CERN Inputting

1. Introduction

This page summarizes the current inputting workflow practices at CERN. The page is not exhaustive, it shows typical examples only.

For the similar practices at SLAC, see the SlacInputting page.

2. Harvesting from arXiv

The workflow is a bit complex due to the requirement to store harvested records both in CDS and in ALEPH500.

The overall schema is presented below. For more details, please see the dedicated CDS wiki page CDS.CERNDocumentServerArxivHarvesting.

3. CERN CLIC Notes: an example of collection cleaning

This example shows how the enrichment of a CERN series such as the CERN CLIC Notes is typically being performed by the CERN library.

3.1. CLIC Notes: overview of actions and actors

This diagram shows an overview of various procedures and actors involved in the cleaning and enrichment of the CERN CLIC Notes collection. The procedures will be described in detail below.

Note that in peak times the library can have up to ten students and apprentices available.

3.2 CLIC Notes: detailed action list

Here is a detailed overview of actions being performed during the CLIC Notes collection enrichment and cleanup. The library maintains this as an MS Excel sheet. (This version is based on Cath's 2008-01-31 file.) Note that the sheet contains a status column that tracks the progress.

CERN CLIC Notes

Actions for CLIC-Notes

Out of these, the "core" actions are (i) the author and affiliation completion with Giva (23), (ii) the publication reference completion (12-13), (iii) the fulltext file completion (14-16). Every core action is specified with a dedicated workflow diagram below.

Note that some actions use programs and scripts referred to under ComparisonSlacFermilabDesyCernEnrichmentScripts page. The overview of used tools:

3.3. CLIC Notes: adding publication references

This diagram shows how the publication references are being completed during the CLIC Notes collection enrichment. (Actions 12/-13/ from the above list.)

Note that an "internal note" tag (595) is used to track success/failure of the enrichment process.

3.4. CLIC Notes: adding fulltext files

This diagram shows how the fulltext files are being collected during the CLIC Notes collection enrichment. (Actions 14/-16/ from the above list.)

Note that the collection is not considered completely cleaned unless all articles have a fulltext file: either an electronic file is found and submitted or a paper copy is scanned. Hunting for fulltext is done in three ways: Google etc, CERN physical archives and rescan, or author asked by email. The choice is ad hoc and depends on the most probable successful way (e.g. year of publication).

The email communication process with authors is currently not assisted by any tool. Note that for some collections such as CERN Theses the communication with authors represented the amount of ~500 emails. A tracking tool would be profitable here.

3.5. CLIC Notes: maintenance, the day after

Once the collection of CLIC Notes is enriched and cleaned, the further maintenance of this collection is performed by the autocheck tool C30 (see SystemDesignBibExport) that alerts cataloguers automatically in case bad things are detected.

4. Importing JHEP publication references and OA fulltext files

For SISSA/IOP, all JHEP and JINST articles are taken with the uploader tool (C4, aka SystemDesignBibConvert). The fulltext file is atached for all OA articles, the link is "SISSA/IOP Open Access article".

This workflow is representative for all importations using the uploader tool (C4).

5. Importing from publishers: Elsevier, APS, IEEE

In the past the uploader tool (C4) was often used to import publication references (e.g. from arXiv). Now the process usually starts at the publishers' web sites and is highly manual (screen-scraping).

For publishers Elsevier, APS, IEEE, we make the search on publishers sites for "affiliation CERN" and current year in all the journals from these publishers and we input new articles and update what we already have in CDS with the publication references. After that we attach the file of the publication for CERN affiliated articles published by APS and IEEE (authorisation for CERN papers), the link is "APS Published version, local copy" or "IEEE Published version, local copy". Other metadata are updated at the same time, e.g. no affiliation in APS, so we need to add it.

We harvest the publishers 4 times in a year, of course we have to eliminate manually what we have done before, but we always make the same search and the same sorting ("most recent first") so not too difficult. We harvest and input manually (more reliable than with an uploader matching because of LaTeX and formulas problems).

Once per year, usually in January, we check all the articles for the previous year, to be sure nothing was missed.

6. Importing conferences via uploader: JaCoW, AIP, etc

Many conference proceedings are imported by using the same uploader tool (C4) as for the JHEP publication references and fulltext files described above, so it is not necessary to present the workflow here.

For example, conferences published by JACoW are imported via a dedicated C4 configuration to treat the "Open Archive Format", and all EPAC, PAC, DIPAC, FEL, ICALEPCS..... contributions are taken.

Another example, for electronic conference proceedings published by AIP, the Library Physicist makes a selection on subject, and then the uploader tool is again used to import them.

7. Importing conferences via hard copy: BEAUTY 2006, etc

For conference proceedings that we receive in the hard copy paper form, such as BEAUTY 2006, a manual process in needed. We look for contributions with affiliation CERN and input new CERN articles, and we update all contributions that we already have (CERN or not CERN) attached with this conference.

8. Inputting authors and affiliations with Giva

For collaboration papers written by many authors, the following procedure is being used to extract the author names together with their affiliations: