When using Classify, there are certain documents which should go to one group, but that group is not included on the list, and there are wrong groups listed. For example, a bill from a vender does not bring up the group for identifically formated bills from the same vendor. Is there a way to “retrain” Classify so that bill (and future similar bills) are linked to the right group?
Only by filing things manually. There is no retraining the internal AI.
- What is your group structure?
- Are you classifying only in the same database?
- Are you classifying based on contents or tags?
I only have one database. All classifying is within the same databail
I am using this for a personal filing system. I have groups (house, medical, taxes, etc.) and subgroups within each of those groups. All documents are filed by manually selecting the correct group from the Classify drop down list. For some items, like an AT&T bill, the correct group never appears. So I have to manually move that item into the correct group. But Classify never learns from that manual filing.
- Is this something you scanned?
- If so, did you do OCR on it?
For the most part, these are not scanned documents, but documents attached to emails. In any event, I always do an OCR on all documents before using Classify and moving them into the appropriate group.
I would make sure your backups are current then do a File > Rebuild Database on it. Don’t interrupt the process or let the machine power down while it’s working. See how it behaves after that.
OK. Will give it a try.
The only possibility is to exclude certain groups or documents from see also and/or classifying, see Info inspector. E.g. if the contents of a document in a group are not really similar to the contents of the other documents in the group, then it’s a good idea to exclude it.