DocumentCloud
DocumentCloud performs erratically when searching memos
Sometimes its entity analysis exceeds standard, other times it falls far short.
DocumentCloud only spotted about half the references of organizations mentioned in these memos from the Obama-Biden transition team -- although it performed better than our annotator on those it did catch.
For three of the six government agencies and organizations, it spotted all of the references from the test criteria and then some, picking up others completely missed by our own test creators. For example, it accurately found 64 references to "Ducks Unlimited" -- the test required it find only 43.
However, the other three groups it missed completely. Some, like the Global Privacy and Information Quality Working Group or its abbreviation, GPIQWG, are hidden in a muddle of acronyms. But it also completely missed "U.S. Department of Justice."
According to developer Jeremy Ashkenas, DocumentCloud's entity analysis service, OpenCalais, uses a complex set of lists and rules to identify names and organizations, and they may simply miss some entities.
A second pass through the system produced identical results.
Product:
//Company:
Investigative Reporters and Editors
//Version Tested:
N/A
//Release Date:
2009
//OS Tested:
//Cost:
Free
//Open Sourced:
Yes
//Demo Available:
No
//Obsolete:
No
Latest Reviews
-
Outwit Hub Pro can scrape almost any database, but learning curve steepens with complex datasets
-
Pdftotext excels at extracting data from conventional tables, stumbles with more complex tasks
-
For Monarch, high cost translates to unmatched spreadsheet conversion
-
DeskUNPDF stumbles with complex formatting
-
Helium Scraper a powerful extraction tool that rewards patience
Latest Tutorials
See more»
