I have a fairly big database (~7000 documents) and when I search them I often look for specific links. Thus, for example if I want to search for Biodiversity and Pesticide I have no problem in doing that in DT. However, if I want to find Biodiversity NEAR/5 Pesticide I get ALL instances of Biodiversity and ALL instances of Pesticide plus the one or two instances of Biodiversity NEAR/5 Pesticide that I am interested in. For a 5 page document, that is no problem. For a 100 page document where biodiversity or pesticide is mentioned dozens of times, it is a real pain.
Thus, is it possible to SPECIFICALLY find, ONLY Biodiversity NEAR/5 Pesticide
I am using the latest version of DT Pro and the latest version of Mac Lion.
Yes, the searches work. (BTW, the NEAR operator isn’t Boolean, strictly speaking. George Boole worked out permutations of the AND, OR, and NOT operators.)
If you use the NEAR, BEFORE or AFTER operator in a query and get a list of search results, be assured that the criteria involved in that operator’s use were met in each document.
You observed a different issue. For the query, ‘biodiversity NEAR/5 pesticides’, all occurrences of ‘pesticide’ and ‘biodiversity’ are highlighted, whether or not the terms met the proximity criterion. That"s confusing, as the highlighting of a query term doesn’t necessarily imply that it met the proximity criterion.
It would be nice if only occurrences of the query terms that met the proximity criterion were highlighted. Unfortunately, to attempt that runs into nontrivial logical issues. The issue is trickier than might seem the case at first thought.
This is exactly an issue I was just searching for in the forums.
So as it stands, there is no way in DTP to identify the actual location in a document(s) when proximity searching? This is disappointing.
For a two factor search [ term1 near/X term2 ], once DTP has identified documents that carry these terms in the required proximity, could the different terms be highlighted in different colors - maybe to eyeball the results? Or have one term highlighted throughout, then you could search the other term - again to eyeball. Neither optimal. (Plus, more terms add more worms.)
I agree with Don - when you have a significant number of documents, it is unwieldy.
Re: Bill DeVille’s comment. If it doesn’t work the way I described, then I think in the help file you should make that clear.
There is a Window’s program, called DtSearch which does exactly what I am asking for here, so it is not an impossible task.
Re KRomig’s comment: I agree. Different colors would help.
Also:
Would (Diversity???pesticide) OR (pesticide???diversity) work, where ? is anything or nothing and the max length is the total number of ? (as opposed to a *, in which case anything goes)
There are quite a few strands on the forum about this now. It is frustrating because at present searching PDFs of more than a few pages is virtually impossible for all but names. It may be non-trivial to make it perfect but some improvement to make our databases usable would be much appreciated.
