Rev2.org

“Finally, something!” are the cheers in Silicon Valley this morning. Powerset, the mega-hyped underdog Google killer semantic play, has finally launched a working demo of its product. While the search engine has been in closed beta for some quite now, this is the first time it has unveiled any of its technology to the public.

What is being unveiled is essentially a search engine based on top of Wikipedia. Allowing you to enter a topic, phrase or question, Powerset uses its semantic algorithm to dig through the keywords and come up with the best matches — or better still, answers — based on the Wikipedia articles. While this is in no way a final product, the demo is supposed to show the glimpse of its algorithm on a test index.

Trying to differentiate from their suggested examples, I tried a few queries, and I am no less than impressed. The first query I tried was “who is Andy Kaufman.” Rightly enough, Powerset was able to extract a mini-bio from Wikipedia, structurally providing a brief description, date of birth, date of date, profession, film roles, etc. What I was more impressed was their ‘factz’ section where key facts are analyzed and put into context. While not totally accurate, I was able to find out names of the characters he portrayed among other things.

The second query I tried was, simply, “Iron Man.” To my surprise, a special character/name supplement was presented again, and this time with more detail. Among the tabs included “Iron Man,” “Film,” “Video Game,” “Song,” “Magazine,” “TV Series,” and “Band,” presumably all taken from Wikipedia’s disambiguations. Going through, I was able to get a good idea of each, although in brief snippets.

For my final query, I decided to try something a little more obscure. This time, my query was “who invented the computer mouse?“. Rightly enough, the first result was Douglas Engelbart’s Wikipedia page, with (in highlighted text) “He is best known for inventing the computer mouse”. All over the results, his name was repeated again and again, too, and tied closely with “invented” and “computer mouse,” so I could be sure it was him just by glancing.

I’ll admit, Powerset’s first display isn’t a bad one. It does the trick, and definitely defeats Wikipedia’s own search in a heartbeat (but then again, that’s not hard to do.) For the general queries, especially those which you’d expect to have a Wikipedia page, the results are more than average, O.K. for slightly obscure ones, and kind of clueless for the more obscure ones. It does not bad for a search engine that has been in stealth mode for 2 years, but certainly, we’ll need more to judge. I’m waiting for more the Google killer till it gets a pass for the hype its received. Oh, and I swear if I hear the word semantic again, I’ll …

Update: we have invites! Head on over to http://viewzi.com/invites and enter “gio” as your referral code.

Visual search engines, or rather, the ones that make a claim to provide alternative visual ways to search, have been on the rise lately — and rightly so, with the advent of Flash and AJAX. A few days ago we reviewed SearchMe, which featured an Apple-like ‘Coverflow’ view of its internally-crawled search results. Viewzi is the latest play in the area and is being touted in the blogosphere as the first to get it right.

The search engine, firstly, differs from the few I’ve seen in that it doesn’t assume that only one type of an interface is suitable for every search. Instead, it presents and allows you to choose from a slew of different interfaces and tools after you make your query. These are slightly optmized to the query, but I didn’t find much of a difference between the ones I tried — perhaps its because of the limited number of search tools available at this point.

There are, obviously, a whole bunch of tools to select from, but I’ll cover the main few that were presented for most general queries (note: unfortunately you need beta accounts to click on the ‘example’ links. If the Viewzi folks want to help out invite-craving Rev2 readers, we’d love some):

Simple Text View (example)
Loads simple/regular text results from meta-querying Google and Yahoo!. While fundamentally the same as normal search results on the respective search engines, they are slightly more edge and ‘2.0′ and sleek with their highlighting, and provide a helpful Compete.com rank for each site.

Web Screenshot View (example)
Similar to SearchMe, the results with this tool are presented in the Apple coverflow style, although not so mimetically. Additionally, since Viewzi is a meta search tool and doesn’t crawl its own index, the results are from Yahoo!, and while drastically more relevant than SearchMe’s, the screenshots are of inferior quality.

