Closed bisoldi closed 9 years ago
Hi, thanks for opening this issue. You are right that stuff such as ages of persons should not be extracted. We will add a negative rule to catch expressions such as "in his/her \d\ds".
Thanks again, we'll keep you in the loop.
Great, thanks! We're just starting to get into implementing HeidelTime in our application and will post any other issues as they come up. I will let you close the issue as you see fit.
I copied and pasted the text of the article found here:
http://www.cnn.com/2015/07/15/europe/germany-nazi-death-camp-verdict/?iid=ob_article_footer_expansion&iref=obnetwork
Into the demo and tried it with all 4 "Document Types". It detected a reference to age (written in an informal manner) as references to a Date.
For example, the sentence "Groening, who's in his 90s, ..." detected "90's" as TYPE: DATE and VALUE: 199.
When written more formally (as one would expect from a professional, mainstream news source) as "Groening, who is over 90 years old" (or something similar), it doesn't detect anything (as I assume it shouldn't), however that's too common of a way of expressing age to be left alone I would think.