Heuristic needed to extrapolate RIF from large database

[egcasco]

Hello everyone,

I need someone who could kindly guide me, or suggest me to whom I should ask, in importing Endnote references from a large database.

This database includes only web-links (27500, about) although it is not homogeneous, in the sense that part of the links are pointing to:
- directly to webpages who contain one or more Endnote reference;
- directly to documents, which may not include their own Endnote reference;
- indirectly to both webpages and documents (including and not including Endnote references), for which an heuristic to discriminate Endnote reference by level of depth is needed. **
- directly or indirectly to webpages which are not relevant.

Although I need only a 1-level in depth analysis, it may be useful to being able to discriminate by levels.

This database is available in any format, such as html, txt, cvs, or whatever.

Thank you all in advance for your replies!
egcasco

** i.e. a link in the db points to a page which contains another link to a page who contain the Endnote reference, this is an example of a 1-level depth discrimination.