Xerox's research arm Monday showcased its latest innovations, including erasable paper and tools that make documents "smart" by adding a deeper meaning to words and images.
Since its establishment in 1970, the research centre, has created numerous technologies available on PCs today, including Ethernet, the graphical user interface (GUI) and the computer mouse. The laboratory, with other Xerox research facilities, is now trying to help its parent company and other start-ups by focusing on printing and other innovations to access, use and secure electronic documents.
Scientists demonstrated paper that can be reused after printed text automatically deletes itself from the paper's surface within 24 hours. Instead of trashing or recycling after one use, a single piece of paper can be used a second time, and reused up to 100 times, said Eric Shrader, area manager at PARC.
Predictions that paper would disappear in the 1970s in favour of electronic documents was false, Shrader said. As the number of electronic documents produced increases, about two to five pages are printed in the office for daily use, like e-mail messages and web pages, which are discarded or recycled after being read. Reusable paper reduces that waste and is environmentally safe, and reuse also helps reduce overall printing and paper costs, Shrader said.
The paper contains specially-coded molecules that create a print after being exposed to ultraviolet light emitted from a thin bar in a printer. The molecule readjusts itself within 24 hours to its original form to delete the print, or heat can readjust the molecule instantly. The molecule was developed by Xerox.
The ultraviolet bar itself is very small, so it can be used in mobile printers, Shrader said. The technology could also be useful for network printing.
For now, the technology prints effectively only in black and white. Although good for everyday prints, ink remains a better option for high-quality prints. Shrader couldn't project when reusable paper or UV printers would reach consumers.
Scientists also demonstrated technologies to make documents more intelligent by providing a deeper meaning to text and images. This is done by cross-referencing similar data and images mined off the Internet and incorporating other sources like e-mail messages and corporate networks.
The company's hybrid categorisation technology, for example, can provide more accurate answers to questions than search engines, said Christopher Dance, manager for data mining research at Xerox. The company is developing the Factspotter data mining software, which takes data sources and builds a structure around images and words that make it easier for users to locate specific information.
For example, if a user adds a travel diary and uploads an image from Cuzco in Peru, Factspotter can categorise a document and provide context to words and images based on information from the diary and similar images mined off the web.
The technology could be used in search engines, but it is more relevant for areas like the legal discovery process during litigation, Dance said.
PARC is also developing algorithms to better secure data on a document through its intelligent redaction technology, which automates the process of blacking out certain parts of a document considered confidential. For example, when there is a legal subpoena of medical records, information like diseases need to be blacked out, said Jessica Staddon, manager of security and privacy research at PARC. Based on rules established to protect the data, the technology weeds out information to blank out, like drug use and mental health conditions.
In medical records, for example, the technology can automatically detect words to redact based on the name of a drug or medical organisations. The current data redaction rates show about 75 percent accuracy, Staddon said.
Data redaction processes are currently inefficient as they require domain experts and hours of manual labour, Staddon said. The intelligent redaction technology computerises the process and provides the expertise based on artificial intelligence software tools and algorithms developed by PARC.