ResearchBuzz

Microsoft has announced that their search engine, Live Search, has moved out of beta along with their Live Local Search mapping service. I am still trying to wrap my head around why they went with the “Live” brand. I don’t know what it means. Live as in real-time? Live as in not-dead? I think we can assume that. Otherwise out there somewhere we’d have “Zombie Search”, a full-text search engine that for any query returns the answer “BRAINS! BRAINS! BRAAAAAIIIINNNNSSSSS…”

Ahem. The Live search is available at http://www.live.com/ , natch, and at the moment offers several search tabs including images, news, and “local”, which is the mapping service. There’s also a new Q&A service, which is short for another Answers service. I think I like this one better than Yahoo Answers, but not as well as Google Answers or Ask Metafilter. I found the Live news search a little annoying; in my tests I found two-year-old content indexed with timestamps of a week ago.

Strangely enough, the stuff I liked at Live.com was on the “More” tab of the search results page. The feeds search was pretty good, though I’d appreciate it if there were a plain link to the RSS feed of a site (instead of an “Add to Live.com” link only.) The Academic search, which provides search results from both free and paid-access content, has extensive results in an interesting format (result list on the left, mouseover interesting ones to see additional details like abstract, journal name, authors, publisher, etc. There’s also an intriguing “Search Macro” offering which allows you to create your own search engine, either by creating uber-queries or by specifying up to thirty sites for Live.com to search (ONLY 30??!! PLBBBTT!) So why’d they shove all the interesting stuff in the More tab?

If you’d like to keep up with the changes and updates to the new Live search, you can visit the blog at http://livesearch.spaces.live.com/feed.rss . There’s also a blog for the Local Search mapping service at http://virtualearth.spaces.live.com/ (take a few minutes and read that blog — this is a developer team which is feeling its collective Cheerios.)

You know about YouTube and all those other videos that host generated content and allow search. ClipBlast is a little different. It aggregates content from sites that are not as large as YouTube and other such sources but carry video content — sites like CBS, Today Show, CNBC, and Expanded Books. It’s available at http://www.clipblast.com/ . (Actually looking at it I see that ClipBlast does index content from YouTube, but they’re pretty picky. They have less than 4,000 clips from YouTube indexed.)

The site starts of with a “Real-time Index” of newly added content which lists small thumbnails of content along with a description, source, and category. There are links to view the clip (on the originating site) share it, save it to a “My ClipBlast” area, and share it via an e-mail message with a link. There’s also a details page for each clip, but it doesn’t do much more than provide an expanded description. It doesn’t provide things I would expect from a clip page, like clip length and file type. (As ClipBlast doesn’t host the content it indexes it may have a hard time getting clip length.)

In addition to looking at clips as they’re added, you can also view a list of editor’s picks, popular clips, clips by category, and clips by provider. A search engine allows keyword search; a search for space shuttle found over 1500 clips from providers including CNN, BBC, CBS, and Reuters Video. If you want to see the second page of results, you don’t click on “next” or a page number or something like that; you click on “older”. To see the last page of results you click on “oldest.”

If you want to save clips to the My ClipBlast tab, you’ll need to register. Registration is nice and basic (user name, password, and e-mail.) While I appreciate statements like “ClipBlast will never sell your e-mail to anyone. Period.” I would like a more overt privacy policy statement please. The one above leaves too many questions, like, “So you won’t sell it but maybe you’ll give e-mail address lists to business partners?”

ClipBlast does have a blog ( http://www.clipblast.com/blog/ ) but it seems to be more of a pointer to video clips than information about updates to the site or new technology. I like ClipBlast, and I like the idea of a site aggregating content from sites that are not primarily video oriented, but there’s more I’d like to see here: more details on the clip detail pages, an overt privacy policy, and maybe more information and updates about ClipBlast’s technology and offerings.

Google has had news search for a long time, but its archives are only 30 days deep. Now Google has added an archive search to the news mix. While it’s always good to have deep archives, and it’s really nice for genealogy searches (more about that in a minute), there were a few things about it that drove me crazy.

Let’s start with the happy stuff. The archive search is located at http://news.google.com/archivesearch . There’s a basic keyword search and there’s an advanced search that allows you to search by date or source as well as restrict your results to free content or paid content restricted to a certain level.

The search is supposed to go back 200 years — to get a search that goes back a good long way try buggy whip. (”Buggy whip? Buggy whip? Three buggy whips.”) Results are listed by — relevance, I guess. Sort your results by date using the “Show Timeline” feature that breaks your results out with text marking years or decades. Within each of these sections news stories are listed oldest to newest, which as you might realize is backwards from the date organizing done in the regular Google News.

Search results include the title of the item, publication name and archive name, cost to access and date of publication, and a small snippet of context. There’s also a “Related Web Pages” link. Try that when you’re looking at a story that’s going to cost you money to access. I found about 25% of the time when I tried it I could access the story for free online via a Google search.

Clicking on the title from the archive search results either takes you straight to the result (if it’s a free story) or to the individual archive’s payment/information page (if the story costs money.) I found that the archive pages tended to have more extensive article snippets. In some cases, like with obituaries, the whole story really was available for free.

Speaking of obituaries, I found that this archive search feature was really good for genealogy. It’s an easy way to search a lot of obituary archives at one time, and for some reason (I thought it was a NEWS archive) this search also finds information on WWI registration cards from Ancestry.com. The search is better — cleaner and faster — than the actual search at Ancestry. In fact this feature got me rather sidetracked…

But there are some things I don’t like about Google’s new News Archive. First is the fact that it seems so separated from the regular Google News. If you go to Google News and do a search that gets no results, does Google News recommend you search the archive? NO! Why not? 30 days is not that long. Further, the fact that the date sorting is one way in Google News, and backwards in the archive search, is going to get frustrating over time. How about newest-to-oldest both places?

I also missed the advanced options available in Google News when trying the Archive Search. Mostly I missed the location: syntax, which would have allowed me to narrow my search results by state or country. That would have made the genealogy searching even better! Perhaps having old stories aggregated into archives makes separating stories by location more difficult.

Don’t get me wrong; I really like this archive search. I can just think of a few ways I’d like it more, and I’m really surprised how isolated it feels from the established Google News.