isidore-en.md 38.5 KB
Newer Older
1
---
2
lang: en
3
description: Presentation of ISIDORE, the search engine for discovering publications, digital data and profiles of social sciences and humanities researchers from around the world.
4
5
6
7
---

# ISIDORE

8
## What is ISIDORE?
9

10
ISIDORE is a search engine for discovering and finding publications, digital data and the profiles of researchers in the social sciences and humanities (SSH) from around the world.
11

12
The full text of several million documents (articles, theses and dissertations, reports, datasets, web pages, database records, descriptions of archival holdings, etc.) and event announcements (seminars, conferences, etc.) can be searched. In addition, ISIDORE links these millions of documents together by enriching them with scientific concepts created by SSH research communities.
13

Nicolas Larrousse's avatar
Nicolas Larrousse committed
14
It is accessible on the Web through the portal [isidore.science](https://isidore.science).
15

16
It also offers scientific social network functionalities. As such, it falls into the category of search engines and assistants and offers many features to organize scientific monitoring.
17

18
Launched on December 8, 2010, ISIDORE is the result of a collaboration between the CNRS "very large equipment" Adonis (2007-2013), the Center for Direct Scientific Communication and the companies Antidot, Mondéca and Sword. It is currently developed, updated and operated by the TGIR Huma-Num.
19

20
References on the history of ISIDORE:
21
22

- Yannick Maignien, "ISIDORE, de l'interconnexion de données à l'intégration de services", Hyper Article en Ligne - Sciences de l'Homme et de la Société, [10670/1.k9lck9](https://isidore.science/document/10670/1.k9lck9)
23
- Stéphane Pouyllau et al, "Bilan 2011 de la plateforme ISIDORE et perspectives 2012-2015", MoDyCo, Modèles, Dynamiques, Corpus - UMR 7114, [10670/1.bqexsj](https://isidore.science/document/10670/1.bqexsj)
24
25
- Philippe Bourdenet, "L'espace documentaire en restructuration : l'évolution des services des bibliothèques universitaires", Le serveur TEL (thèses-en-ligne), [10670/1.lnieuv](https://isidore.science/document/10670/1.lnieuv)

26
## How does ISIDORE work?
27

28
ISIDORE harvests textual metadata and full text, enriches them and then indexes them. It uses the metadata of the documents as well as the full text. The goal is to analyze this information in order to enrich the document, to link them to the concepts of the scientific vocabularies (thesaurus, etc.) and to link them to the authors' identifiers (ORCID, IDRef, IDHAL, VIAF, etc.).
29

30
Several enrichments are performed:
31

Nicolas Larrousse's avatar
Nicolas Larrousse committed
32
- Semantic annotation: the words present in the metadata of the documents are compared to the entries of the vocabularies through an algorithm based on a morphological analysis of the terms. If an equivalence is found between a term from the document and an entry in one of the vocabularies, then the resource will be linked to that vocabulary entry. The vocabularies are multilingual and aligned with each other. Thus, the semantic annotation is multilingual.
33

34
- Disciplinary categorization: ISIDORE uses a semantic classifier that, after being trained on a reference corpus, categorizes all documents in ISIDORE into the SSH disciplines of the MORESS vocabulary. The classifier is trained with the help of the manual categorization completed by researchers in [HAL](https://hal.archives-ouvertes.fr/) when depositing their publications.
35

36
- Detection of the authors: ISIDORE detects the authors of the documents and enriches the author form (first name and last name) with the help of international (ORCID, VIAF, ISNI) and national (IDHAL, IDRef) author identifiers.
37

38
ISIDORE indexes, in its search engine:
39

40
41
42
43
44
- Document metadata;
- The full text (if it is available in open access) ;
- The semantic annotations ;
- Disciplinary classification;
- Author enrichment and normalization.
45

Nicolas Larrousse's avatar
Nicolas Larrousse committed
46
More information is available on [the "Vocabularies" page](https://isidore.science/vocabularies) of ISIDORE.
47

48
### Can ISIDORE index multilingual documents and data?
49

50
Yes. Since 2015, documents and datasets in English, Spanish
51
and French are indexed, enriched and linked to scientific repositories by ISIDORE (metadata and full text). For full text outside these three languages, it is indexed in the language of the document but in that cas no enrichment is done.
Nicolas Larrousse's avatar
Nicolas Larrousse committed
52
For more information, you can consult our blog post on the subject: [Isidore speaks English, sino también español et toujours en français](https://humanum.hypotheses.org/921).
53