Basic Photo View (example)
This is the image search function of Viewzi. Searching Riya and Flickr, the tool gathers images and presents them in a more navigatable format. Although, it links back directly to the original image as opposed to dynamically enlarge it unlike most other visual image search tools I’ve seen, which I thought was a little odd — if that’s the point, why not just use Flickr?

4 Sources View (example)
One of the strikingly different tools out there — and in a good way — this is like meta search 2.0. It searches Google, Yahoo!, MSN, and Ask, and lays out screenshots from their results with corresponding labels as to which search engine it came from. For ones presented in multiple, or all 4, the screenshots are stacked. I’ll admit, it’s no less than neat.

Video x3 View (example)
One of the more nichely useful tools I found, this one searches through video search engines — namely YouTube, Blinx, and Veoh — and presents the results in a scrollable landscape video reel-like format. For a video searcher, it’s not a bad searching alternative as it is. I also love how the Blink search results are animated.

Everyday Shopping View (example)
This is one of my personal favorite tools. It searches shopping sites — Amazon, eBay, Walmart, and Target — and lays out the results in a simple, standard format with the picture and the price. For a shopping searcher, this could prove immensely helpful. And in the interface lickability factor, it doesn’t score too badly either.

Conclusion
After reading through the hype and testing out the tools individually, I have to admit, Viewzi lives up to it. Something I love is that as a search engine, it’s focus is not on relevancy or providing the best search results, but on the different ‘views’ and how best to present them using search algorithms already out there. And it works fantasically. I’ll admit, there are a few clunky tools — such as the Photo View and Screenshot one that I’m not sure add a whole lot of value — but there are some real jems.  And the best thing is, as more get built, it’s going to keep growing with more new tools. For a visual searcher, for the 2.0 searcher, and for the advanced searcher, Viewzi isn’t just worth a try — it’s worth the replacement.

A few months ago I wrote my first article for another blog about Google Apps and its potential to change the way small businesses address their technology needs. Google released Google Sites a month or so ago, so for my first article for Rev2, I thought I’d revisit the topic and take a look at the impact it could have.

After spending the weekend playing with the capabilities and building a Site to address a specific problem my wife is dealing with at work, I can say without a doubt that this is another watershed moment in the evolution of web based applications. In a few hours I was able to assemble a series of spreadsheets, calendars, forms and gadgets to create a completely customized application that solved a problem that she has been struggling with for years. It isn’t quite as functional as we would like, but I’m sure that as the features evolve with time we’ll get all the capabilities needed.

So, why is this so significant? Their are several reasons, but I’ll touch on just a few. First and foremost, Google is again giving away something that we have been accustomed to paying for. Google may know better than anyone exactly what Chris Anderson means when he talks about ‘economics of free‘. They are letting anyone use the applications for basic domain services, alleviating a major strain on small business infrastructure.

With Google Sites, they are letting the users tie their applications together to create custom interfaces for their data and collaborate more effectively. In my simple Site I hooked up multiple spreadsheet and calendars to create a simple call log and staff schedule tracker. By combining the information in them with the functionality of gadgets you open up endless possibilities for using the data.

This functionality, while not the same currently, could one day mimic that of companies like LongJump and Coghead. What’s more interesting is that some small business owners might be much more comfortable with the idea of setting up Sites with spreadsheets and documents than trying to create a ‘database driven web application’ for one simple reason - they have already been doing it for years. All of sudden taking those things online and creating a dynamic application doesn’t seem that complicated anymore.

Google is allowing people working together to address process issues and apply custom solutions using tools they are already comfortables with. This is a very liberating concept for a lot of businesses, especially if it means that a smart team doesn’t need a developer to get the features they need.

