h a l f b a k e r y
Naturally low in facts.
idea:
add, search, annotate, link, view, overview, recent, by name, best, random
meta:
news, help, about, links, report a problem
account:
Browse anonymously,
or get an account
and write.
Login
Create account.
|
|
|
An advanced search option in Google to search only for certain kinds of texts. We already have taxonomies like Google News and Google Scholar but those seem to be based on the sources rather than the characteristics of the texts.
Have the indexed websites categorised by the structure of texts like:
-
poetrical
- lyrical
- journalistic
- marketing
- business
- discussion
- scientific
- essay
- instruction
- encyclopedic
and so on.
Combinations should also be possible. Like business + journalistic.
Google Custom Search: Google Picks
http://www.google.c...xamples/GooglePicks
Try one of these custom search engines, look for "refine results for" above your results. [krelnik, Mar 20 2008]
[link]
|
| |
Who decides how a site is categorised? |
|
| |
This is currently under development by Google, though in a much more sophisticated form than you suggest. Early prototypes are already being tested by a number of international intelligence agencies to create profiles of suspected terrorists and terror organisations. Coming soon to a search engine near you. |
|
| |
Search the acronyms NGIR, PROSE, SIB & ALVIS for more info. |
|
| |
Forced to guess, I'd say I read it in Newsweek about 18 months ago. |
|
| |
It would be useful to limit searches to certain categories.+ |
|
| |
It already is implemented by Google, but only appears in "custom search engines" right now. Click my link, pick one of the search engines featured and do a search. Look at the top of your results page for "Refine results for". How many refinements you get, and how detailed they are, depends on the author of the search engine. I don't think they allow combinations at this time. |
|
| |
What GooglePicks does is linking to specialised search engines. Like the Google News and Google Blogs, it judges content by the sources. |
|
| |
To answer angel also, I propose that that the content is automatically categorised based upon characteristics of the text itself. Wether the source is known for that particular kind of text or not. |
|
| |
The challenge is to code all conventions for certain texts into a formula. The Haiku search engine would be the easiest to implement, I assume. |
|
| |
(-) We don't really know what these vague categories mean, and have no chance to check whether or not a search engine does a good job implementing them. Given a choice, I'd rather have more underlying linguistic characteristics. (Is it written in first person, second, or third? What's the sentence length or "reading level"? Which syntactic roles do my search terms appear in? ...) |
|
| |