54
### How often is ISIDORE updated?
55

56
57
58
59
60
IISIDORE is updated, incrementally, on average once a
month. Why this delay? In addition to harvesting and indexing documents, 
ISIDORE enriches them with concepts from
scientific repositories (thesauri, taxonomies, etc.). This semantic enrichment is
automatic and allows us to offer you suggestions for reading to help you discover documents other than those 
61
you were looking for. This requires a certain amount of processing and calculation time.
62
63
64
The updates of the documents with which you are associated, which will thus be proposed 
 in your user account as documents to be claimed, will also be done monthly.
 
65

66
## How to use ISIDORE?
67

68
ISIDORE offers several tools to search, discover, collect and organize the contents it indexes:
69

70
### The isidore.science portal
71

72
The [isidore.science](https://isidore.science) portal is a website in three languages that provides a [relevance search engine](https://isidore.science) that can be used with several query methods.
73

74
75
76
- By default, ISIDORE searches for all the words in a query posed by the user by removing empty words ("of", "the", "the",  "the", etc.);
- It is possible to search for a document with a complete sentence or a group of words by using quotation marks around the sentence or word group, for example: "direction of consciousness" will search for exactly this expression. Thus, in this case, the "of" will not be considered as an empty word;

77

78
79
80
81
#### Search operators
Several boolean search operators are available in
ISIDORE. Note that the syntax of the operators is important in
ISIDORE, they are always in UPPERCASE (e.g. AND):
82

83
84
- AND: the intersection will find the terms (or set of terms) common to the query.
    For example:
85
    - consciousness AND gender
Nicolas Larrousse's avatar
Nicolas Larrousse committed
86
    - "cold war" AND migration
87
- OR: the union will find the terms belonging to both sets of terms, or to one or the other.
88
89
    For example:
    - "semantic web" OR "web 3.0"
90
- EXCEPT (NOT): the exclusion will reduce the noise by excluding terms. For example:
91
    - revolution NOT French
92
93
94
95
- NEAR(n.): the NEAR(n.) operator (i.e. "close to") will link terms by indicating a value "n." of proximity between them. It works like an AND with n. word(s) between the terms. The value "n." indicates the number of words that separate the two terms. NEAR also works without the value n. and is in this case equal to a NEAR(10), i.e. 10 words between the searched terms (standard spacing).
    - house NEAR(4) nobility : searches for house and nobility with
        a proximity of 4 words

96

97
#### Sorting of search results
98

99
By default, in [isidore.science](https://isidore.science), the results are sorted by semantic relevance. It is possible to change the sorting of the search results to:
100

Nicolas Larrousse's avatar
Nicolas Larrousse committed
101
- sorting by novelty
102
103
104
- sorting by author's name in alphabetical order
- sorting by author's name in reverse alphabetical order
- sort by ascending date
Nicolas Larrousse's avatar
Nicolas Larrousse committed
105
- sorting by decreasing date
106

107
VVery soon, two more options will also be available:
108

Nicolas Larrousse's avatar
Nicolas Larrousse committed
109
- sorting on the title by alphabetical order
110
- sorting on the title by reverse alphabetical order
111

112
### Advanced Search
113

114
115
An advanced search is also available at [https://isidore.science/as](https://isidore.science/as) and also accessible from
the first page of the [portal](https://isidore.science/as).
116

Edward Gray's avatar
Edward Gray committed
117
### Personal space for researchers
118

Nicolas Larrousse's avatar
Nicolas Larrousse committed
119
Isidore.science offers a personal space for researchers allowing them to:
120

121
- collect, classify and organize the documents found;
122
123
- gather all their scientific production in order to edit it in a personal profile page;
- follow the productions of colleagues;
124
- record and publish queries and their results for monitoring purposes;
125
- create bibliographies that can be exported to Zotero.
126

127
### The APIs of isidore.science
128

Edward Gray's avatar
Edward Gray committed
129
The [isidore.science search engine APIs](https://api.isidore.science) are available through the GET method on HTTP or HTTPS.
130
They provide a fast, accurate and reliable query service for ISIDORE data with advanced search features (auto-completion, spell checking, multi-criteria, boolean and faceted searches, sorting, aggregation of answers, etc).
131

