ResearchBuzz

Google is now offering several different alternate views for its search engine. You can see the announcement here, but you actually have to check out the search views at http://www.google.com/experimental/index.html. You’ll notice this is a Google Labs page. There are several “experiments” you can join here; the Alternate Views is the first one. (You can only join one at a time — bummer.)

Anyway, there are several views you can choose from. Pick a query that’s provided on the experiments page. When you choose one you’ll go to a result page that has several tabs of result types to choose from. The first tab, the “List View” appears to be the standard search result view — a vertical list of results occasionally with a screenshot or image.

The second tab, the “Info View” tab, looks like the List View but with a twist on the right side of the results. Instead of the standard snippets for each result (which usually contain the search term and some context), you can specify if you want to see dates, measurements, locations, or images. You can also filter the first three options — choosing to see only the year 1971, for example, or only the location Houston.

For the search terms I tried, I didn’t get much luck with anything but images. I had to make up a term that got plenty of results for all the Info View options (first man on the moon). Google did pretty well detecting the measurements and dates, but occasionally got it wrong (The date “31″ and the word “Month” together was taken as a measurement; TV shows like 48 Hours and 60 Minutes were not recognized as such.) It wasn’t enough to make the view unusable.

The Timeline View pulls date information out of results and presents a timeline at the top of the results page. Click on it and your search results will be limited to the timespan you chose. Even with my first man on the moon, search, I found the timeline was pretty skimpy. Doing a search for World War II presented a much more populated, albeit narrowly-focused, timeline.

The last item on the tabs was to me the most impressive. The “Map View” presents the results on a Google Map. Your results are listed on the left and then marked on a map on the right. Click on a map marker and you’ll get a list of Web pages that refer to that location. So cool! You can also filter your results with an additional location so you get an even smaller map. For example, you could search for “Moon exploration” and you’ll get a map of the US. You can narrow your Map View by adding Houston and you’ll get a map of Houston with search results taking you all the way down to street addresses.

(Unfortunately it looks like the Map View is limited to Earth. I did a search for “Moon exploration” to see if I might get a map of the moon instead, and I didn’t.)

Though I remain surprised that Google doesn’t do more with its standard search results (like introduce JavaScript-based page previews and the ability to filter by content like Flash and RSS feeds, I do like what’s happening here. Now if it can just do a bit more with the plain vanilla List View…

Wondering what people are passing around at Google Reader? ReadBurner, at http://www.readburner.com/, is here to help. The site monitors “several hundred” linkblogs (there’s a link to show them, but it gave me a 404 error) and filters for the ones that are shared by multiple people on Google Reader.

You can review currently-popular items, look at “upcoming” items (ones that are getting hot, I guess), things that are popular this week, and things that are popular over all time. (RSS feeds are available for everything. Happy sigh.) You can get lists in several different languages, in several different levels of detail (from nothing but a headline and a link to getting the complete item within the list (very cool.) Even just the basic text and summary shows a lot of details.

The items listed here tend to be very tech-heavy, but I got no problem with that. If you’re less interested in the items themselves and more in the people behind them, you can check out the top linkbloggers, sources, and authors. (Cory Doctorow, what a surprise! :->) Clicking on a name gives you a list of what they’ve linked/generated/contributed.

From What’s on the blog, it looks like this service is fairly new, but I like this. Absolutely chunking a few of these in my feed reader. Check it out!

Hey, it’s not in beta — THAT’S different from Google. Wikia, the much-buzzed Google competitor, finally launched in alpha at the beginning of the week. Unfortunately, with the model that’s being used, the suck is built right in. The only thing that’s going to tell with this engine is time.

Instead of Wikia.com, the new search engine is available at http://alpha.search.wikia.com/ . If you go to the about page, you get an overview of the search engine and also the admission, “We are aware that the quality of the search results is low.” Um, yeah.

I started with a search for Hawaii. It’s a good example search because you can get all kinds of results but you can reasonably expect an official result as the top one. In Wikia’s case the first result is for The University of Hawaii’s athletic page. Not ideal but not awful. Awful is the fact that the third result had a very keyword-loaded description (and didn’t appear official at all), and the official state site comes in only after that. For experimental purposes, I also tried the “Strawberry Shortcake” test. The front page of the results appeared to be spam-free, but the ratio of 80s-cartoon-character-to-recipes was very high. (Using lowercase instead of capitalized in the search didn’t make any difference — I don’t think Wikia is case-sensitive.)

The idea with Wikia is that we, the users, are to give feedback on the search results we get — rating them 1 to 5 stars. With that in mind there’s a little set of stars next to each result and we’re encouraged to rank what we see. (When I tried to rank I got a popup that the stars were only there for testing and were not being kept permanently. When I tried it again a little later I didn’t get that popup. So I don’t know.) The theory is that eventually, given enough people and enough results, the search results will improve. Unfortunately, it’s just launched, there’s no huge amount of feedback driving results, and thus the suckage being built in and all.

Even if I were assured that my ranking counts, I would still be ambivalent about ranking results. Yes, hinky sites get one star. Official sites, for the most part, get five stars. But what about the ones in between? Every site I looked at to rank I kept thinking, “It depends.” It depends on why the a searcher was looking for “Hawaii”. It depends on why they were looking for Strawberry Shortcake. In addition to ranking individual results, Wikia invites you to create miniarticles to create definitions, clear up ambiguities, etc. but I don’t understand why you want to duplicate the effort of Wikipedia. Now, if you invited me to tag Wikipedia pages relevant to my search, I’d be all over that.

In addition to getting Web search results, you’re also given lists of Wikia users who match your search. For example, if I do a search for cow I get one user under the “People Matching “cow”" result list. But I get no information about that person, and I don’t get any reason as to why they match my search. I’m afraid I don’t understand the point.

I’m afraid Wikia left me not so much cold as bewildered. I didn’t understand the extra miniarticles, I didn’t know why Wikipedia wasn’t more closely integrated — I mean, so many people are working on it, why not do some crossreferencing? — I didn’t know why I got one user result for cow, I didn’t know if my rankings were accepted or not, I didn’t know what evaluation process I should be performing to do the rankings in the first place — eventually it got all philosophical and I began wondering why we were all here to begin with. (Thankfully this is ResearchBuzz, not ExistentialismBuzz.)

I’m going to go back and work with it more later in an attempt to Get It.