RSS Feed

New feature proposals

Doug just reminded me that the "100 1 " at the beginning of thesaurus entries is completely ignored by VantagePoint.  I had thought the weighting code was still working but rarely used...which explains why some of my more advanced cleanup macros never worked very well.

I've had several customers ask me about creating groups from their own list of terms in a text file.  For the customer, I'd like to simplify the format.  I know we've had discussions in the past about making the "100 1 " optional, but I don't know if we actually followed through.

If we wanted to keep some power in there, there are several variables (besides weighting) we might want to turn on:

1- Case sensitive or not (default NO)

2- Treat as regex or not (default YES)

3- Anchor (contains, begins with, ends with, or exact match)

Anything else?  Do we need different toggles for "power format" vs. Simple format.  Do we keep using "**" or switch to something else?  XML?

It would be easy to add a second supported thesaurus format simply by using a different top-level-term indicator (~~ or ~-~ or whatever instead of **).  Then that new format wouldn't have the 100 1 on each sub-item line.

Whoops, forgot...

Also, we already have case sensitivity switches built in, but it would be easy to add a regex or not switch to the whole file as well.

XML would also work, but I prefer not using that here simply for easy of implementation, although it would be a more flexible format for future changes.

How much of an impact will this have on the Thesaurus Editor?

For the thesaurus editor, I suppose we would need to add toggles (checkboxes) for turning on/off the regex and case options.

Also, what level are we talking about setting these options?  Thesaurus file, top-level item, or sub-item?  For simplicity, I was thinking at the file level, but I can see how the item level would work also.  I'm just trying to imagine a situation where I would be mixing things.

While we're pitching ideas for thesaurus format revisions, I'd like to be able to set an 'Allow multiple matches' option for 'Full match' thesauri.

With this option 'on' in a 'Country-to-continent' thesaurus, 'Russia' would match both 'Asia' and 'Europe' top level items.

Would running a thesaurus that makes use of the weighting code have a high performance (speed) cost?