Project10X has just published a "Web 3.0 Manifesto". It's a kind of sequel to their Semantic Wave 2008 report released in January this year. Mills Davis, Managing Director of Project10X, told us via email that the new manifesto "reveals how semantic technologies will drive product and service opportunities in the next stage of the internet."
You can download the Executive Summary here. We got a look at the whole report and it is packed full of great data, including the two top 10 lists of Semantic Web opportunities detailed below.
Also in this post we're giving away 3 free tickets to Defrag for the best suggestions in the comments for 'web 3.0' apps. See below for more details.
Top-10 List [of Semantic technology opportunities] -- Consumer
1 Interest networking
2 Semantic social networking
3 Semantic bookmarks
4 Semantic search & QA
5 Semantic desktop / webtop
6 Semantic blogs, wikis
7 Semantic identity management
8 Semantic mobility
9 Semantic email & IM
10 Reality browsing, avatars, & context-aware games
Top-10 List -- Enterprise
1 Information sharing
2 Semantic search, discovery, & navigation
3 Semantic mashups and composite applications
4 Semantic infrastructure / middleware SSOA, SBPM, SWS, virtualization, policy-based computing
5 Semantic business intelligence
6 Semantic ERP applications CRM, PLM, SCM, HRM
7 Semantic governance, compliance, & risk
8 Semantic web sites, wikis, collaboration, interest networking, & collective knowledge systems
9 Semantic advertising, marketing, personalization, & customization
10 Intelligent systems knowledge-based research, design, engineering, simulation, planning, scheduling, optimization, & decision support.
Totally unrelated to Project10X's report, but we just happen to have 3 tickets to the Defrag conference in Denver Nov 3-4 to give away. To win one, simply leave a comment in this post detailing what kind of 'Web 3.0' app you wish to see developed. You can use the above top 10 lists as inspiration, or wing it ;-)
The best 3 as chosen by our editors will win a full ticket to Defrag.
UPDATE: Thanks everyone for the great comments. The 3 winners of the Defrag passes are:
Scott Brinker
Jesse Wilkins
Edward Benson
Comments
Subscribe to comments for this post OR Subscribe to comments for all Read/WriteWeb posts
So, to suggest an interesting Web 3.0 app all we have to do is puck a buzzword of the day and slap "semantic" in front of it. It doesn't actually matter what a "suggestion" really means. Well, let's see what we can come up with; here some of my proposals:
Semantic SaaS.
Semantic AJAX.
Semantic ice hockey.
Semantic financial crisis.
Semantic US president (now that would be a killer).
Semantic lolcats.
...
Posted by: Berislav Lopac | October 16, 2008 2:28 AM
Berislav, lol! But you make a very good point...
Posted by: Richard MacManus
|
October 16, 2008 2:51 AM
hmm.. what about
Semantic Ad Network
Posted by: john | October 16, 2008 4:10 AM
I'm pretty excited by opportunities with APML. All kinds of apps could be developed around that. It's still very early days for semantic tech.
Posted by: Chris Lake | October 16, 2008 4:34 AM
That manifesto is, to be polite, BS. As Berislav indicates I can just invent a new “app” by appending Semantic to everything. That said figure-1 - the evolution of the internet - is useful. In the context of that chart I would say the Semantic Web is about connecting information, not connecting knowledge.
Knowledge is actionable it is what agents (human or computing) obtain from information. To enable computing systems to connect information we need to imbue that information with meta-data so that it can be interpreted across silos/applications. This is already happening albeit ad-hoc.
Yahoo! Fire Eagle is a really good example. The entire service is about representing “my current location.” As a result any application that retrieves information from Fire Eagle knows, because someone programmed the semantics, that the information represents a person’s current location. It doesn’t matter to what level of precision my current location is described is still represents my current location. Enough.
Some Semantic applications:
1. A television that when I switch it on with my mobile phone (as the remote), automatically shows “my personal channel” because it knows it is me, that I’ve tagged some video links that flowed through my twitter stream as “To Watch.” along with my Netflix queue is etc.
2. A Photo Frame that is smart enough to know that when my best mate’s ex (also a friend) comes into the house that it doesn’t show any photos from my mates Flickr stream.
Ok, so maybe those applications are Ubi Comp.
3. When analysing/researching the markets for investment options I can discover the inter-relationships of people, companies, manufacturers, performance, listings, market news, greenness etc. easily Mapping these relationships to provide meaningful knowledge efficiently. This is what the Metaweb’s Freebase Parallax can/could easily do this given the structured information and an ability to map the vocabularies used to refer to the same concept across domains.
Essentially Semantics structures information so that it can be more easily connected. The connection allows us to obtain useful knowledge more efficiently in the same way that Google connected keywords to webpages.
Posted by: Robert O'Brien | October 16, 2008 4:37 AM
If you imagine semantic apps marriaged with mashups like microsofts popfly, then it's plausible that one could achieve an application that takes a suggestion, any suggestion really, like "travelling, with low prices, to places with good weather forecasts" and get served an app with a list possible cheap travel destinations with a good weather forecast attached, listed in order, and letting you start booking it from there.
This is in my mind where semantic services has to go.
Posted by: Mikael Bergkvist | October 16, 2008 4:45 AM
A quick follow on from my previous comment. A useful service that would fit the definitions of the manifesto under the “Semantic Cloud Computing” category would be a Cloud based Triple Store. Essentially a refinement of Amazons Simple DB service to store a three tuple instead of the just the attribute-value combination currently. Obviously the query facilities would suit the needs of accessing a graph too.
Posted by: Robert O'Brien | October 16, 2008 5:03 AM
Create 3d avatar/game environment with the same basic rules that understands how to interface with contacts from your gmail contact list, your facebook or myspace friends or your second life contact list depending on the context of web page (in an activeX or FF plugin) or extended 3d virtual environment like Second Life or There. Data Access is made through central webservice calls that describe the environment and download interface code (C#/Silverlight CodeDom, Java, LSL, JS, etc).
Posted by: Mike Simon | October 16, 2008 5:22 AM
Since I am a writing teacher and grade a lot of student papers, I'd like a semantic web application that does all that grading for me. Seriously, though, I hope that this would help some people I know, break out of their rather stiff ideas of what web and web design are.
Posted by: pz | October 16, 2008 5:45 AM
I'd like to see a Semantic version of Xbox Live to find people to play games with. If it could select people with similar interests and game play patterns then I might have a better time gaming online.
Posted by: Bukola Ekundayo | October 16, 2008 6:27 AM
Semantic recommendations and syndication - think mixing expert profiling, semantics, RSS, and Amazon-type recommendations to produce feeds of information that could be subscribed to by users or pushed out by e.g. an enterprise RSS system.
Posted by: Jesse Wilkins | October 16, 2008 6:48 AM
Possibly a semantic translation engine. Using natural language processing the engine could understand denotation and connotation in all languages. This would make the translation of full texts and long strings of words extremely precise.
Posted by: Ben Walker | October 16, 2008 7:16 AM
What I would like is an "interest system for business opportunites detection".
e.g. back in 2002, I owned a webagency. In order to find leads, we had built a system linking financial result to a website and its age (looking for old tags or old design). We are doing the same with good results with my new venture (http://www.m--x--m.net).
This kind of system creates a lot of value by pairing up potential vendors to customers. It has the same kind of power than Google Adwords, except it is more adapted to offer a person does not know yet it needs.
This works for consumer and companies. There are a lot of issues to be solved (e.g. privacy, technology) but it can bring tremendous values to SMB while solving the spam problem (if you know who might need a plane ticket, why send a spam offering to sell it to everyone).
Posted by: Nicolas Toper | October 16, 2008 8:02 AM
Semantic You - having the internet and technology work towards your needs/desires. For example, with the current economic situation and people being laid off, how could your experience and knowledge be leveraged? How could you convey your full history to the world as opposed to the conventional resume? Today we are using Linked-IN and other social networking tools to provide some information but this is just an electronic way of doing something old. Take it to the next step and automatically match employeer/employee.
