4 ms·
I think you nailed it...today, many sites blur the line between "spam" and "content". Ridiculous viagra ads are obvious spam, and Wikipedia is obvious content -
by LiveTheDream 17y ago
I think you nailed it...today, many sites blur the line between "spam" and "content". Ridiculous viagra ads are obvious spam, and Wikipedia is obvious content - but what about eHow? I may very well find a good answer on that site, but the site is littered with ads, the quality is usually terrible because it is mass-produced with quantity in mind instead of quality, and lots of answers are basically duplicates because fresh content is good for seo. What about a site whose only content is "top X" lists, or one who ranks high but the actual content is hidden behind a paywall?
That blurry distinction between useful content and spammy content is tough for an algorithm (and a person) to sort out.
- sunkencity 17y agoI suspect that what they do is that they game the algorithms so that they can insert their hit before the actual content, the way it's weighted currently makes it possible to get sites with links to content to rank higher than the actual content, which is wanted in some ways (like the HN discussion is an important complement to what it links to) but with the spam sites it's just littering down the interwebs.
- _delirium 17y agoYeah, these sort-of-but-not-really-content sites have really increased the cognitive load of using the internet. Of course, internet content was always of variable quality, but up until the late 1990s, maybe early 2000s, most random pages on some subject were written by someone who actually had some sort of interest in it. Given a geocities page on some random subject, it might well be crazy, but it was usually in good faith: the person really was saying it in their own voice, because they wanted to, not because they were being paid some pennies per paragraph to write it. The rise of a few specific sites with decent, non-spammy info is the only thing that's really saved its usefulness for me. Wikipedia alone accounts for a huge proportion of the times when a Google result is actually a good one. And for travel information, I go directly to a few sites (mainly Wikitravel) and don't even bother with Google.