Powerset

Powerset is a natural language search engine that can use everyday phrases and grammer to conduct more accruate web searches by understanding the search query and the pages it indexes. Parsing phrases and grammer theoretically produces better results because the egine has a better understanding of the searches intended goal than with just keywords alone. For instance, a Powerset search for “politicians who died in office” returns information on the subset of politicians who died in office, rather than a group of pages that ranked highly with the phrase.

Powerset presentation begins: talk about semantics and search, “we parse the web”. Natural language search.

Announcement: Powerset labs, where users can explore tech demos, share ideas, feed the learning engine and “improve your search karma”.

Demonstration of natural language queries with a social voting style feature. Touches of other sites

Demonstration of Powermouse (see screen shot), information is pulled from Wikipedia into a semantic index.

TC40 attendees will be amongst first in private beta.

Overall: tough sell in the search vertical, but interesting take. Great start to TC40.

Cognitive Code

Cognitive Code makes artificially intelligent user interfaces. Their main product is the SILVIA (Symbolically Isolated, Linguistically Variable, Intelligence Algorithms) platform, which can add a human-like artificially intelligent interface to nearly any digital device. The SILVIA platform can learn and converse in natural language to carry out tasks for the user. Potential applications include children’s digital toys and personal assistants.

Flagship product: “silvia platform” Symoblically isolated linguistically variable intelligence algorithm. Laymens terms: AI.

Demonstration with AI on the screen, the AI system is having a conversation with one of the Cognitive Code. A couple of bugs in the live demo, but pretty cool.

Uses include embedding in toys, phones, websites “unlimited uses.” First major target market is “smart toys.”

Clever idea, if they can pull it off we’re seeing the future of toys.

CastTV

CastTV is trying to build one of the web’s best video search engines by creating a rich index of contextual data about videos and an easy to use interface for searching them. The engine pieces together context for a video based on it’s metadata, the content surrounding it, and the content of pages linking to the video. Notably, CastTV also searches paid video searches such as Apple iTunes. Their user interface allows users to sort results by shows (to weed out non-relevant stuff), host (such as itunes, CBS Innertube, etc to focus on a favorite service provider), by date, relevance, prices, etc.

Presentation begins: CastTV doesn’t host videos, they index them.

Britney Spears video search compared, Google, Yahoo and CastTV: CastTV results are pitched as being better, more accessible etc

Colts Titans next example. CastTV is using smart clustering for results, pulling video from MSM and user generated content. Nice results, even if I have no interest in American Football

FAROO

FAROO is a peer-to-peer web search engine that has no centralized index and crawler. Each web page visited by users is automatically included into the distributed index. Ranking of search results is based on a distributed usage statistics of the web pages visited by FAROO users, which leads to a more democratic, user centric ranking. FAROO also shares advertising revenues up to fifty percent with its users. The search engine uses privacy-protected behavioral targeting to increase conversion rates.

Interesting concept, P2P in a strict sense. Results are only pages that have been visited by users…I cant’ help but think the SEO crowd is going to love this

The presenter claims that the algorithms actually prevent manipulation: he doesn’t know the people I know. Nice results though.

Indexing via a desktop P2P client, demonstrated version on Windows. Faroo beta opens today.

Viewdle

Viewdle is a white-label platform for indexing, searching and monetizing video. The technology they are developing lets video producers algorithmically extract metadata from news, shows, movies, and Internet video. This is much more effective than the old method of text-based metadata indexing. Viewdle’s most notable feature is their facial-recognition technology that can create a create a “real-time index of true on-screen appearances”. They plan on building one of the largest databases of people-in-video references. Reuters is currently testing out Viewdle’s technology with their videos news inventory by letting people search their catalog for specific people.

Demo starts with 2 minute demo video. Slick, we’ll see if we can get a copy.

The facial recognition is always an interesting concept, but I’m remind of Riya. More Britney Spears examples, although they are pulling data from others in the video as well, it looks a step from previous tech, particularly give it’s video they are scanning, not just pics.

Product: Top Chance, scans on criteria, including date. Popularity search includes total video time and when. Platform (presuming API) will also be released shortly to plugin widgets etc.

Expert Panel: Ryan Block Chris Anderson, Marc Andreessen, Om Malik, and Marissa Mayer

First question Marc Andreessen to Powerset, great question, how do you break out, API’s etc. Good response.

Chris Anderson: what are the advantages of the various products to the user

Faroo responds first: we are by the user, for the user, it’s good because “they are doing the search together”

Om Malik to Faroo: most P2P systems people turn off, how do you overcome that, also how do you seed the network?

Faroo: it’s not a problem…not a particularly good response.

Marissa Meyer wants to know about the video search startups, scaling etc…classic

CastTV: we’re scaling, focus. Viewdle “we reference a point” hence can scale to billions, using “fusion engine”

Discussion continues around AI and natural language tools.

Jason asks Om: which one is the most viable. Om: CastTV. One to last: Cognitive Code. Middle of the road pick: Powerset. Faroo is “interesting,” Viewdle will be “acquired soon”

Jason Calacanis to Marissa Mayer: will people switch away from Google. Reply: most people use more than one search engine according to stats. Google’s advantage is being a one stop shop. JC: what did you think of CastTV, MM: nice interface, clustering for duplicate issues is good tech.

Marc Andreessen: I don’t want to be obsessed with distribution…but I am, how do companies deal with it
Powerset: we’re very aware of this…uploading to users (???), embeding on external sites (Google custom search style I’d think).

Conclusion: speaking to Nick and we agree that CastTV was the winner in a very competitive group, good tech which just works with a practical use. Cognitive Code had the coolest product, but the demo wasn’t great which lost it for them.

Crunch Network: CrunchGear drool over the sexiest new gadgets and hardware.

There are a couple key issues they’re all struggling with as they try and generate the greatest amount of ad revenue. There’s still some uncertainty about where to put the ads (pre/post/interstitial?). Even the type or length of the ad is up for debate. A recent study found longer ads were more effective at branding, while conventional wisdom has cast doubt on users sitting through the longer plugs.

After deciding on the format, determining the content of the video in order to generate relevant ads is yet another tough problem. It’s also a dire matter for big brands that don’t want to risk being associated with inflammatory content. Finally, these ad platforms will need publishers, advertisers and a marketplace to trade in.

Here’s a look at what people are doing in video advertising:

Definitely the team to watch, YouTube is treading carefully, experimenting with text ads running along the bottom of the video that users can click on for a full video ad. They’re going to be testing the system with some of their top content producers and word on the street is that the terms are pretty good. Revver splits ad revenue 50/50 with publishers. They run ads at the end of viral videos, which might mean that people are still paying close attention after watching the main content. However, this also means they lose some precious real estate to help drive traffic to other videos on their network like YouTube does. Revver filters the content themselves, tying in the appropriate ads. Similar to Revver, VideoEgg helps publishers deliver and monetize their video inventory. It’s a very hands on approach suitable for larger brands that have tight control over the quality and context of their content. They serve up over 20 million videos daily across their EggNetwork. Ads show up alongside lead ins to other videos as well. ScanScout’s technology scans each video and determines content, with ads delivered contextually to match each scene. They run text ads along the bottom of the videos based on context derived from audio analysis and user behavior.

They’re like adsense for video, tying contextual text ads based on the content of a video. It looks similar to what YouTube is aiming for. When videos play, Adap.tv digs up relevant Amazon products and Looksmart ads to populate an ad bar on the bottom of the video at key moments. They use tags and other meta data, as well as speech to text translations to find out what the video is about.

AdBrite was one of the first to overlay ads on videos with their InVideo platform. Adbrite has created an embeddable video player similar to YouTube. If we choose to show a video on TechCrunch, we can use this embeddable player, and at our option it will include Adbrite ads and our logo as a watermark. Anyone who takes the content and embeds it on their own site will show the same video, with the same ads and watermark. And all click backs on the video go to the original site.

The most interesting ad play, BroadRamp wants to make everything you see on your video a possible point of sale. See a t-shirt you like? Just click the video to buy it now. Tagging or programmatically generating the links to products from the video may not scale or prove too difficult. Their core business is still video content delivery systems, however.

Formerly Podzinger, Everzing searches audio and video. Since they don’t own the content they can’t insert ads on the video content, but their speech-to-text transcription means they can help solve the problem of finding out the subject of a video.

A video search engine like Everyzing, Blinkx analyzes videos speech and meta data to tease out the content of the video. They also claim to use visual recognition as well. However, Blinkx has also leveraged their technology to launch adHoc, contextual advertising based on the content of the video.

Another video search engine currently running in private beta, Cast.TV looks at a video’s meta data and surrounding links to determine more context around the video. We’ve been impressed with how well it works. They haven’t discussed plans to incorporate advertising, however.

Coming up with a kick ass, scalable ad platform solution for social video that satisfies the needs of publishers, advertisers, and viewers is only a piece of the problem. While finding the most effective format will take a lot of testing until consumers reveal the most effective methods, the platforms will also need video content to monetize. Since well defined video properties with targeted content can work with sponsors on established video ad networks, the ideal market for these platforms remains effectively monetizing the jumble of amateur viral video floating around on social networks and YouTube. However, YouTube, which currently owns the lion’s share of video on the net, seems to be taking their time developing the solution in house.

