Bug 104020

Summary: Extract dates for KOrganizer automatically from emails
Product: [Applications] kontact Reporter: Kurt Bennater <curdyben>
Component: generalAssignee: kdepim bugs <pim-bugs-null>
Status: RESOLVED DUPLICATE    
Severity: wishlist CC: kde-2011.08
Priority: NOR    
Version First Reported In: unspecified   
Target Milestone: ---   
Platform: openSUSE   
OS: Linux   
Latest Commit: Version Fixed/Implemented In:
Sentry Crash Report:

Description Kurt Bennater 2005-04-16 17:36:54 UTC 
Version:            (using KDE KDE 3.4.0)
Installed from:    SuSE RPMs
OS:                Linux

I often receive emails containing the time and date of an appointment. When dragging them to the Calendar icon, a "new event" window pops up with the title of the mail. It would very handy if KOrganizer could scan through the body of the email for keywords like "today", "tomorrow" or dates and numbers which could indicate the time of the appointment. This will be not easy because it has to deal with non-standardized and language-dependent formulations.
In case you decide to implement this convenient feature, perhaps a spell-checker could decide first which language test yields the fewest number of errors; this could indicate the language of the email. Then the program would scan for keywords from a (user-defined) list for that given language in order to detect date and time.
If the result is not unique, it could either display a list with possible results or just decide for the first one. The result should preset the appropriate values in the "new event" dialog. Is that feasible?
Comment 1 Kurt Bennater 2005-05-02 18:48:59 UTC 
Some people from Stanford have actually been looking into an even more ambitious project, the fully automated extraction of events from emails, see the first two links at http://nlp.stanford.edu/courses/cs224n/2004/ . At least the extraction of dates seems to have worked well (as explained in http://nlp.stanford.edu/courses/cs224n/2004/jblack-final-report.pdf ).

A rudimentary perl tool for date extraction and a parser for normalizing the output can be found at http://www.shlrc.mq.edu.au/masters/students/rdennys/PerlAsst3.html ,
other perl tools include http://search.cpan.org/~mcmahon/Date-PeriodParser-0.03/PeriodParser.pm
(understands even expressions like "tomorrow") and for multi-language support: http://search.cpan.org/~gbarr/TimeDate-1.16/lib/Date/Parse.pm
Comment 2 Kurt Bennater 2005-05-09 15:38:40 UTC 
The KDE desktop search engine Kat includes code which can identify the language a document is written in. Perhaps the same works for emails? So I guess, at least some important parts for this feature are already there...
Comment 3 Dotan Cohen 2009-05-30 18:26:36 UTC 

*** This bug has been marked as a duplicate of bug 75663 ***