In the move from SPIRES to INSPIRE the handling of references and citations is not yet complete. Naturally we are working on fixing this and present below a list of all the problems we know about. Some are problems that were always in SPIRES some are new to INSPIRE.

1. Citation links displayed with a paper don't give reliable results or references that are visible on a paper's reference list don't show up as citations for the cited paper. This is a problem of the indexing function of INSPIRE, not the data. Fixing this is a top priority for us and we are currently working on it. We will post here when it is fixed.

2. Extracted reference list on a temporary entry misses lots of references. This is a data problem, often caused by our system looking first for %%CITATION and then upon finding it, ignoring all other kinds of references. This is how SPIRES always behaved but it is more noticeable now because, due to a staffing shortage we are not able to process papers as quickly and also because %%CITATION is not displayed in INSPIRE formats (so people have not used it in citing newer papers). This problem of missing references has now been fixed. We are going to re-introduce %%CITATION in INSPIRE as we expand our coverage of citations to books, conference papers and unpublished preprints, something SPIRES never did. Of course, some missing references have nothing to do with this, our script simply failed to find them. We are currently working on a more robust reference extractor.

2. (variant by KS) Extracted reference list on a temporary entry misses lots of references. This is a data problem, often caused by SPIRES looking first for %%CITATION and then upon finding it, ignoring all other kinds of references. This is how SPIRES always behaved. INSPIRE is much more flexible making %%CITATION in most cases unnecessary therefore the tag is not displayed in INSPIRE formats. But during the transition period references are extracted by SPIRES which failed badly when using reference lists created from INSPIRE. This problem of missing references has now been fixed. We are going to re-introduce %%CITATION in INSPIRE as we expand our coverage of citations to books, conference papers and unpublished preprints, something SPIRES never did. Of course, some missing references have nothing to do with this, our script simply failed to find them. We are currently working on a more robust reference extractor.

3. Citations to conference papers disappearing. Several years ago in SPIRES we began to assign a pseudo-journal publication note, Conf.Proc., to conference papers that would enable us to track citations to conference papers. This worked quite well and, as an example, we were able to track the 700+ citations to Salam's famous, unpublished paper http://inspirehep.net/record/53083?ln=en. In the transfer of records from SPIRES to INSPIRE, this Conf.Proc. publication note was not handled correctly and the citations did not connect to the papers. This has now been fixed. We are working on new reference extraction methods that will allow us to track citations to conference papers more robustly; we will do this retrospectively.

4. Self-citations can't be removed. In SPIRES the citesummary format displayed citation results for an author with all self cites removed. This feature was not duplicated in INSPIRE but will be soon.

5. There has been a large reference discrepancy between Spires and Inspire in many journals with Volume letters in the REF code. Example is NUCIA,A46,1. Inspire requires these to be exact. Over the years, Spires has had written more flexible search functions where it could capture more flexible entries of the cite, such as NUCIA,46A,1, or more problematically NUCIA,46,1. In many many cases, there are colliding volume numbers and page numbers for different A,B,C,D or non lettered volumes where Spires counts references it should not. Much of this was cleaned up in Spires over the summer, but some discrepancies remain that have to be handled by more detailed attention. Details were documented here: https://twiki.cern.ch/twiki/bin/view/Inspire/SlacSpiresInspireReferences

6. Incorrectly formatted cites in Spires with ambiguous journal codens do not get counted in Spires. If they have a valid volume and page, this reference will be counted in Inspire. In some cases, this counts correctly, in many cases it does not.

7. We have continuing problems with delays or missed updates from curation of records in current Spires and the passing of data to Inspire. This will go away, of course, when we have completed the curation tools in Inspire. The symptom of this problem is to see the correction appear in the Spires record, and not in Inspire. We will not turn Spires off until this is corrected.

8. We have a number of records that are blocked from updating into Inspire due to special processing of large author lists, or are otherwise curated currently in Inspire (CMS notes). This will create a discrepancy in cite counts, for some records Inspire will be correct, and Spires incorrect. This is being addressed.

-- HeathOConnell - 31-Oct-2011

Edit | Attach | Watch | Print version | History: r4 < r3 < r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r4 - 2011-11-01 - KirstenSachs
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Inspire All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2023 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback