 20% of the total pages.
Most of the search community believes advances in web search from Google and others will now take place incrementally, by squeezing a bit more from Google's Pagerank, or by tuning relevance, or indexing hard to find files. But for the next leap to happen, many forward thinkers believe a new architecture must be built.
Deep water search engines like WebSonar could be used to remedy this perplexing problem.

You can copy/paste the iReader content. Use the mouse pointer to select the text, then use Ctrl-C to copy it. Right click to get to context menu to copy from menu makes the preview disappear.

Let's see the summary from iReader about this article:

• Following by true graphical interface and last decade lot of people viewing that graphical interface least in part through browser
• Large part for legal reasons intended to keep Googlers at bay
• Throw wording that carry no real information words like
• Adding vectors in column yielding single vector representing column in multidimensional space defined words articles in database

Hmmm... I've seen executive summaries that are more informative. It looks like iReader needs a few more revisions to be useful.

iReader would probably be better incorporated into a search engine instead of as a client's web browser. (That is, if they get it to work). Imagine seeing your search results and iReader's summary instead of what Google now shows you - a few bits of a sentence surrounding the key word you were searching for.

Or maybe it could be a way to cull spam links. Right now, search engines use a mixture of links to a page or what words appear on a page. Both of these methods are vulnerable to spam. Imagine if a search engine generates an iReader summary, then uses that to evaluate the value of a webpage during a search. Nonsense verse used to generate links would fail to provide a good iReader generated summary. Such a page could then be tossed out of the search results.

I just saw an analogy between iReader idea and this video:
http://www.youtube.com/watch?v=lBvaHZIrt0o

Control Data Corp. Now there's a name I haven't heard in a long time. Even worked for them once.

Eh, I don't think this would be useful for tech savvy people, who know what to look for on the internet. It could, however, be useful for my mom (for example). Summaries of webpages could be a nice feature.

Check out OS X; they have has a feature in the services menu called summary, that sounds very much like iReader.

Output for generated for this article - judge for yourself... :)

# Following by true graphical interface and last decade lot of people viewing that graphical interface least in part through browser
# Large part for legal reasons intended to keep Googlers at bay
# Throw wording that carry no real information words like
# Adding vectors in column yielding single vector representing column in multidimensional space defined words articles in database

Output for generated for this article - judge for yourself... :)

# Following by true graphical interface and last decade lot of people viewing that graphical interface least in part through browser
# Large part for legal reasons intended to keep Googlers at bay
# Throw wording that carry no real information words like
# Adding vectors in column yielding single vector representing column in multidimensional space defined words articles in database

Here's an example using http://www.google.com/firefox - Firefox default start page, "About Mozilla" link on that page:

* Global community dedicated to building free, open source software products and technologies

* Programmers, marketers, testers and advocates around world working to ensure that Web remaining open sharing public resource

* Open source software products and technologies offering free-of-charge to people everywhere over 40 languages

Pretty nice!

This sounds very similar to a technology Google acquired Orion Algorithm from Ori Allon. The release at that time said something like:

The results to the query are displayed immediately in the form of expanded text extracts, giving you the relevant information without having to go the website.

So far we havent seen Orion release, wonder what Google is doing with that algorithm !!