Wolfram|Alpha Celebrates First Anniversary with Some New Features

Happy birthday, dear Wolfram|Alphaaaaaaa…. happy birthday to youuuu….. Search engine Wolfram|Alpha put up an interesting blog post Tuesday about its first anniversary and the way it has changed over the last year. The search engine also announced a few changes.

The home page is a bit different, pretty but still simple. If you’ve never quite “gotten” W|A, check out the examples by topic, so you can get an idea of what Wolfram|Alpha can do. If you really want to get under the hood, check out the still-incomplete entity index, which shows you very specific examples of what W|A covers in different categories. (This is still under development but it’s fascinating and I can’t wait to see how it fills out.) The home page also has settings now, too, though it’s just for background settings (the blue one is nice) and whether W|A shows hints or not. Looks like it relies on cookies to keep these settings.

There’s also some new content; the site now offers street maps; searching for something like Sydney Opera House shows, in addition to information about the structure itself, a street map to where the structure is located. There’s also several ways to search for diseases — pulling up that URL will let you calculate disease risk, look at the incidence of disease in populations, get information on specific diseases, and more. I did find that I had to play with my searches a bit to get some of these results. And of course I knew a long time ago that the phrase random disease works.

W|A also announced that when the search engine doesn’t know the answer to a question, it’ll will try to find the “nearest” query to interpret. It doesn’t work all the time, but W|A is working on making this better. I’ll need it, because I’m still not great at figuring out Wolfram|Alpha’s syntax sometimes, though I find myself using it more and more.

In fact, I’m using it so much that I find myself actually looking for a couple of features, though neither one of them is probably what W|A is made for. First of all is an expansion of random words. You can search W|A for random word and get a word with definitions, synonyms, etc. But though the definitions include the parts of speech, you can’t search for, say, random noun. I wish you could; it would be a handy tool for Mad Libs or generating random queries for Flickr. You also can’t stack random queries, either, which is a shame. Wouldn’t it be a great creativity tool for writers if you could run the query random first name, random surname, random occupation, random city and get all the answers on one page?

Happy birthday, Wolfram|Alpha. You’re not getting as much attention as you probably deserve, but it hasn’t stopped you from evolving in new and useful ways. Keep it up!

Ever Wondered? Site Helps You Answer “Life’s Little Mysteries”

If you have kids or like “Ripley’s Believe it Or Not” You’ll enjoy Life’s Little Mysteries, a Web site that answers all those random questions that you constantly wonder about. It’s available at http://www.lifeslittlemysteries.com/. It’s a terrific reference site, though it does in my opinion suffer from a lack of sourcing.

Life’s Little Mysteries is sort of like a question engine but human-powered, with extensive articles to answer the questions. So the data pool isn’t particularly wide, but it is pretty deep. The front page divides its answered questions into sections, like Body & Mind, Animals, and Just Plain Strange. Ancient mysteries (“Do Fish Sleep?”) exist side-by-side with more contemporary questions (“How Are Oil Spills Cleaned?”)

There is a keyword search available; a search for baseball found six results, including “Why Is Baseball Spring Training in Both Florida and Arizona?” and “Are Left-Handed People Smarter?” (the keyword search is a full-text search, so you will find questions that are far afield.) Questions are answered by articles that range from a couple hundred to several hundred words; terrific detail. I was surprised to see that while many articles had links, some that didn’t didn’t offer much sourcing. For example, “Why Is Baseball Spring Training in Both Florida and Arizona?” offers several quotes and dates, none of them are sourced in the article that I can see.

I bring this up because of the quality of the site wranglers, which you can read about here. Several of the staffers have journalism or science degrees, and that’s one of the reasons I decided to cover this resource. So I know research has been done to create these articles, so why not source it?

Though the site has just launched, it looks like there have been articles added since at least March, so there’s plenty to see here. I just wish there was more bibliography available.

Embedding Tweets With Blackbird Pie: Why?

I will be the first to admit that sometimes I don’t “get” stuff that I come across on the Internet. And such it is with Twitter’s new offering, Blackbird Pie. The whole time I was trying it, I was thinking, “And this is better than a screen shot how?”

