Tuesday, September 18, 2007

Semantic Language Search....

I stumbled upon Powerset while reading a post on the New York Times blog Bits. Naturally I had to check out the Powerset Blog where I was delighted to read posts from Marti Hearst, an associate professor in the School of Information at UC-Berkeley, and Doug Cutrell, a Powerset engineer.

Language has always been fascinating to me, and one of the things that has been frustrating with search engines is having to think through a series of search terms in order to find useful information. I got a crash course in search terms when I worked in online marketing, and after you do it enough, it becomes second nature. You start to wonder what is wrong with people who can't seem to find useful information on the Web. One word searches are out! It's the two, three, sometimes four-word phrases that bring up the results you want. The trick is to find the right combination, or search phrase.

Now working as a communications consultant, language has taken on a different meaning. Writing procedures requires being exact in language, directing the end user through a specific set of steps to achieve a specific end result. A systematic execution, if you will. Search, if you think about it, is random. You have an idea of what you're looking for, and type in words you think apply to the idea and continue doing so until you find useful information. Most people call that trial and error. With practice comes improvement, but in the end, it is still trial and error.

Powerset seems to me to be taking a more systematic execution approach to search, first by removing the guesswork involved and second by appealing to the natural use of language, as in full statements, questions, etc. instead of a random mix of keywords that may or may not produce the desire result.

Makes me wonder if the days of audio search, where I will be dictating a procedure and have to search for something online simply by speaking the phrase or sentence, is closer than originally thought.

Monday, September 10, 2007

Google Book Sharing Not the First

There's this post on the Google Blog about enhancements to its book search, namely the ability to create a "virtual bookshelf." That's all fine and dandy, and it being Google, undoubtedly the searchability is good.

But this isn't the first "virtual bookshelf" that you can share with others, and I doubt it will be the last. There is another website called Sheflari that lets you find, collect and share books in a "virtual bookshelf," and exchange comments, reviews and suggestions with other users. You can embed your Shelfari into your website, blog or online profile as well.

I've been using Shelfari to being a digital catalog of all the books I have read. It's turned out to be a bit of an undertaking as I have lists strewn about of the 100s of books I have read, and a journal of sorts I received as a gift to chronicle the books I have read, favorite passages, books I want to read, etc. There is something more gratifying about updating my Shelfari page and then seeing who else has read the same book, what they have to say about it and making a connection.

Google's addition to its Book Search has the same potential, and is yet another example of Google knowing anything and everything about you. Maybe you don't mind, but I have second thoughts about putting all my eggs, as it were, in one basket. That, and there isn't the information overload with Shelfari as there is with Google Books.

Wednesday, September 5, 2007

Time, Money and Resources

Google and the American Blind and Wallpaper Factory have settled the trademark infringement lawsuit. And then there is the article about news wires like the AP agreeing to license news feeds to Google.

All of this fits well with an article in The Economist: Who is Afraid of Google?

The answer seems to be no one. Google has demonstrated, on more than one occasion, it has enough time, money and resources to fight any and every lawsuit hurled at it. Google also has the time, money and resources to impose its will on governmental agencies, and play an increasingly larger role in the shaping of public policy to fit its own agenda, only under the guise of improving information flow for the masses.

There is its post on patent reform, its lengthy posts on the telecom spectrum, all of which sounds very much like running commentary on governmental policies that pertain to technology, except that the commentary is coming from a source that millions of people blindly trust.

This issue came up on the "L" ride home from class last night, and at a dinner party I attended over the weekend. My generation and younger have no qualms about freely posting information about ourselves to the Internet, or freely sharing information via searches with engines like Google. And, if you think about it, if you use any of Google's free services, Google has even more information about you. I have yet to read somewhere that says something along the lines of Google employees not being able to read your documents or your email or access Google Checkout account information.

Makes you wonder exactly what kind of security, if any, they use to monitor employees access to customer information, as it were. For all I know, a Google employee has been joyfully reading my email, and checking every so often to read the next chapter of my novel.

What it boils down to is that people in my generation and younger have decided to make a trade off: less control over personal information for convenient accessibility anywhere. Using something like Google Docs & Spreadsheets, or ZoHo (which actually has a much better interface, more options and is more user-friendly) gives me the convince of being able to work from any computer with a browser and an Internet connection, but in return, I have to provide what amounts to personal information to which ever online document client I choose.

With all the news in recent weeks about databases being hacked, such as Monster and Loyola University-Chicago, I wonder if a few of these aren't warm-ups, testing hacks to get at the wealth of information stored in search engines. Some might say no place is safe, and that it is only a matter of time.

Which makes you wonder, would search engines notify you of the possibility of compromised information, or are we really better off using Google Docs & Spreadsheets just to keep notes or reminders, and not to keep anything useful or important? Afterall, with Microsoft Word documents, you're major security concern is just your laptop, which is more often than not completely in your possession.