Metasearch engine Zuula released a big update last week, what it is calling “Zuula 2.0″. It may be 2.0 but it’s still noted as being in beta; you can try it at http://zuula.com/.
My biggest complaint about using metasearch engines is that you end up seeing the same results over and over again. I’m jazzed to see that Zuula is conquering that with a feature called “Dynamic Pruning.” Here’s how it works. You do a search and get a result page that shows tabs for results from a variety of search engines including Google, Bing, Exalead, Gigablast, and Mojeek. The first tab, Google, shows all the engine’s search results. (You can use the Preferences option on the right of the search result page to change the order of the tabs.) The second tab, Bing, shows only those results which didn’t appear in Google.
The hidden results can easily be revealed with a click, and Zuula keeps each individual result’s order number so you can easily get a sense of where you are in the search results.
This means I can do a general-term Web search and quickly browse across several search engines’ worth of results without seeing the same three or four sites over and over and over again. That is exciting.
If you can’t think of anything to search for, Zuula can help you out there too with real-time search suggestions. Start typing “strawberries” and Zuula kicks right in (what the heck is a Strawberry Shortcake Makeover?) As far as I can tell, though, this new feature works only on the Web property of Zuula’s search.
Zuula does have several types of information to search, including Web (of course) Blogs, News, Jobs, and a new category called Tags. Tags searches seven collections, including StumbleUpon, Delicious, and Blinklist.
As I noted I don’t do a lot of metasearching because I find it extremely repetitive, but with that Dynamic Pruning feature, Zuula may just change my mind…
Footage.net, which is a metasearch search engine for stock, archival, and news footage, announced yesterday a new release of its search form and a new industry directory.
The search engine, available at http://footage.net/, lets you do a keyword search for footage and specify whether you want online screeners and text results or just online screeners. You can also narrow your search results by category. I did a phrase search for roller coaster and got over 2000 results from about 20 suppliers, though when I narrowed that down to search results that had online screeners the number of results dropped to just over 250.
Search results include a brief description, supplier, and links to play (if there’s a screener) and e-mail for more information about the clip. One time I could not get a screener to load.
I actually decided to mention this site because of the related directories. There’s an archive directory at http://footage.net/ArchiveDirectory.aspx which lists archive video companies from ABCNEWS VideoSource to WPA Film Library. There’s a supplier archive which seems to echo the archive directory with a few more additions. The directory I really want to see, though, is the researchers directory, which is supposed to be at http://footage.net/ResearcherDirectory.aspx. I say “supposed to” because at the moment it is only
You might not think so but metasearch engines for stock footage are serious time sinks, especially if you restrict your results to online screeners and use weird keywords. The time lapse video of the melting chocolate Easter bunny is still freaking me out…
When I first got the e-mail about Stinky Teddy, I almost threw it away. I mean, Stinky Teddy? That could be anything. But I read a little more and I was reassured. Stinky Teddy is in fact the name of David Hardtke’s daughter’s beloved stuffed animal. David Hardtke took the name and adopted it for his search engine that collects buzz from all over the Web. Stinky Teddy, complete with cute logo, is available at http://www.stinkyteddy.com.
Stinky Teddy is what I guess you would call real time metasearch. Enter a keyword and you’ll get search results from Bing, Yahoo, VideoSearch, Twitter, Oneriot, and Collecta. These different search types are choosable from the front page via checkboxes; I did a search for superbowl.
Stinky Teddy provides the results stacked on top of each other; Web, News, Images, Twitter, Popular Links, etc. There’s also a “Buzz-O-Meter” that measure how fast tweets are being generated and posts added to OneRiot. My searches brought me discussions on Twitter, fairly current news, and overview Web sites for context. The videos didn’t seem particularly “real-time,” but that was the only spot that seemed weak.
If you don’t like what results are being gathered, you do have some options; a brief preferences page allows you to specify what kinds of resources are being searched, as well as whether you want adult content, where you want to get your “gossip streams” (Twitter, Collecta, and/or OneRiot.) There’s also an option to specify whether you want the gossip streams to influence your search results — I think this means that that concepts that appear in the gossip stream will influence the other search results. I tried turning this preference off and on (it defaults to on) and I found that it made more difference in the video search results than anywhere else — maybe it was my search terms.
Stinky Teddy gets points for the Buzz-O-Meter. It gets more points for not having any Google Web search resource in its search (not that Google is bad. It’s just that too many search tools start and stay with Google search, and don’t explore other sites and possibilities.) the site is planning to look at other things to add to its search lineup, and I’m looking forward to that. I like the search results here but would like to see more.
Can you stand another metasearch engine? Sure you can. So check out Search3, which launched last month at http://www.search3.com. (And I’m not even going to tell you it’s in beta because you probably know already.)
This metasearch engine has two sets of searches — Web and images. The Web search finds results on Google, Bing, and Twitter (?). Well, that’s the default, anyway. Each column of search results on the page (there are three columns; I can’t believe they did that and it still looks nice) has a pulldown menu that allows you to replace each column’s search results. Google, Bing, and Twitter are the ones initially visibile but you can have Yahoo or eBay as well.) I’m glad you have the option to change the search results columns; except for the most general searches I wouldn’t have much luck using Google and Twitter at the same time.
The image search finds results on Flickr, Google, and Bing. (In this case the only image search option not available is Yahoo.) thumbnails show screen shot or avatar previews, depending on what resource you’re looking at.