You can get an overview of Blackbird Pie at http://media.twitter.com/411/fresh-baked-tweets and the tool itself at http://media.twitter.com/blackbird-pie/. Briefly:

You might find an interesting quote on Twitter that you want to, well, quote. You can quote it the regular way, with, like, quotation marks, or you can take a screen shot of the quote. (I don’t know why this developed as a method — do you take a screen shot of a Web page every time you quote it?) Or you can use Blackbird Pie.

With Blackbird Pie, you enter the URL of a tweet (Direct URLs look like this: http://twitter.com/FakeAPStylebook/status/13449746429), and it spits out a huge line of code. Copy that line of code to your Web site, and you may or may not get a live version of the quote (that is to say, the direct URL is clickable, the Tweeter’s account is clickable, etc.) I say “May or may not” because this tool is not meant to work with some sites.

Let’s see if it works for me.

.bbpBox{background:url(http://a3.twimg.com/profile_background_images/62706301/fakeapbg.gif) #EBEBEB;padding:20px;}p.bbpTweet{background:#fff;padding:10px 12px 10px 12px;margin:0;min-height:48px;color:#000;font-size:18px !important;line-height:22px;-moz-border-radius:5px;-webkit-border-radius:5px}p.bbpTweet span.metadata{display:block;width:100%;clear:both;margin-top:8px;padding-top:12px;height:40px;border-top:1px solid #fff;border-top:1px solid #e6e6e6}p.bbpTweet span.metadata span.author{line-height:19px}p.bbpTweet span.metadata span.author img{float:left;margin:0 7px 0 0px;width:38px;height:38px}p.bbpTweet a:hover{text-decoration:underline}p.bbpTweet span.timestamp{font-size:12px;display:block}

Do not confuse the Mexican holiday “Cinco de Mayo” with the recent condiment stock trading scandal, “Sinkhole de Mayo.”less than a minute ago via HootSuite

If you saw a snarky tweet from FakeAPStylebook, it worked. If not…

The person who developed this, Robin Sloan, said he did it because “we just think it’s a pain to take screen grabs of tweets.” My setup makes it incredibly simple to take screen shots (which is why most of my writeups include screenshots nowadays). Further, I’m a little worried about the external image reference to Twimg.com. If Twitter goes down will that make the tweet impossible to load? And what happens if FakeAPStylebook deletes this quote?

I like Mr. Sloan’s idea, and if it were something like embedding a live, refreshing Twitter List I would be right there (why oh why oh WHY don’t Twitter Lists offer RSS feeds?) but I don’t understand the need to embed one tweet at a time. I just don’t get it. It’s probably me.

Answers.com Pushes Answers to Twitter

Answers.com, which you might remember from WikiAnswers or ReferenceAnswers, announced last Wednesday a new alpha feature called “Hoopoe,” which ties in with Twitter. Hoopoe’s Twitter account is at http://twitter.com/answersdotcom.

Here’s how it works. Either send a tweet to @AnswersDotCom or write a tweet with the hashtag #AnswersDotcom or #hoopoe . Answers.com will send an autoresponse with a snippet of information and a pointer to the answer on its Web site.

The first thing I did was send a tweet to @AnswersDotCom asking “What is a hoopoe?” No more than a minute or so later I had the tweet you see above. The URL ponted me to an Answers.com page that provided information from a dictionary, Columbia Encyclopedia, Western Bird Guide, and a huge Wikipedia article with lots of images. I am now very confident about my basic hoopoe knowledge.

Next I tried a more abstract question, “When is the sun coming up tomorrow?” I tried that twice with #AnswersDotCom and #hoopoe. Be careful about including extraneous text in your question; Hoopoe will try to answer all of it. (It would be great if it only tried to answer the text set before the hashtag. That way you can warn all your followers you’re playing footsie with an autoresponder.) AnswersDotCom couldn’t answer the question, referring me to the site instead.

Thinking that perhaps that question didn’t provide enough data (like where I was) I asked a simpler reference question, “What is the square root of 12?” I got an answer from Answers.com but it was incorrect. And it completely ignored my similar question from a few minutes later — I think it searched for “Square root of” and gave me the first thing it could find in its database.