That leaves becoming a destination, partnerships, or acquisition as possible outs. Video search sites like Blinkx and Everyzing are currently monetizing their search pages, but can’t take full advantage of their platforms by embedding ads into the content they link to. While these sites offer deeper video search, existing as a destination site is also a tough path that goes up against established web properties like Google, Yahoo, and AOL. In a slightly different way of going it alone, AdBrite has been going directly to publishers with their InVideo player. Adap.tv has been testing out partnerships, trying their platform out on MetaCafe.

As with most ad platforms, advertisers and publishers will be trying them out for effectiveness. In the end, the startups that can deliver the most return to these two will win out.

Crunch Network: MobileCrunch Mobile Gadgets and Applications, Delivered Daily.

adap.tv	AdaptiveBlue	Aggregate Knowledge
CastTV	Critical Metrics	Jangl
Pando Networks	SodaHead	Spock
Wize	ZapMeals	ZenZui
	Zing

Adap.tv – They’re like adsense for video, tying contextual text ads based on the content of a video. When videos play, Adap.tv digs up relevant Amazon products and Looksmart ads to populate an ad bar on the bottom of the video at key moments. They use tags and other meta data, as well as speech to text translations to find out what the video is about.

AdaptiveBlue – Makers of the Blue Organizer, a Firefox bookmarking and tagging add-on that parses web pages, adding contextual information where appropriate. For instance, if you go to a web page about a band, Blue Organizer’s right-click menu will show you more info about the band drawn from sites like Odeo or Wikipedia. The plugin also has smart links that let you easily push the link to services like Digg or LibraryThing. More coverage of the recent feature additions here.

Aggregate Knowledge – One of the more established companies at the event, they work with online commerce sites to provide personalized recommendations by looking at user’s collective behavior. They just closed a large round of financing and are rumored to be profitable after a little over a year in operation.

CastTV – A video search engine that pieces together context for a video based on it’s metadata, the content surrounding it, and the content of pages linking to the video. The service performed well in our earlier review. They recently raised a $3.1 million round of financing from DFJ.

Critical Metrics – A music recommendation service that aggregates music reviews from around the web. Each song includes an audio and optional Youtube sample and purchasing options from services like Yahoo Music, iTunes, or Rhapsody.

Jangl – They specialize in anonymous phone communication. A Jangl is a real phone number lets people call you with knowing your real number. The first time someone calls you they have to leave a message and request permission to connect to you directly. You can ban a number at any time as well. They just recently launched a service that lets you generate a Jangl number for any email address and leave a voicemail for that user and number for a callback. Calls are served over a VOIP bridge, so it also makes long distance calls cheaper.

Pando Networks – Desktop peer to peer file sharing service Pando speeds up file transfers by torrenting files and buffering them over their higher speed network of servers. At Supernova they’ve announced their Pando Publishing Platform that lets users easily publish to the web with the cost savings of peer to peer. The platform gives publishers the benefits of P2P video streaming amongst their users and CDN peering service. P2P streaming lets users view video incrementally instead of after a lengthy download. Their CDN peering service will let turn a regular CDN server into a supernode that will save on bandwidth by balancing load between users and the main servers. They’ve already lined up content partners NextNewNetworks, Blip.tv, and Rever.

SodaHead – A polling destination site that lets users poll their friends, SodaHead experts, or strangers. Polls are embeddable widgets that can be voted on at the destination site or any page featuring the code. Polls also feature comments so users can express opinions that don’t fit into any of the options.

Spock – A people search engine that automatically aggregates information linked to a person along with support for updating contacts.

Wize – Wize is a site that tracks expert and user product reviews across the Internet and churns them through an algorithm to create a single, 1-100 “WizeRank.” Earlier this year they got a $4 million round. We have a review of other review services here.

ZapMeals – Adding a little levity to the event, Zapmeals is a startup spoof (e.g. the fake) that aims to be a marketplace for meals, hooking up hungry stomachs with nearby home cooked meals or caterers. You choose your cook based on a member rating system and their fleet of couriers would deliver the meals to your home.

ZenZui – A new way of surfing the web on your mobile phone browser, Zenzui economizes on your phone’s screen space by displaying sites and services as icons on a grid display. You can scan the 36 slot grid using your numbered keypad and zoom in for more detail on a specific service. We covered their launch here.

Zing – Zing enables mobile music players to connect to music libraries over WiFi. They’re currently powering the SanDisk devices for Yahoo Music and Pandora.
You can see a Wink powered group for the event here.

The 12 real start-ups were hand-chosen from more than 130 applications. StumbleUpon, one of the 2006 Connected Innovators, has already enjoyed great success as a newly acquired eBay business. We have high hopes for more great success stories from this year’s crop of Connected Innovators.

Disclosure: While these companies were selected from 130+ applications, they were required to pay a fee to participate once selected. As a partner to the conference, TechCrunch received a percentage of that fee.