132
Each request to the engine is submitted by means of a URI pointing to a specific web service. The response is a stream in XML (default format) or JSON format.
133

Edward Gray's avatar
Edward Gray committed
134
The [isidore.science API web page](https://api.isidore.science) details all the commands available for the different services available.
135

136
### Enriched metadata for *Linked Open Data*.
137

138
ISIDORE's metadata, ontologies and vocabularies are available in a triplet repository [RDF (Resource Description Framework) or *TripleStore*](https://en.wikipedia.org/wiki/Resource_Description_Framework), thus placing ISIDORE data in the *Linked Open Data*. A web interface for querying using the SPARQL language and browsing the ISIDORE graph is available via:
139

140
141
- A documented SPARQL query interface and presentation of the ISIDORE data model: https://isidore.science/sqe  
- The basic Virtuoso software interface: https://isidore.science/sparql
142

143
In the ISIDORE *TripleStore*, the main vocabularies for structuring information are:
144

145
- RDF and RDFS
146
- Dublin Core Element Set
147
- Dublin Core TERMS
148
149
150
151
152
153
154
- SIOC
- FOAF
- OWL
- SKOS
- ORE
- DBPEDIA

155
(The complete list is available at <https://isidore.science/sparql?nsdecl>)
156
157


158
### Complementarity between ISIDORE and Zotero
159

160
#### Use from ISIDORE of the Zotero connector to feed its bibliographic database
161

162
ISIDORE is compatible with Zotero. The references of documents can be imported on two levels as soon as the user has installed [the Zotero connector](https://www.zotero.org/download/) in his browser:
163

164
- On the page listing the results of a search,
165
- On the page listing the results of a search, in the page displaying a document.
166

167
#### Using the ISIDORE search connector from Zotero
168

169
Zotero (Linux, MacOS, Windows client) uses search engines to search or complete bibliographic references directly from the Zotero interface. We propose here two ISIDORE connectors for Zotero that make it possible to use ISIDORE from author search.
170

171
172
173
174
By adding ISIDORE to Zotero you can:

- complete references from a search on the author's name: this is the "ISIDORE, help me find what he/she has published."
- find documents in which the author is cited: this is the "ISIDORE, what do you have on the author?"
175
176


177
These [connectors and installation documentation are available on the TGIR Huma-Num GitLab](https://gitlab.huma-num.fr/spouyllau/ISIDORtero).
178

179
### Use of RSS feeds
180

181
ISIDORE can propose its research results in the form of RSS feeds in order to feed scientific monitoring software (including Zotero for example), research notebooks, etc. The RSS feeds created in ISIDORE are updated, like all the contents of the search engine, approximately once a month during the general update of the ISIDORE contents. Thus, it is possible to follow, from Zotero, the update of the ISIDORE documents resulting from the registered queries.
182

183
To do so, access your personal space (login required), and click "My queries" to see your registered queries:
184

185
![My Image](media/isidore.png)
186

187
For a registered query, you have to click on the pictogram "Request
Edward Gray's avatar
Edward Gray committed
188
RSS feed of the query" available on the right ![My Image](media/isidore-rss-001.png){: style="width:170px"} and to copy the link with ![My Image](media/isidore-requeteRSS.png){: style="width:120px"}.
189

190
The copied link is in the form: `https://isidore.science/feed/lt3913`.
191

192
If your browser is equipped with a module for reading RSS feeds, this link can be used directly in your browser.
193
For our example, we will continue with Zotero.
194

195
196
In Zotero, you have to choose: New feed > From URI:

197

198
![My Image](media/zot-001.png){: style="width:60%;margin-left:20%"}
199

200
Then add the url of the feed provided by ISIDORE (N.B. When using Safari under MacOS, take care to remove the mention "feed:" from
201
202
the url). Then paste it in "URL" of the Zotero RSS feed creation window, example below:

203

204
![My Image](media/zot-002.png)
205

206
Then you have to give a title to your feed, for example:
207
"isidore.science - Query on ...".
208

209
## What can be found in ISIDORE?
210

211
### Organization of documents and data in ISIDORE
212

213
ISIDORE contains several million documents in SSH that are harvested, enriched with scientific references and indexed. They are organized into:
214