But I’m getting ahead of myself. There is still a lot left to add before this is truly a competitor to online application service providers, but Google could close the gap fast. One major missing point is interactivity with embedded page elements. This will come with a more robust permissions system, but not being able to enter data on a spreadsheet (without a corresponding form) or add events to an embedded calendar certainly hold the overall functionality back. When they enable this, and if they’re able to integrate things like email processing and dynamic data mash-ups, we’ll have a whole new way to process business information.

If there’s such a thing as a Google fanboy, I probably fit the mold, but Domain Apps is worth the praise. I think it’s hard to argue with free applications that are powerful enough to run a business on. It’s time to take a close look at how to incorporate this new capability in your work.

Ed. Note: We have 250 free Xoopit invites for Rev2 readers! Head over to this link and enter “Rev2″ as your invite code.

San Francisco-based Xoopit unveiled in a private beta this morning, also announcing a $5 million Series A round led by Accel Capital. The startup aims to convert your inbox — a simple and traditionally unaltered form of communication — into a more exciting and usefully presented social network (though what it really is, is better e-mail.)

The service works in three forms. With your Gmail account, iGoogle homepage, or directly on the web. The signup process consists of entering your Gmail account information (currently the only e-mail service supported) and installing the plugin — a cool thing about this being that your e-mail login (e.g. Gmail) IS your Xoopit login. Meanwhile, Xoopit spends a few minutes excavating  “precious photos, videos, and files” from your inbox. In my case, it must’ve taken around an hour so.

With the e-mail excavated and plugin installed, heading to your Gmail inbox reveals a whole new kind of dimension to your e-mail — a Gmail 2.0 or rather what Gmail could be in the next three years. A top bar gives you an expandable line of thumbnail previews of the latest Photos, Videos, and Files in your inbox, which you can search through, as well. Clicking on the respective links takes you to a more deeper view — for example, in the same way that your inbox hows you the conversations you’ve recently exchanged, clicking on “Files” shows you an inbox-like view of all your recently received files, same with “Photos” for photos, etc.

This in itself is a really useful way of getting a grip on the contents on your inbox and seperating out an e-mail from its contents, but that’s not where it ends. Heading to Xoopit’s own website and logging in reveals a full-fledged social network with the whole Facebook-style News Feed of your recent items, but produced without you or anyone else having to take a single action.

For example, I can see that recently, I’ve been sent 10 photos by my mother, 3 files by a college, 3 photos by a PR person, an .RTF “FYI” receipt from a recent order I made online, and so on. To be honest, this is the most dynamic, useful view of my e-mail that I’ve ever witnessed. Listed under “People” (friends list) are all my contacts with whom I’ve recently been in touch with, and clicking under the name I’m able to browse through tabs to see the stuff they’ve sent to me recently.

Xoopit is starting small, but I can without a doubt say they’re onto something. If this kind of presentation and interaction isn’t the future of e-mail, then I’m not sure what is. Frankly, I’m surprised no one’s thought of this idea before — it seems obvious after Xoopit that social networks have undoubtedly mastered the “here’s your life” user interface, and there is a whole world to conquer with e-mail in this respect. I certainly hope with their $5 million round they’re able to get somewhere with this concept, because this is how I want to live through my e-mail, darnit!

Calendars — and their formats — have had to deal with proprietary solutions since day one of the personal productivity era, let alone the web. Finally with services like Google Calendar and devices like the iPhone, we see a lot of convergence, but it’s still far from a standard solution consumers would prefer. Enter FuseCal.

FuseCal has two primary functions: they allow you to save — or scrape — events from the web directly onto your personal calendar, and for the advanced lot, publish calendars on their own websites. In a way, I like to think of it as a “FriendFeed for calendars/events.”

The primary action with FuseCal is importing calendars. You can enter the URL of any iCal file or web page, and and FuseCal will go ahead and harvest the data — doing its best to import it into your personal calendar at FuseCal. Obviously, calendar files are imported without any trouble, but where FuseCal shines is in scraping normal data. Once you’ve gathered your events, you can save your whole calendar feed into almost every service available — namely Windows Calendar, Outlook, Google Calendar, Yahoo! Calendar, and iCal.