Crunch Network: CrunchBase the free database of technology companies, people, and investors

The traffic figures support this. Comscore reports that the site has 40 million monthly unique visitors and is growing at about a 50%/month rate.

Truveo says this is a combination of their approach to search as well as the fact that copyright holders are becoming more diligent about pulling down unauthorized uploads on YouTube and other sites. The result is that YouTube is no longer the single place to find good video clips. People are turning to search to find Daily Show clips on the Comedy Central website, for example, instead of just looking for them on Youtube.

The company is also partnering outside of AOL to get broader distribution for the engine. Brightcove, Clevver, CSTV, Flock, Pageflakes, PureVideo, Qwest, Search.com, Sportingo, Netvibes, Widgetbox and YourMinis all now use Truveo for video search, joining existing partners AOL Video, AOL Search, Excite, Infospace.

There’s a new kid on the block, though, called CastTV, which recently announced a $3.1 million Series A round of financing from Draper Fisher Jurvetson. They haven’t launched yet, but the results from demos show it to be as good or better than Truveo. Given that AOL already owns Truveo, look for a quick acquisition of CastTV, probably by Google, if the technology is as good as the demos suggest.

Crunch Network: CrunchBoard because it’s time for you to find a new Job2.0

We first wrote about the company, which is led by husband and wife team Edwin Ong and Alex Vikati, last October. See our post and screenshots here.

AOL arguably has the best video search technology through their acquisition of Truveo in early 2006. Google also has a growing video search engine that includes both Google Video and YouTube videos.

Since many videos posted on the internet contain little or no associated meta data to describe what’s in the video, services like AOL’s Truveo and CastTV look at surrounding text to help determine what the video is about, and make it more searchable. Truveo does a great job with this. CastTV is much better based on the demos I’ve seen. Alex and Edwin won’t disclose all of the technology behind the service, but part of the trick is that they are able to track videos through multiple links on a site, collecting metadata along the way. And they also parse the code on the video files as well, gathering additional information about the content. If tags are available for the videos (such as YouTube tags), these are indexed as well. The final step is actually even more interesting – CastTV will take the data they are able to collect about a video and search the web in general for additional data. If there’s a close enough match, CastTV adds that information to the content metadata. And unlike other video search tools, CastTV indexes movies and shows from iTunes and other for-pay services.

The company says they will enter private beta next month; full launch will come this summer.

Crunch Network: CrunchGear drool over the sexiest new gadgets and hardware.

Now they are preparing to launch their second company, CastTV. It is an ambitious effort focused entirely on video search. And that’s not a bad place to be. As we’ve recently written, rich media search is a very hard problem to solve.

A notable success in video search is Truveo, which launched in 2005 and was acquired by AOL in early 2006 for an estimated $50 million or more. Truveo had a unique way of searching video content. First, finding video can be hard, and Truveo does a good job of locating video on indexed web pages. Second, Truveo looks at text surrounding video links, as well as metadata included in the video file itself, and makes intelligent assumptions about the video.

CastTV is taking many of the successes of Truveo and building on them. Search itself is significantly better than any other video search we’ve seen, and the user interface, which includes really smart filters (more on that below), makes it easier to find what you’re looking for, fast.

Edwin and Alex won’t disclose all of the technology behind the service, but part of the trick is that they are able to track videos through multiple links on a site, collecting metadata along the way. And they also parse the code on the video files as well, gathering additional information about the content. If tags are available for the videos (such as YouTube tags), these are indexed as well. The final step is actually even more interesting – CastTV will take the data they are able to collect about a video and search the web in general for additional data. If there’s a close enough match, CastTV adds that information to the content metadata. And unlike other video search tools, CastTV indexes movies and shows from iTunes and other for-pay services.

All of this doesn’t mean much unless the search performs well. In tests I was able to find full versions of movies and tv shows that simply didn’t show up in Yahoo, Google or Truveo search. A free episode of 24 at a Myspace Burger King page. Full versions of Pirates of the Caribbean 2 with price comparisons ($14.99 at CinemaNow, $9.99 at iTunes), as well as available formats (WMV and Quicktime). Searches for trailers, movie and show clips, and user generated stuff all yielded equally impressive results.

Results can be sorted by shows (to weed out non-relevant stuff), host (such as itunes, CBS Innertube, etc to focus on a favorite service provider), by date, relevance, prices, etc.

Look for CastTV to launch sometime in the Fall, and in my opinion be acquired shortly thereafter. In the meantime you can sign up on their home page to be emailed when they launch. If you’d like to see founder Alex Vikati demo the product in person, make sure to attend Jeff Clavier’s Search SIG event next week in Silicon Valley. She’ll be there, along with other hot video startups.

Additional screen shots below.

Crunch Network: MobileCrunch Mobile Gadgets and Applications, Delivered Daily.