a web of your own
Google hasn't moved it's Personalized Searching feature out of beta yet, but I've been using it for awhile and it looks like the beginning of a good idea. Enter in a profile of things that interest you, fairly general categories like 'gardening' or 'electronics' - these will be stored in a cookie. Then search as you normally would, and Google returns a page of results that is very similar to what the normal Google search would give, but at the top of the page there is a slider bar that lets you control how personalized the results are. As you slide the bar over to the right, you see some of the search results move lower in the list, or even fall right off the page. Crank the thing all the way over to 'max', and the results that remain are a combination of the search terms you entered and the profile you created, and should therefore be (theoretically) more useful to you.
I should *DO* this next thing, instead of pontificating about it - I think that this sort of personalized searching is the next killer app. Google is taking the first step in the right direction, but there are two factors working against them: the large number of people using their service, and the extent to which checkboxes in a number of categories can capture the essence of a person's interests... there's just too much to store in a little cookie.
Sci-fi authors have long envisioned a future where the technology that surrounds has become very personalized - the morning paper only shows you items that you will find to be of interest, your web browser spiders the web collating useful links in the background while you sleep, and any advertisements you see are for products or services that you might actually need. Amazon.com has been working on that last bit for awhile: based on your previous purchases, it often recommends other things that might interest you. I spent an hour once going through the recommendations at Amazon, clicking 'Not Interested' on anything that didn't apply to me, and since then I rarely see a recommendation that misses the mark.
RSS aggregators have already made some progress into personalizing the web... you decide which feeds to subscribe to, which is a good first-order winnowing of the data, though not quite as useful as deciding what sort of information you wish to see - the RSS model is more of a trust model, in that you trust the feed source to deliver the type of information you would like to see.
I've thought quite a bit about a desktop app that would scour the web, storing interesting articles for you. When you read an article that interests you, you could give it a higher rating, and similarly less interesting articles would receive a lower rating. These ratings would then further refine future searches. I don't think it would take long before the system was proficient in rerturning only those results that interested you.
Of course, one of my greatest interests lies in learning, in being exposed to new ideas, so I suppose the software would have to randomly insert results from outside of my declared (and computed) areas of interest. It would still be able to use previous knowledge to weed out pages that were high in advertisements or annoyances (like injudicious and excessive use of the tags).
I should *DO* this next thing, instead of pontificating about it - I think that this sort of personalized searching is the next killer app. Google is taking the first step in the right direction, but there are two factors working against them: the large number of people using their service, and the extent to which checkboxes in a number of categories can capture the essence of a person's interests... there's just too much to store in a little cookie.
Sci-fi authors have long envisioned a future where the technology that surrounds has become very personalized - the morning paper only shows you items that you will find to be of interest, your web browser spiders the web collating useful links in the background while you sleep, and any advertisements you see are for products or services that you might actually need. Amazon.com has been working on that last bit for awhile: based on your previous purchases, it often recommends other things that might interest you. I spent an hour once going through the recommendations at Amazon, clicking 'Not Interested' on anything that didn't apply to me, and since then I rarely see a recommendation that misses the mark.
RSS aggregators have already made some progress into personalizing the web... you decide which feeds to subscribe to, which is a good first-order winnowing of the data, though not quite as useful as deciding what sort of information you wish to see - the RSS model is more of a trust model, in that you trust the feed source to deliver the type of information you would like to see.
I've thought quite a bit about a desktop app that would scour the web, storing interesting articles for you. When you read an article that interests you, you could give it a higher rating, and similarly less interesting articles would receive a lower rating. These ratings would then further refine future searches. I don't think it would take long before the system was proficient in rerturning only those results that interested you.
Of course, one of my greatest interests lies in learning, in being exposed to new ideas, so I suppose the software would have to randomly insert results from outside of my declared (and computed) areas of interest. It would still be able to use previous knowledge to weed out pages that were high in advertisements or annoyances (like injudicious and excessive use of the tags).
1 Comments:
-stan
return to front page