| |
  |
| Radar Networks Launches Twine Semantic Web Tagging Service |
|
 |
|
posted by Editor
on Monday October 29, @10:31PM
|
|
|
|
|
This article in MIT Technology Review describes Twine, a new web service being launched by startup Radar Networks that registers W3C-compliant metadata for user data such as e-mail, web searches, calendar entries, notes, spreadsheets, documents, and presentations. In addition to relying on manual tagging by users to identify information, Twine applies natural-language processing and learning algorithms to extract key concepts from text so that submitted data can be tagged automatically. This may help to overcome some of the metadata credibility challenges that critics of the semantic web have raised. Twine also uses social graph analysis to assist users in searching for information based on the connections between people and their information. In this scheme, the system establishes the statistical relevance of one user (or node) to another -- the "farther away" a node is, the less relevant it is to a particular user's search.
Twine is currently accepting registrations for beta testers, and plans to open fully to the public in 2008.
|
|
|
|
|
| Penumbra Enables Peer-To-Peer Networks Directly Via WiFi |
|
|
|
posted by Editor
on Monday January 29, @07:26PM
|
|
|
|
|
Penumbra is a prototype of a broadcast-based wireless network protocol that enables peer-to-peer networks to be constructed directly via WiFi connections. Normally, WiFi is used by mobile computing devices to communicate with "hot spot" access points that are physically connected to the Internet. With Penumbra, mobile computing devices can use WiFi to establish peer-to-peer connections directly with each other. Packets sent to a peer running Penumbra can be forwarded to other peers that are in range of that peer, and so on. The result is a free, distributed 54Mbps network that is based entirely on WiFi connections. Penumbra is designed to coexist with existing WEP- or WPA-encrypted private networks, so small changes are needed in the operating system's WiFi drivers to allow unencrypted packets to be sent and received without disturbing any existing association (see modified Linux drivers). An early version of a Usermode Daemon is also available, although it does little more than upload and store encrypted packets right now. An RFC for the Penumbra protocol is here.
|
|
|
|
|
| Will Social Profiles Be Killer App For The Semantic Web? |
|
|
|
posted by Editor
on Wednesday November 01, @06:18PM
|
|
|
|
|
|
As a potential successor to the World Wide Web, the Semantic Web benefits from worthy design goals and distinguished credentials due to its sponsorship by the father of the original Web, Tim Berners-Lee. But standards are only as useful as the applications that are built on them, and the mainstream adoption of the Semantic Web has been limited by the lack of compelling applications that fully exploit its capabilities. Matt Robson writes that social networking tools based on the Semantic Web could deliver such breakthrough applications: "With social profiles, every profile contains the same kind of information. Unlike the unstructured or semi-structured web, social profiles are highly structured, both in terms of representing what the components of the profile actually refer to, and also representing explicit friendship and group belonging relationships. In a sense, this isn't just symbolic guesswork. Rather, there are real object representations, real fields with meta-data, and real relationships between objects, and other sets of objects. Further, the benefits of the Semantic Web are more readily achieved with social networking, since users participating in social networks generally take on the burden of "structuring themselves", whereas when people search the web with current tools, they usually aren't structuring the results, or tagging components". Some pointers to resources on social networks and the Semantic Web are here.
|
|
|
|
|
| Using Grid Computing To Pick Stocks |
|
|
|
posted by Editor
on Monday October 02, @06:49PM
|
|
|
|
|
Anonymous Coward writes "I have read a news piece on gstock.com the recent grid computing project analyzing stock behavior. I have seen many grid computing projects, I even contributed my CPU to some of them but this is the first time I see grid computing being used for something so "practical" as increasing profits from stock trading. When I leave work at night and see the big empty office buildings I always wonder how much of humanity problems could have been solved if all those idle desktops were put into use. We could have been cancer free, rich with friends in galaxies far far away." GStock harnesses thousands of volunteered computers into a virtual supercomputer that generates daily Buy and Sell recommendations. Its developers claim 70% accuracy based on performance to date.
|
|
|
|
|
| Seven Barriers To Success For The Semantic Web |
|
|
|
posted by Editor
on Monday July 24, @06:29PM
|
|
|
|
|
|
This article on xml.com describes seven barriers to the adoption of the Semantic Web:
A lot of the semantic data being created today is inaccessible or unusable through the formal Semantic Web standards, or is simply not useful;
There is little interest from developers in building Semantic Web applications (as shown by googling the terms "rdf" and "semantic web", which produces barely enough hits to rise above noise level compared with established development standards);
The Semantic Web's relative complexity represents a high barrier to entry, compared to systems like AJAX, which are built up from simpler technologies that are easy to adopt;
It is not yet clear enough to typical users exactly what problem the Semantic Web solves;
The Semantic Web depends on the contributions of people, which are inherently undependable (and the use of trust metrics to maintain the integrity of tags will only add to its complexity);
The Semantic Web would be much easier if there a central, top-down ontology were adopted, which is unlikely to happen in the egalitarian culture of the web;
The Semantic Web vision is fundamentally predicated on philanthropy, and until some sort of killer application appears that exploits Semantic Web standards, there will be no economic incentive for organizations to share their metadata.
|
|
|
|
|
| Browster Previews Web Pages Behind Links |
|
|
|
posted by Editor
on Tuesday June 13, @03:57PM
|
|
|
|
|
Browster is a browser plug-in that provides real-time previews of the links on a given page. Any time the user hovers over a link, a preview of the full web page (not a thumbnail) pointed to by that link pops up (see screenshot). Also, the mouse wheel can be used to automatically jump from one link preview to the next. Browser is a free download, and currently works with Internet Explorer and Firefox.
|
|
|
|
|
| Surfing The Web While Offline: Is It Possible? |
|
|
|
posted by Editor
on Monday May 01, @06:30PM
|
|
|
|
|
MIT Technology Review has a brief review of Webaroo, a program and service that lets users search and browse web pages without a connection. Webaroo's servers continously scan the web, analyze web pages, and automatically select the subset of pages with the most content value in the least storage size. These pages are then assembled into topic-specific "web packs" that are transmitted every time the user syncs with the Webaroo service. The reviewer found that the concept works, but demands a great deal of disk space, even as the resulting coverage is somewhat sparse.
|
|
|
|
|
posted by Editor
on Tuesday April 11, @04:35PM
|
|
|
|
|
|
zeitgeist writes "Mapping Cyberspace is a neat overview of numerous attempts to visualize the complex datasets and interconnections in cyberspace, from the first hand-drawn maps of Arpanet, to hyperlink maps, to visualizations of USENET discussions, and more. (I'm not on the publisher's payroll; I just spent two hours in the bookstore looking at it.:-)" Yes, this is the definitive resource for web visualization. The authors have also published a more recent book called Atlas of Cyberspaces, which has a companion web site with an impressive gallery of all kinds of web visualizations, including network topology, info landscapes, info spaces, web site maps, and surf maps.
|
|
|
|
|
| Meta-Surfing With Bitty Browser-in-Browser |
|
|
|
posted by Editor
on Wednesday March 15, @12:51PM
|
|
|
|
|
|
The Bitty Browser is a fully functional browser instance that can be embedded in web pages. Similar to picture-in-picture for television, visitors can use the Bitty browser to navigate the web without leaving the current site. Some of the possible applications include live navigable blogrolls; embedded search boxes; and remote feeds of live content. There are several ways to use Bitty, such as by adding some script code to a particular page, or personalizing Google.
|
|
|
|
|
| Linux Desktop Distributor Collaborating On Developing Semantic Web |
|
|
|
posted by Editor
on Wednesday March 01, @07:05PM
|
|
|
|
|
|
Linux distributor Mandriva has announced that it will participate in the NEPOMUK Social Semantic Desktop project sponsored by the European Union. The goal of the project is to apply semantic web concepts on local desktops so that users can annotate and link arbitrary information across different media types, file formats, and applications, allowing them to transform ideas into semantic information. The implementation will involve integrating wiki technology with an annotation mechanism based on Resource Description Framework (RDF). The project has a total budget of 17 million euros, of which 1.8 million euros will go do Mandriva's Edge-IT subsidiary. As a key provider of Linux desktop distributions, Mandriva is in a good position to drive the adoption of the semantic web in mainstream productivity applications.
|
|
|
|
|
|
|