The secondary action with FuseCal makes it possible for bloggers, website owners and publishers to publish their calendar onto their webpage. Give it a name, an initial page URL, and FuseCal will give you an embed code for your badge. Here, visitors can save your events into their own personal calendars — great for conferences and the like.

FuseCal is a great utility for a lot of users. Of course, one may only use it to gather feeds and then subscribe to them in Google Calendar, but the functionality it offers is useful and unseen. For people who attend a lot of conferences or events from the web, and people who simply want to converge their calendars, FuseCal is great.

At Rev2, I’ve sold my own ads since day one (except for, of course, the AdSense blocks.) This is because while I realize the power of an effective sales force, I also think a small to medium-sized publication doesn’t need one to sell 4 - 10 spots, let alone adhere to the hefty 50% cut put forth by some ad networks. And it’s worked out superbly. Now, Adregate just made my job easier by a milestone.

For small to medium-sized publishers, bloggers, and website owners (such as myself), the service allows you to manage, literally, every ad spot and campaign that ever happens on your site. After signing up, you can specify different ad spots that you have available (e.g. leaderboard, 125×125 button, medium rectangle, etc.) Of these, you can either “port” over your existing campaigns — if you have any — or provide potential advertisers with a media kit.

The service has a range of plans, supporting 1 website and 50,000 impressions in its free one to everything unlimited in its $200/month. While the charge may seem hefty, for some publishers, the backend control panel and the value added to both parties may make it more than worth it. Adregate delivers fairly rich stats and I can’t emphasize enough how many aspects of your ads it lets you control.

The service, by being more of a tool and less of a “tried and tested” ad network of sorts, fills a huge niche in the small to medium publisher space, and I may just look into spending the weekend porting over all my campaigns to it. And I’d recommend all fellow little-guy bloggers to do the same!

Remember mixtapes? Yeah, those retro cassettes you had to play in these retro cassette players and rewind back to the start to listen again. Well, here’s mixtapes 2.0: Muxtape. The service offers a super simple way to create mixtapes from your MP3 files and share them with friends.

After the simple signup procedure, your only instructions are to select up to 12 songs from your hard drive from MP3s no larger than 10MB and hit “Upload.” Once you’re done uploading, the file is taken and added to your Muxtape — for example, here’s mine.

Muxtape is a service that’s so simple, useful, and to-the-point, that no detailed article is needed to elaborate it any further, yet it’s important enough to cover. Go create your Muxtape!

Mountain View-based Boorah, which launched at last year’s DEMO conference, is a site I recently learned about that aims to garner a community around restaurant reviews in major cities. The site follows the concept of Yelp and a number of new startups aiming at building a community around the local market.

Supporting most cities in America, as a user (or searcher — looking for restaurant reviews) — there a number of ways you can use the site. Of course, being a contender in the local space, the site is organized by cities and regions. You can click through one — for example San Francisco — and a page is presented with the latest restaurant reviews in the area, pictures of restaurants, top rated restaurants, most popular cuisines, discounts, among more. Needless to say, if you’re looking to discover restaurants in your city, the Boorah page should contain everything on your city’s page.

Another way to use Boorah is through it’s search functionality. You can search by category/theme/keyword (eg., “Cheesecake factory”,”Pizza”,”Romantic”) and by Neighborhood, City, Zip, or Address, and up come results that are organized by recommendations, buzz and ratings (of course, relevancy is a minor factor in the local search game.) This immensely useful, and I think is what makes sites like Boorah shine — for example, I was able to discover the 10 best Pizza restaurants in San Francisco just by searching through it.