Posted by: Larry | October 16, 2008 8:13 AM
I really would like to see Semantic mobility app to be developed. It's a trend for handheld device to handle more stuff, since it is personal, secure and convenient.
Posted by: Keven | October 16, 2008 9:07 AM
I think it is too easy to assume Web 3.0 will be the semantic web. More than likely, it will be a combination of different technologies and trends.
My Web 3.0 app:
The Virtual Web. A virtual world similar to Second Life with the virtual shopping aspects of Kinset and a similar interface for browsing web portals and popular blogs where you walk down a street filled with houses that represent popular blogs on a topic -- like a street housing ReadWriteWeb, Mashable, TechCrunch, etc. And all of this would run through a browser, of course.
Talk about personalizing the web and making it one big chat channel ;)
Posted by: Dan | October 16, 2008 9:18 AM
Top-down semantic web applications are cool, but ultimately for the Semantic Web to realize its full potential, businesses and publishers need to have incentive to embed authoritative semantic data in their sites.
So for my Web 3.0 app, I'd like to see the search engines -- Google, Yahoo!, Microsoft in particular because of their commercial scale -- provide that incentive (and to do it in a way that will help make search a better experience).
Two ideas for that:
1. A SemanticRank, analogous to PageRank, that rates sites based on how good their semantic metadata is and how it links to and is linked to other references on the web. This rank would help sites improve their position in search engine results. Not to replace PageRank, but to complement it. I know, there are spam issues to be dealt with here, but I think the folks who could implement this could be smart enough to have the algorithm and weighting balance that.
2. A visual indicator on search results that the pages have semantic data. Maybe even shortcuts to some of the semantic relationships that are exposed on that page.
Both of these ideas are about giving more visibility to companies and publishers who make the effort to treat semantic data as an explicit part of their search marketing and content strategy.
Once critical mass is achieved with lots and lots of sites including semantic data, the potential applications that can be built to leverage it will blossom like Spring.
Posted by: Scott Brinker | October 16, 2008 10:10 AM
@Scott
Pretend you are "The Wolf" (or any other SEO asshat). Do you realize how great an unstandardized, non-centralized, and essentially ad hoc system is for me"
The need for a "semanticRank" (however poor an analogy) is strong. This idea itself requires lots of thought and iteration.
Posted by: Eric | October 16, 2008 10:23 AM
I work for a denver web design company and I had this tool exists. Good to know now.
Posted by: Luis Pastrulo | October 16, 2008 11:05 AM
@Eric
You're right, it's not a trivial problem to reward the deployment of good semantic data while minimizing semantic spam ("spem"?). And you're also right that it will require lots of thought and iteration.
But, asked for my Web 3.0 app wish, I'd like to see some of those iterations begin.
I don't think the analogy of SemanticRank to PageRank is necessarily a poor one, as one of the best ways to leverage the relational aspect of semantic metadata -- and combat SEO, um, black hats -- will likely be to look at the implied authority/value of semantic data by its internal and external link structure.
I had another idea, which you'll probably choke on as well: ad networks -- take Google AdSense for an example -- could give publishers a larger cut of revenue if the sites include good semantic data that can be used to improve the accuracy of so-called semantic advertising. I know, I know... lots of potential for abuse there too. But I don't believe it's impossible to establish the right incentives.
Posted by: Scott Brinker | October 16, 2008 12:00 PM
1) Semantic Flash app for robotic vacuum cleaner? Flash app connects to internet, then founds new heuristic algorithms for cleaning, optimal use of energy, etc... Why Flash? Adobe have cash enough to put Flash and AS3 in my fridge, which lead us to:
2) Fridge that collects data from net, and prepares recipes for dinner, takes care about my health, making me better developer and my wife better housewife.
3) Calculating next big things and trends.
4) Calculating financial trends
Posted by: krdr | October 16, 2008 12:05 PM
How those all semantic web applications (I'm referring only at these) will be developed?
It seems hard for somebody to start a service based on "semantic_something" because the technology isn't there yet...and if it is cost a ton of money (Powerset put a lot of work and money in thier technology...some other companies too).
I think we need to have the tools first, right? And then start building something "semantic"...
Web 2.0 happened also because the tools were there (free or low cost)...everybody who wanted and learned a bit could use them.
How's that for "semantic"?
Posted by: Mircea | October 16, 2008 1:20 PM
Develop the tools first so people can build on top of them...
But what do I know? I might be wrong though...
Posted by: Mircea | October 16, 2008 1:21 PM
Cisco's big insight was to foresee the need for network plumbers & plumbing equipment before the need was materialized in the marketplace. Similarly, I think one of the first killer apps that will arise for the Semantic Web will be "information plumbing"
A lot of data will always be stored in databases both for legacy and performance reasons: triples are good, but aren't appropriate for everything. But that doesn't mean we can't use the Semantic Web to query all those collections of databases as a single, merged, graph-based data source.
The Cisco of Web 3.0 will provide software that allows me to group together a bunch of data sources I want to query and treat them as a single graph query endpoint. Web Services and relational databases get wrapped in a graph query language like SPARQL, and light-weight ontologies describe their structure as triples. Rules are applied to normalize different data structures when data flows through these wrappers (epoch time -> UTF, for example).
This killer App will have a benevolent dictator that defines the "blessed" ontologies for Books, Contacts, Products, Prices, Flights, and so on, to keep arguments over semantics from drowning the technology before it spreads. But anyone who wants to can create their own ontology and use that instead, just as anyone who writes a wrapper for a relational schema or web service can map onto any ontology they choose.
The result is that I can rope together { Amazon, Half.com, eBay } web services as a consumer and query all three at once. Or as a small business owner I could rope together all my parts suppliers catalogues. As an academic I can simultaneously query the ACM, IEEE libraries, but also include a database of papers my department keeps. The user is in control over the point of aggregation instead of having to wait for a company to hand-code a solution (like a movie's near me search) to do it for them.
The key here is to realize that solving the plumbing problem takes us half-way toward realizing the semantic web. If we were all still trying to figure out how to wire up and configure our home networks, we wouldn't have had the time to figure out how to stream video to our homes. The first killer Web 3.0 apps is the one that solves the plumbing problem enough that all the other killer apps down the road don't have to re-invent the wheel.
Posted by: Edward Benson | October 16, 2008 2:04 PM
what kind of 'Web 3.0' app do I wish to see developed ?
Leverages ‘indexed content’ via an ingestion process that extracts relationships of grammar (parts of speech, natural language processing, entity enrichment, linguistic semantic analysis ) combined with Social Network statistical algorithms to generate contextual connections and interrelationships across enterprise data repositories.
Uses knowledge network mapping to analyze, visualize, and discover hidden patterns and relationships within and across large sets of unstructured data
Posted by: Steve Ardire | October 16, 2008 5:35 PM
Richard,
As for Web 3.0 and defrag, the most crucial Web 2.0 element that really need to be defragged is the online identities. We have too many identities in varied Web 2.0 sites. It is hard to manage these identities, let it alone to manage the information associated to these scattered identities. This identity defragment would be a critical Web 3.0 application.
Semantic technologies will certainly be helpful and useful to approach this goal. However, I think it might be too quick to claim that Web 3.0 equals to Semantic Web. We need first to solve several other problems such as the redefinition of online identities before we may really reach the era of Semantic Web. Web 3.0 is simply too soon to realize the whole vision of Semantic Web.
Yihong
Posted by: Yihong Ding
|
October 16, 2008 8:18 PM
this Semantic Crap is Boring beyond Belief* Maybe it will be a Minor Element of the next Web iteration but WEB3D.0 is the way to go!!
I think c0ol Apps like Cooliris.com + Neave.tv are the way to go!!
