Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Search Still Sucks

techcrunch.com

Post reply on HN
Print view View original

41–46 of 46 posts

Re: Search Still Sucks

#41

Post by pasbesoin »

vanessafox wrote: post #38
pasbesoin wrote: post #35

If "the customer's always right", then, for the current search situation, you have to ask: Who's the customer? -- P.S. Google: I'd accept, gratefully, as a stopgap, being able to specify multiple "site:foo" parameters. I don't know what the computational (or financial) impact to you would be. (So far, I've held off trying browser extensions to accomplish this, but I guess I can start looking for one having a good rep…

If you want this ability for one-off searches, you can use the OR operator. site: http://news.ycombinator.com OR site:stackoverflow.com I don't know if there's a max # of sites you can add. I tried it with three and it seemed to work great. http://www.google.com/search?q=site:searchengineland.com+OR+... (Yes, I did an ego search. Isn't that the best way to know if the results are likely accurate? :) If there's a set…

I've tried that [before], but with multiple parameters, I simply got zero results. I then enclosed the site:foo OR site:bar within parentheses, leaving the rest of the query outside parentheses. Whichever site:foo came first would be matched, but only it appeared to be matched. Subsequent site:bar, etc. parameters did not produce any matches in the results.

I'll have a look at your links, in a minute -- probably should have done that first.

EDIT: Well, your second link sure works. I wonder what I was doing wrong. When I saw someone else('s or s') comments online to the effect that it didn't work, I assumed that was why my own experience failed.

I'm in serious need of some coffee, at the moment. Thanks for correcting my mis-impression; I look forward to straightening it out in my mind and taking advantage of this feature.

Re: Search Still Sucks

#42

Post by pessimist »

noonespecial wrote: post #19

Travel searches, for example, are a joke, and startups like Gogobot are popping up to try to fix that. This is how Google is finally toppled. Not by a "google killer" or another giant like Bing, but by a thousand tiny niche services that get "X" search right.

This is true, but it has always been true. Google has always sucked for product review searches, local service vendors (try finding a plumber anyone?), travel searches etc. The odd thing is why people continue to use it for all these things. Perhaps the ads?

Re: Search Still Sucks

#43

Post by vanessafox »

pasbesoin wrote: post #41
vanessafox wrote: post #38

Earlier quoted context omitted.

If you want this ability for one-off searches, you can use the OR operator. site: http://news.ycombinator.com OR site:stackoverflow.com I don't know if there's a max # of sites you can add. I tried it with three and it seemed to work great. http://www.google.com/search?q=site:searchengineland.com+OR+... (Yes, I did an ego search. Isn't that the best way to know if the results are likely accurate? :) If there's a set…

I've tried that [before], but with multiple parameters, I simply got zero results. I then enclosed the site:foo OR site:bar within parentheses, leaving the rest of the query outside parentheses. Whichever site:foo came first would be matched, but only it appeared to be matched. Subsequent site:bar, etc. parameters did not produce any matches in the results. I'll have a look at your links, in a minute -- probably shou…

It could be that you were using a lowercase or. It has to be all caps OR.

(lowercase or just uses "or" as one of the search terms, which I think actually is considered a stop word and is then ignored.)

And I'm always a supporter of the more coffee recommendation.

Re: Search Still Sucks

#44

Post by chris_j »

tomkarlo wrote: post #37
chris_j wrote: post #9

Earlier quoted context omitted.

I know what you mean about searching for local vendors or buying something; my experience is that a Google search isn't great for either of those. I still find Google pretty useful for coding help, though, and personally I don't tend to go straight to somewhere like Stack Overflow. Could you give an example or two where Stack Overflow gets you an answer more quickly than a Google search? Here's my example: In gdb, I…

Mea culpa. SO was definitely my weak example there... if they had a better on-site search it wouldn't be an issue but I guess they mostly just optimize for Google search. So a lot of the time I end up Googling tech issues, but then clicking on the SO link because I know there will be multiple answers on that page instead of just one.

No worries, I also notice that SO comes up in Google results quite a lot and that's generally how I access it as well so I also do a Google search for tech issues and then end up clicking on the SO link.

Your post yesterday got me thinking about why Google is still pretty useful for coding questions but not so for searches of a more commercial nature. I guess it isn't worth so much money to game Google for the former (Experts Exchange notwithstanding).

Re: Search Still Sucks

#45

Post by photon_off »

mooism2 wrote: post #8
photon_off wrote: post #6

Search engines would benefit enormously from using data from social bookmarking sites. If you haven't lately used Delicious to search, I highly recommend it. The breadth of search isn't as far-reaching, but all of the results have essentially been pre-filtered by however many users have bookmarked it -- it's unlikely that many people would bookmark a spammy site or site with crap content. I think Google results could…

If black hat seos started to believe Google was paying specific attention to Delicious bookmarks, Delicious would be flooded with fake accounts faster than you can say "link farm". History of search: 1. New company releases new search engine, better than the competition because it pays attention to a previously ignored signal. 2. Seos work out how to game the signal. 3. New search engine is now as crappy as the searc…

You're quite right. But pretty much every strategy a search engine uses is susceptible to black hat SEO. I think there's something to be said about having the system use how others are reacting to your comments rather than just what your content is and who links to it. Yes, you could farm out Delicious accounts and have them bookmark your site, but from a SE perspective it's easier to remedy this problem (Delicious requiring captchas periodically, looking for accounts which have an uncanny amount of similar bookmarks in a span of time, etc) than it is to come up with strange rules about content that the black hat is free to change.

Judging by the votes, apparently it is more valuable to point out why a certain idea might not work, even if all of the alternatives have the same problem, than it is to suggest something with exciting possibilities. Nothing against you, personally -- your comment is quite right and I agree gaming is a major concern.

Re: Search Still Sucks

#46

Post by photon_off »

vanessafox wrote: post #39
photon_off wrote: post #6

Search engines would benefit enormously from using data from social bookmarking sites. If you haven't lately used Delicious to search, I highly recommend it. The breadth of search isn't as far-reaching, but all of the results have essentially been pre-filtered by however many users have bookmarked it -- it's unlikely that many people would bookmark a spammy site or site with crap content. I think Google results could…

There has been a bunch of research on use of delicious tags in particular. See for instance: http://heymann.stanford.edu/improvewebsearch.html If I recall correctly, Yahoo tested using this data a few years ago and found the signals not to be as useful as others they used at the time. Of course, Bing and Google have been working to include more social signals in rankings: http://searchengineland.com/what-social-signa…

Really cool study, thanks for sharing. I don't think the research proves that social media wouldn't be a worthwhile signal. They really don't look into how bookmarking data could be applied to results, they just look at how bookmarked sites relate to search results and the web at large. The strongest conclusion is that Delicious users only bookmark a sliver of the web's content, and while the researchers considers this a weakness. I would consider it a strength.

The way I see it, search engines are entirely inclusive, while bookmarking sites are selectively inclusive. For certain ultra-specific queries, I'd much prefer the search engine approach -- I want as much breadth as possible. For other not-so-specific queries, I'd much prefer the result sites to be vouched for by people. Looking one step further, probably for a decent chunked size of my search queries, I'm OK if the number of sites it is searching over is only about 10,000,000 (my estimate for how many sites on Delicious have been bookmarked by more than 20 users).

What I find exciting is that the two can be effectively combined. Take all the normal results that a search engine would show, but give it a boost based on the log of how many times it's been bookmarked. It really is that simple.

Post reply on HN