215
216
217
- Research documents and data (archives, raw materials, photographs, films, datasets, statistics, etc.), identified in the ISIDORE ontology by: http://isidore.science/class/primaires
- Published documents and data (articles, books, dissertations and theses, reports, etc.), identified in the ISIDORE ontology by: http://isidore.science/class/secondaires
- Scientific events (conferences, study days, etc.), identified in the ISIDORE ontology by: http://isidore.science/class/evenementielles
218
219


220
For a large number of SSH disciplines, ISIDORE makes it possible to search documents coming from the main publication platforms worldwide, as well as a large number of digitized collections from national, university and
221
municipal libraries.
222

223
For advanced search uses, the [ISIDORE advanced search](https://isidore.science/as) offers, for example, the possibility of searching for documents between two dates and by discipline or by collections.
224

Nicolas Larrousse's avatar
Nicolas Larrousse committed
225
The main publication platforms (journals and books) present in ISIDORE are:
226
227
228

- OpenEdition
- Cairn
229
- Perseus
230
231
232
233
234
- Erudit
- Oapen
- Redalyc
- Scielo Books

Edward Gray's avatar
Edward Gray committed
235
The complete list of collections containing publications can be obtained by querying [the ISIDORE Triple Store](https://isidore.science/sqe) with the [following SPARQL request](https://isidore.science/sparql?query=SELECT+*+WHERE+%7B%0D%0A%3Fs+rdf%3Atype+%3Chttp%3A%2F%2Fisidore.science%2Fclass%2FCollection%3E.%0D%0A%3Fs+rdf%3Atype+%3Chttp%3A%2F%2Fisidore.science%2Fclass%2Fpublications%3E.%0D%0A%3Fs+dcterms%3Atitle+%3Ftitre%0D%0A%7D+ORDER+BY+ASC%28%3Ftitre%29&format=text%2Fhtml&debug=on&timeout=0) :
236
237
238
239
240

```
SELECT * WHERE {
 ?s rdf:type <http://isidore.science/class/Collection>.
 ?s rdf:type <http://isidore.science/class/publications>.
241
242
 ?s dcterms:title ?title
} ORDER BY ASC(?title)
243
```
Nicolas Larrousse's avatar
Nicolas Larrousse committed
244
The main digital libraries (municipal, national, etc.) present in ISIDORE are:
245
246

- Gallica
247
- Selene
248
249
250
251
252
253
254
255
256
257
- E-rara
- NuBIS
- Octaviana
- Burgerbibliothek
- Berkeley Library Digital Collections
- Argonnaute
- BNE
- Cornell
- Didόmena

Edward Gray's avatar
Edward Gray committed
258
The complete list of collections containing archival holdings and book collections can be obtained by querying [the ISIDORE Triple Store](https://isidore.science/sqe) with the [following SPARQL request](https://isidore.science/sparql/?default-graph-uri=&query=SELECT+*+WHERE+%7B%0D%0A%3Fs+rdf%3Atype+%3Chttp%3A%2F%2Fisidore.science%2Fclass%2FCollection%3E.%0D%0A%3Fs+rdf%3Atype+%3Chttp%3A%2F%2Fisidore.science%2Fclass%2Fprimaires%3E.%0D%0A%3Fs+dcterms%3Atitle+%3Ftitre%0D%0A%7D+ORDER+BY+ASC%28%3Ftitre%29&format=text%2Fhtml&timeout=0&debug=on) :
259
260
261
262
263

```
SELECT * WHERE {
 ?s rdf:type <http://isidore.science/class/Collection>.
 ?s rdf:type <http://isidore.science/class/primaires>.
264
265
 ?s dcterms:title ?title
} ORDER BY ASC(?title)
266
267
```

268
### Indexing of the main data platforms in SHS
269

270
ISIDORE harvests and indexes the contents of many SSH data platforms, allowing researchers to group all their data in their user profile. We encourage researchers, for their research programs, to use platforms offering open interoperability devices and protocols to present documentary and scientific metadata.
271

272
The main data platforms (sources, archives but also publications) are harvested by ISIDORE.
273

Edward Gray's avatar
Edward Gray committed
274
The complete list of collections can be obtained by querying [the ISIDORE 3store](https://isidore.science/sqe) with the [following SPARQL request](https://isidore.science/sparql/?default-graph-uri=&query=SELECT+*+WHERE+%7B%0D%0A+%3Fs+rdf%3Atype+%3Chttp%3A%2F%2Fisidore.science%2Fclass%2FCollection%3E.%0D%0A+%3Fs+dcterms%3Atitle+%3Ftitre%0D%0A%7D+ORDER+BY+ASC%28%3Ftitre%29%0D%0A&format=text%2Fhtml&timeout=0&debug=on)
275
276
277
278

```
SELECT * WHERE {
 ?s rdf:type <http://isidore.science/class/Collection>.
279
280
 ?s dcterms:title ?title
} ORDER BY ASC(?title)
281
282
```