Cheers! ;)) Peace*
Posted by: BillyWarhol | October 16, 2008 8:36 PM
Seems like there are some doubters out there about what semantic web technologies can actually deliver in the real world. The bottom line is that semantic technologies have a lot to prove themselves out in an open world scenario (i.e. www), but they become increasingly relevant in a closed world where the semantics are more easily understood and known, which is where zAgile begins.
zAgile offers products today utilizing semantic web technologies. We enable software engineering teams to unify any software delivery environment, which typically consists of many tools with different terminologies (bug tracking, version control, project management, testing tools, etc.) and processes/methodologies. The zAgile repository interfaces with all other tool repositories, and the zAgile application provides an "umbrella" integrated view of everything in the software delivery process. All of is contextual unification is not really possible today without semantic technologies. And unifying teams is particularly important with distributed development teams. For distributed development teams, you also need collaboration tools such as wikis and instant messaging. zAgile semantically integrates those tools as well so that you can achieve an information repository across all your software delivery and collaboration environments. Check out www.zagile.com for a 10 minute video overview and fill out the information form for more info.
Posted by: Andrew Lampitt | October 16, 2008 11:31 PM
So far web 3.0 looks like a bit of a waste, mind you web 2.0 is mostly just buzz as well.
Posted by: Christopher Ross | October 17, 2008 4:50 AM
I'd like a Semantic Backchannel.
If developed, I would probably then want a Semantic Backchannel Aggregator.
Posted by: Michael R. Bernstein
|
October 17, 2008 2:18 PM
Many things I've been said, and my vote goes for the grading device. I'm in the middle of essays, and I need this one now!
Seriously, one thing that I've been dreaming for a long time would be to have all the rushes from a film (or series), directors' cut, fanfic and everything graded by the makers and possibly some viewers, tagged, images and motion, and the speech heard, etc. all used to allow people to make they own version of a film, more or less automatically.
The easiest feature (not very Web 3) would be to let directors make one movie, possibly rated R-to-PG. If a scene needs to be censored for younger viewers, replace it â but allow a room of adults to see what was intended for them. Higher on the relevance scale, allow cornier version of a movie, or more action packed ones; some with a twisted ending, and some with lame jokes; one with the writers' mad idea just in the middle that makes the whole thing feel worst then Lost, and one my ex could understand. After that, allow users to manipulate th
Posted by: Bertil Hatt
|
October 17, 2008 5:00 PM
Frankly, I'm surprised to see part of the list as it is. If we're eventually going to decouple form from function and apply that to communications then we need to stop thinking in terms of how we're sending a message, e.g., I'm going to call Lisa, email her, send her a video or text message, etc.
Imagine sending a video message and allowing the recipient to decide the form it comes in. Inherent in video (as we record it today) is audio. Add voice recognition software and you have audio translated into text.
Why do this? Well, you might prefer picture and video messages from attractive people or people who have something visually interesting to offer and audio messages from people with radio voices. Perhaps you're concerned about battery power on your mobile device so you choose text.
Anyway, that's my two cents and yes I'm familiar with the traditional concept of unified messaging.
Posted by: Lisa Morgan | October 17, 2008 5:08 PM
Thanks everyone for the great comments. The 3 winners of the Defrag passes are:
Scott Brinker
Jesse Wilkins
Edward Benson
We'll email you 3 about the prize.
Posted by: Richard MacManus
|
October 18, 2008 12:42 AM
I know I'm late to the comment game, but can I just say how much I hate the Nova Spivack 4 quad diagram being used to describe: Web 1.0, Web 2.0, Web 3.0, and Web 4.0.
I think that it gives the impression that many Semantic (big "S") enthusiasts have that the Social Web was a big detour. It's saying that we have to stall progression on the Social Web front and go back to connecting information.
How many of you suffer from "social media fatigue"?? I think that semantic technologies should be helping to better connect human beings together, not just figuring out how to empower human beings or machines to connect together information. Remember, as Web 2.0 has proved, the world wide web isn't just a bigger and better online library, it is also the new telephone and the new campfire.
The next generation of the web has to be as much of a revolution as the Read/Write web. We need the next generation of the web to not only be smart, but to also be more human.
Marta Strickland
Editor, ThreeMinds
http://threeminds.organic.com
Posted by: Marta Strickland | October 20, 2008 1:19 PM