DonationCoder.com Software > Find And Run Robot

How do you handle large amounts of files? Are tags possible?

(1/2) > >>

sabot7726:
Hi all!

I've recently started using FARR at the 911 dispatch center I work at. Over the years, several thousand documents worth of contact information, policies and procedures, maps, daily recaps etc have stacked up. Initially all this information was in binders all over every nook and cranny of the room, but we've recently digitized it all and stored it on a shared network drive.

Right now FARR does a great job of finding something if we happened to include it in the filename or folder it resides in, my question is if there's a way to add tags to the files or folders so more stuff shows up while searching?

For example if there's a document called noisepolicy.doc: Some people who only vaguely know what they're looking for might try searching for construction, animals, party etc, but the doc they're looking for might only show up if they type noise.

How is everyone handling large amounts of files where you may not necessarily know the exact file name?

d4ni:
You might want to install something like Windows Desktop Search or Google Desktop Search; those will also scan the contents of the files.

Shades:
There is also a free (commercial and non-commercial alike) search engine available from IBM together with Yahoo! in case you are not allowed to 'cripple' each workstation with a desktop search client, especially when the information required is not stored on them but on a server. To me server storage sounds like the most logical thing to do because of ease of backup and those desktop search clients don't "(re-)index" the (LAN) network to a standstill. After all 911 is a service where every second counts, is it not?

The free search engine can be found here.

An open source variation of a search engine is also available in the form of Lucene from the Apache foundation and is based on Java. Configuring this one will be harder than the IBM option (which promises to be up and running in minutes), but it does not have the limit of 500.000 documents.

EDIT: added link to Lucene project

mouser:
This is actually a really good question.

In FARR you can build lists of files inside aliases so they are always easy to find by typing the alias (keyword) name. [note you can easily add files to an alias group by right clicking on a result or opening alias and dropping files into it]

But this isn't a perfect solution to cases where you'd like to have multiple tags per file, etc.

I could pretty easily add feature that let you combine aliases as tags by doing +aliasname so that would get you part of the way there.


The desktop search tools aren't going to really help you much with this.  As you've stated clear, you really need a tool that lets you put tags on files.

Perhaps what we really need is a proper system for integrating FARR with a general purpose file "tagger" tool.  If someone finds a free file tagging tool that might be interfaced to FARR let us know.  Alternatively

d4ni:
Desktop search tools aren't going to help with tagging, true. However, I do believe they will have a much higher hit rate than FARR as they also scan the contents of files. Actually, they can guarantee that higher hit rate as they also search the file name. In the example he mentioned about the noise policy; the document itself will most likely contain the keywords 'party', 'construction' etcetera, hence the file noisepolicy.doc will pop up when searching, right?

My point is, manually tagging the files is a very time consuming task, and perhaps not even needed when the tags you want to set are already present in the text :) I am questioning the effectiveness of FARR in this situation versus some kind desktop search tool with the advantages I mentioned (which does not take away from the fact FARR is awesome for launching stuff ;))

Navigation

[0] Message Index

[#] Next page