I'm spending too much time correcting the Bayesian transaction matcher

[John Ralls]

On Mar 1, 2013, at 10:38 AM, Chris Lonsberry <chris.lonsberry at gmail.com> wrote:

> Anyone? Am I the only person who has experienced the Bayesian matching
> being mostly wrong? Is there an easier way to correct?
> 
> Chris
> 
> 
> On Fri, Feb 22, 2013 at 8:25 AM, Chris Lonsberry
> <chris.lonsberry at gmail.com>wrote:
> 
>> Hi all,
>> 
>> I am the type of user who imports all of my transactions rather than
>> entering them by hand.
>> 
>> I have been trying to train the Bayesian transaction matcher, but it is
>> getting to be a rather painful process. It seems to match incorrectly about
>> as often as it matches correctly.
>> 
>> That wouldn't really be much of a problem if it were easy to fix. I don't
>> mind categorizing my transactions by hand. BUT, the method I have been
>> using to correct the imported transactions is painfully slow. I have to
>> double click each one and then scroll to find the right expense account and
>> click again to select it.
>> 
>> If, on the other hand, I just let it import everything to Imbalance, I can
>> move through the transactions on my keyboard and have them all properly
>> categorized in about 1/10th the time.
>> 
>> So, am I doing the corrections wrong? Is there a more efficient (ideally
>> keyboard only) way to accomplish this task?
>> 
>> If not, I might have to reset the Bayesian matching so that it always puts
>> everything in Imbalance and accept the fact that I will need to manually
>> assign them all.

No, you've got it about right. There's no other interface.

The bayesian matching works primarily off of the description field. Are many of
the descriptions supplied by your bank/credit card company similar?

Regards,
John Ralls