I don’t use metasearch engines that much but I liked this one for its speed (very, very fast results) and for the fact that it managed to get three columns’ worth of search results on one page without making me claw my eyes out. Clean and quick.
I covered Molu the Meta Search engine a couple of years ago, but recently got a note from the site letting me know that a lot of things had changed. When I covered it before I didn’t like it very much — it was in alpha, didn’t work in Firefox, didn’t get me a result count, and basically got on my nerves. Today it’s in beta and I like it a lot more.
Getting search results from other parts of the Web is as easy as clicking on other parts of the tree. I was less impressed with the image search results because they simply gives you thumbnails with absolutely no context (Where did this come from? When? What size is the original image?) but you do get four resources to search from. The video search had several search resources as well.
I went through some of the other categories, including news, books, and answers, and while some of the categories are great, with a lot of extras added to the results (like News and Books) the result pages of the multimedia categories need more information attached to them. Even a title page with the thumbnail would be helpful.
Aside from the lack of context on the multimedia results, and the fact that some of the category searches were rather slow, I liked Molu. I can see myself using it for Web and news metasearch, but I’d probably skip the multimedia. Worth a look.
There’s a new search engine that allows you to narrow your search by country and by languge before running a metasearch that gives you results from Google, Yahoo, MSN, and local news sources. Glearch is available at http://www.glearch.com/ .
The front page has a map from which you can pick a country, or you can use the search engine form underneath. The languages are separate so you can search for Chinese-language pages from Congo Web sites if you really want to. (You won’t FIND anything, but…) What comes after the country, language, and query word form is a nice touch: reference information. When you pick a country the page will load with an overview of that country (CIA World Factbook) along with additional pointer’s to the country’s business, tourism and travel, culture, etc.
I did a search for Canadian Web pages, in English, containing reference to the word economy. Glearch provided the top ten results from each of the three search engines in the first paragraphs, but it didn’t feel particularly Canadian. I mean, I had at least two results from Wikipedia, a result from CNN, and only a few results for demonstrably Canadian (that is, .ca) site.
The news tab was much better, with news from The Globe and Mail, Toronto Star, and Yahoo! Canada. But I was pretty disappointed in the Web search.
I tried again, only this time I searched for the word economy in Iceland Web pages (again in English.) This time I had better luck. The first page result was for Iceland.org, and there were several .is pages in the Web search results. However I didn’t get any NEWS results! Sigh.
The non-English results were even weirder. I did a search for Web pages from French sites, in German, with the word gehts. This time I got at least two pages with the Germany country code (DE) and one with the Austrian country code (AT). Honestly, I don’t understand how Glearch is delineating these sites.
Here’s the thing: when the results were solid (like with the Iceland search) they were SOLID. There’s enough non-overlap of search engine content that I appreciated getting results from three different engines. And I liked the reference materials at the bottom of the search pages. But sometimes the results were all over the map, to the point where I couldn’t even figure out what the search engine was doing.
I think you could get some immediate use out of the quick reference on the site, but expect to do some experimenting with the search results.
Icerocket recently launched a “big buzz” search that allows you to pull keyword search results from several different sources onto one page, with the option to refresh as often as every minute. I like that. What I REALLY like, though, is the fact that you can incorporate your query into an easy-to-use URL.
The URL syntax is http://keyword.icerocket.com . If you want to use multiple keywords separate them with dashes: http://many-keywords.icerocket.com (Blake at Icerocket tells me they haven’t gotten to integrating quotes yet.) For example, say you want to keep up with what Tim Geithner is up to. You could use this URL:
Icerocket will return a page with results from all over the social Web about Mr. Geithner. From top to bottom on this search results page I got results from Icerocket’s blog search, Twitter, FriendFeed, Icerocket’s News Search, and Flickr. (In other searches I ran there was a group of video results between Friendfeed and News.) An option on the right side of the page allows you to set an automatic refresh rate (the default is no refresh) and there’s a nav at the bottom for multiple pages of results.
There is a tremendous amount of information on this one page. All the text-based results are tagged with how recently they occurred, and the Twitter listings include tiny little icons and extensive extracts (often the entire Tweet!) The multimedia results don’t have quite as much information, but you can at least mouseover the screen thumbnails for item titles. The video results do show how recently the video was added.
I knew this would work well for words and names but I wanted to try something which I figured would be a little more obscure, so I plugged in:
#cil2009, as you might know, is the hashtag for the Computers in Libraries 2009 conference, which is happening RIGHT THIS SECOND.
I was pleasantly surprised to see that the page I got was not limited to Twitter. I also got results from Friendfeed, blogs, and even images (though a couple of the images didn’t look like they had anything to do with CIL2009.)
Using stock symbols brought more mixed results. Searching for http://bgp.icerocket.com/ brought results that had nothing to do with bookselling company Borders, but http://bgp-borders.icerocket.com/ brought incredibly focused results with a business-y tone (though no images or video.)
Actually slanting main keywords with secondary keywords could keep me busy for hours; http://shaq.icerocket.com/ is one kind of search result, while http://shaq-twitter.icerocket.com/ is COMPLETELY different. http://used-books.icerocket.com/ leads to cozy discussions about reading lists and recent purchases, while http://used-books-cpsia.icerocket.com/” leads to complaining about federal legislation.
And the thing is, even with these tweaks, you still have a good URL. It’s readable, relatively short (you could go nuts with the queries, of course, but you don’t have to) and it’s easy to pass on.
I like this. I like this a lot. Well done Icerocket.