A Boorah representative, Sanjay Chakraborti, alerted me about some impressive milestones they crossed recently:

• 1,000,000 Restaurant Reviews
• 225,000 Restaurants
• 45,000 Restaurant Menus
• 30,000 Restaurant Pictures
• 6,000 Discount coupons for various restaurants
• 4,000 Restaurants with Online Reservation capability

Unlike Yelp, what I like about Boorah is that they’re aiming for a niche, yet hefty market, and do a great job of serving that, as opposed to chasing others. So the next time you’re looking to go for a romantic dinner with your girl friend, use Boorah!

There’s a positive and negative to everything, and one of the negatives to Twitter remaining as simple and open and bloatless as it is, is that it can get very hard to follow Twitter conversations and keep a tap on who-started-what-discussion and who-responded-to-which-tweet-which-responded-to-the-discussion-starter. You get the idea. Quotably is  a new service that tries to solve this problem for all Twitterfolk.

If you’ve used API-based services like Digg Suggest before, Quotably will seem familiar. You enter your (or a) Twitter username, and up come all the conversations and discussions that person started, and took a part in on Twitter. The tweets are organized in threads, much like comments on Reddit or a discussion board.

For some of the featured Twitter users and to get an idea of what the service looks like if you don’t use Twitter, check out the Quotably pages on Chris Pirillo, Leo Laporte, Robert Scoble, and Veronica Belmont. And if you’re already a Quotably user, you can follow Quotably on Twitter to get the latest updates on feature releases.

The way I look at it, Twitter a huge 3-Dimensional world, and a service like Quotably gives you some 3-D glasses to look through. Sure, it’s possible to do without — you’re not going to be blinded without it — but to get perspective and see the big picture and the complete experience, it helps if you’re wearing them.

What I’d really love from Quotably, or anyone else for that matter, is a leaderboard for the Twitter conversations of the day. A lot of times, memes and interesting discussions are spread by a small community of users (like the Twitter Color Wars) and the only way I’m acquainted by them is by going through the tweets of the people I follow and trying to get a sense of it. Instead, I’d love to get the “Top 10 Twitter Conversations for Today” (by length/number of replies or otherwise) and just read through it in my daily coffee break.

There’s been a lot of activity lately in the search space, and equally so in the semantic web. Every day, we learn about a new search engine that’s trying to search X by using Y, or a new startup that has the buzzword “semantic” all over it. Today, I learned about Swotti — a new startup that combines both search and the semantic web — and is the most practical  player I’ve seen in either fields recently.

A self-proclaimed “opinion analyzer,” Swotti goes with the AskJeeves-esque theme with a mascot dog waiting to give you advice. As Swotti claims on the homepage:”Hello, I’m Swotti. I read each day 3 million opinions about products. I can learn the good and bad experiences of users. Let me advice you!”

Swotti aims specifically at products and places and objects and things. For example, while it won’t recognize “how to clean up your room, type in “iPhone,” and you’ll get a page full of images, bar graphs on the varied opinions (”I love,” “I like”), categories and rating charts (”Design,” “Usability”), and a list and snippets of actual web pages where people have made a comment on the device (”I highly recommend it,” “I hate it,” or “it’s incredibly innovative.”)

Unlike Wikipedia and Mahalo, the search engine is obviously fully technologically driven and almost 100% lackfull of human input (unless of course, you count the actual analysis which is 100% human input.) As a result, what’s really amazing about Swotti is that pages/results are built and compiled on the spot, at that very moment. When I typed in queries like iPhone, socks, Taj Mahal, and The Office, it was able to automatically recognize their specific type of product/object and present relevant images and tags and opinions all without human input — something that totally amazes me.

When you look at Swotti, you realize stuff like this is undoubtedly where the web is heading. Sure, humans can do it better — and when it comes to search engines like Mahalo humans are instrumental in their core vision — but that’s not to say that computers are far off of doing an equally important job directed at a seperate purpose. Ultimately, it’ll be the combination and input of both that prospers, but for now, it’s fun to watch them heading towards a collision — and yet being far from it.