Access important files from anywhereSeptember 22, 2006 9:08 AM PDT
Search engine for personal files to be bundled in PCs
- Related Stories
-
A virtual vault for personal data
July 20, 2005
PCs bundled with the company's application will appear before the end of the year, said CEO Tom Garland, though the company wouldn't say which computer makers will include the program. Lexar is putting it on USB drives.
EverNote will also release a mobile version of its software before the end of the year. A new version of the desktop software will come out later this month.
The company's application lets consumers search a broad array of files on their desktop: PowerPoint slides, Excel pages, Word documents, handwritten notes and even photographs. An integrated optical character recognition tool translates road signs and other words in photos for the search engine.
A search on "mustard," for instance, turned up a photo containing a jar of Heinz mustard. A search on a person's first name turned up a photo with a conference badge, handwritten notes containing the name, and e-mails containing the name in type.
The trick is that EverNote doesn't search everything on a hard drive. Consumers have to tag the items they want to save and be searched later. Thus, even heavy users may only have a few hundred items in their EverNote database.
Microsoft, among others, has for years tinkered with desktop search. These search engines search entire hard drives, which makes the task significantly harder. EverNote defends its practice of only searching a small subset of hand-selected data by noting that individuals, in most circumstances, will not want to search the mass of stuff on their hard drive.
"You wouldn't want to search all of your spam," said Stepan Pachikov, president and chief product architect at the company.
Consumers can attach descriptive terms to a saved item to make it easier to find later. The application, however, will automatically classify the document by type (handwriting, e-mail, photo, a clip from the Web, etc.), date, location where it was created and other factors.
Time, it turns out, is the most important bit of identifying information, Pachikov said. People generally remember when they took a photograph or had a meeting. Thus, they often search by an approximate date.
"The second most important is location," he said. "If you met me and wanted to find a photograph, you wouldn't think, 'Was it France or Australia?'"
Third, people rely on the type of media, which means they will start a search by focusing only on handwritten notes if they remember the information was taken down by hand.
Another future version of the product will be for the Web. Consumers will clip what they want to save and post it to a Web site, sort of like a collage.
See more CNET content tagged:
search engine,
photograph,
search,
note,
hard drive
- Bryan
http://www.BryanCFleming.com
Everywhere else? *nix users can use the built-in (nowadays) utility known as [i]locate[/i] (which is a mini search engine in and of itself... it uses a database to save time and I/O while searching).
Windows and OSX users I guess may have a use for it, but really... isn't this just a utility for people too computer-illiterate to use what's already there?
The key to the article is that this app supports tagging of items to be searched. This allows a user to 'search' items that were typically excluded by other engines since other engines require data be represented in a format supported by the engine. For instance, imagine that I want to locate a song I really like that I've dumped into my mp3 collection. I might tag it as 'mom's favorite song.' Later, when I search for stuff related to mom, this song pops up.
Of all the files on your disk, it turns out that your desktop search engine only supports indexing / searching a fraction. Evernote claims users don't care about searching through most of the stuff on disk anyway.
The problem with this app is that *users* are required to do their own tagging - something that will never happen. This is a problem with the new 'web 2.0' search engines that allow users to build a catalog of items to be searched - something that probably won't catch on either.