So if you’re looking for reference-type information, Answers.com’s Twitter service is fast and good. And #hoopoe is a hashtag I could actually type on my cell phone without too much trouble. On the other hand if you want math questions or almanac questions answered, you’ll have to keep waiting for Wolfram|Alpha to come out with a twitter service.

Cuil Launches Cpedia, Web Aggropedia

Poor Cuil, a victim of what I like to call “Teoma Syndrome.” Teoma, for those of you who weren’t geeking out on search engines ten years ago, was a search engine that launched in 2000. It got TONS of publicity. Lots of people made noise about it. It was the alleged Google killer. Ask.com bought it in September 2001, spent some time working on it, and then it just kind of… faded out of the public consciousness. It’s still at http://www.teoma.com/ if you want to try it. It’s not a bad engine, if you can ignore the smaller data pool and the fact that for every page of ten search engine results you get ten sponsored results (five above and five below.)

When Cuil was launched there was a similar level of fuss, a lot of press, and then poor reviews and some concern about results. Cuil had interesting ideas, but couldn’t compare to Google at launch.

But here’s the thing. If Cuil hadn’t been so constantly and overtly compared to Google, it could have stunk at launch and that would have been unfortunate but okay… it would have had the breathing room to get better. As it was, there were bad reviews at launch, it wasn’t Google, and it didn’t get the traction it could have. And that’s one of my complaints about this “Google killer” business.

The thing that kills Google — one week from now or ten centuries from now — will not look anything like Google. Yahoo and Ask are not threatening Google right now, Facebook and Twitter are. And if something like Facebook gains ascendancy, it won’t be a Facebook clone that overcomes it. It’ll be some site or service that we can’t imagine now, like BrainColorSearch, a USB-powered portable MRI that you hook to your computer while exploring the Internet. The portable MRI device scans your brain constantly as you search and each item of Internet content has an aggregated brain scan associated with it. Searchers using BrainColorSearch match for relevance and for a map that most consistently matches their brain (or how they have declared they want their brain to look — a specified mood.) Searchers end up exploring a Web that makes them the most neurologically comfortable. (Soon tweeners begin “Brain Bombing,” hacking the MRI devices to search for content items that generate bizarre, almost impossible MRI activity maps…)

Okay, that example was a bit goofy, but you get the idea. The next company to displace the leader will not be leader+10%, it’ll be something completely different and (at least for a while) strange.

Anyway, what was I talking about? Oh yes, Cuil. Cuil has moved away from being a Web-type search engine and has announced a change to something different, an attempt to aggregate “instant reference” pages. Sort of like Mashpedia, which I reviewed last week. Cuil’s new effort, Cpedia, doesn’t seem to go to the breadth of Web properties that Mashpedia does, but I found the results more relevant in some cases with nice clustering.

Start your search at http://www.cuil.com and enter a topic. I used Benjamin Franklin as my test search with Mashpedia so I repeated it here. As you can see all the content in the result is from Web search — you won’t find Flickr, YouTube, etc here except as the byproduct of a Web search. The results were a little odd; the first one was “huh?”-inducing — I couldn’t figure out why it ranked higher than the Wikipedia article — but the rest of it was good content. There were other tabs across the top of the page to provide more targeted content — Benjamin Franklin Parkway, Benjamin Franklin House, etc.

There were also related topics on the right. The query Benjamin Franklin spawned such related topics as “Gentlemen Scientists,” “Continental Congressmen From Pennsylvania,” and “United States Presidential Candidates, 1808.” These headings show a list of topics which, when clicked, take you to a new page of search results (though in the same browser window, so open some tabs if you want to go off exploring.)

At the top of the topic list is a window that opens real-time results. I didn’t get any for Benjamin Franklin, so I tried cows and I only got one for that, so I tried Padres and then I got a bunch, from Twitter and other news services. A slider allows you to specify how recent you one the information to be, from last hour to last day. I am not normally one to comment on design, but it’s annoying to have the main search page designed vertically, and the streaming results designed horizontally.

I can imagine using Cpedia in a complementary fashion to Mashpedia. I’d come to Cpedia first, to explore content topically and zero in on the names/descriptions/topic headers that really find me what I want. Then I’d take those to Mashpedia and use them to explore certain parts of the Web (Flickr, YouTube, Twitter) more deeply.