April 25th, 2012

Ann Vole

The 3 sides of web 3.0

The 3 sides of web 3.0
"web 2.0" had three sides: 1) content was sourced by the users rather then the publishers. An example is Wikipedia were the content is written, edited, and validated by the users. Another is YouTube were the users uploaded stuff and voted on it based on use. 2) the content was linked no matter what the source. Browsers did lots of that via cookies and search engines. For example, I can search for videos on a topic and will get videos from a wide range of possible sources 3) it has not happened quite as much as predicted but resources are not located by URL any more but rather by key words. It sort of happened with browsers featuring a search engine space beside the address bar. The original vision was all internet information to be well labeled and always stay in one spot. This vision is not practical as the content is not made by experts.
So, how is Web 3.0 different then 2.0? Web 3.0 is where the content is created by computers for individual users. Of course there is also the issue of personal privacy but individualized content can still be created using artificial intelligence. The three sides to doing this are 1) the text and other content needs to be labeled by meaning. Because the content it is based on has been created by users via web 2.0, computers need to figure this content out. One example is new translation programs that analyze translated documents to figure out the probable meaning of words and phrases. People did not create the programming and translation, computers did. YouTube is trying a pair of programs for subtitling videos. One listens to the audio and tries to determine the language spoken and makes subtitles of the dialog in text form. The other can take that text, and translate it to other languages. I tried using this pair to translate a Korean film into English subtitles. The results were hilariously bad but I was watching the film with English subtitles already added so I could compare results... and it is still amazingly close except for using financial terms (for an animated film staring animals). 2) Artificial intelligence needs to be used actively during browsing. One example is the advertisements presented on your page based on the limited information available to the web page server (last visited page, link source, user's location and operating system and browser, timing of text entry and movement of pointer). Another example is the option to find similar images in an image search result. Of course the computer has no idea what the image actually is but it can guess based on the image file details and the use of that image within the web page and all the text and links near the image within the page. 3) user-based programming will be used to enhance the experience of the user. This also has to be available on wimpy devices like cell phones, reader devices, vehicle navigation systems, and gaming platforms. This is where cloud networks and creative use of available CPU resources within networks becomes the new hardware science. Internet-connected gaming worlds are an example of this where all the images are created on the user's computer but the content is made up of actions of users around the globe in real time.