283
Please feel free to report any new source to us.
284

285
#### Can data deposited and documented in NAKALA be referenced by ISIDORE?
286

287
Yes, data deposited and documented in NAKALA can be
288
289
accessible in ISIDORE. NAKALA offers as standard the [OAI-PMH](https://en.wikipedia.org/wiki/Open_Archives_Initiative_Protocol_for_Metadata_Harvesting) interoperability protocol which allows for the harvesting of document metadata, and therefore for referencing, enrichment and indexation by ISIDORE.

290

291
292
However, referencing by OAI-PMH harvesting is not
automatic for the moment, in particular to allow users to prepare and organize their data and
293
data and metadata. To be referenced, simply request by email to be indexed in ISIDORE via <isidore-sources@huma-num.fr>.
294

295
#### How will scientific articles and images deposited in the HAL, HAL-SHS and MédiHAL open archive be accessible in ISIDORE?
296

297
All the files (PDF, illustrations, photographs, audio and video) deposited and documented in the open archive HAL, including HAL-SHS, as well as MédiHAL are automatically referenced in ISIDORE and indexed at the level of their metadata. All these documents and their notices are thus accessible through the various query interfaces of ISIDORE.
298

299
#### Can the data deposited in the Didómena (EHESS) warehouse be referenced by ISIDORE?
300

301
Yes, [Didómena](https://didomena.ehess.fr) (the research data warehouse of EHESS) offers OAI-PMH interoperability. Be careful, harvesting is not automatic. For your collection to be referenced, please provide us with the OAI-PMH access point via <isidore-sources@huma-num.fr>.
302

303
#### Can data deposited in Calames (ABES) be referenced by ISIDORE?
304

305
Yes, descriptions of archival holdings cataloged in [Calames](http://calames.abes.fr) (the catalog of archives and manuscripts of French university libraries) are indexed in ISIDORE. However, the EAD-XML standard, used in Calames, does not always allow an optimal documentary indexing: this mainly concerns the richness of the metadata. This is due to the logic of the EAD-XML standard in the encoding of information in the levels of description of the collections.
306

307
#### Can the data deposited in the Data.sciencespo warehouse be referenced by ISIDORE?
308

309
Yes, the data deposited and documented in [Data.sciencespo](https://data.sciencespo.fr) (Dataverse) offer interoperability in OAI-PMH. They are harvested automatically by ISIDORE.
310

311
#### Can the data deposited in the COCOON platform be referenced by ISIDORE?
312

313
314
Yes, the data deposited and documented in [the COCOON platform](https://cocoon.huma-num.fr) offer interoperability in OAI-PMH. This platform is automatically harvested by ISIDORE.

315

316
#### Can files and documents deposited in the European Zenodo platform be referenced by ISIDORE?
317

318
319
320
Yes, it is possible for ISIDORE to reference the files and
documents deposited and documented on the platform
Zenodo](https://zenodo.org).
321

322
323
324
325
The referencing is based on the principle of OAI-PMH harvesting on a
set of files and data (and thus their metadata) corresponding to one or more
identifier(s) corresponding to the "communities" identifiers in Zenodo (see https://developers.zenodo.org/#sets).
We can also group several Zenodo identifiers in the same
326
ISIDORE collection, allowing the depositors of several corpora 
327
328
deposited in Zenodo to group them in ISIDORE to give them more
visibility.
329

330

331
332
To add your Zenodo repositories in ISIDORE, [please send us
the URL
333
OAI-PMH](mailto:isidore-sources@huma-num.fr?subject=%22Je%20souhaiterai%20faire%20moissonner%20mes%20dépôts%20Zenodo%22)
334
of your repository (see <https://developers.zenodo.org/#oai-pmh>).
335
336


337
## How do I get data referenced by ISIDORE?
338

339
340
There are several ways to get data and documents referenced by
ISIDORE:
341

342
343
344
- Submit your data via [an XML stream of standardized metadata and using the OAI-PMH protocol](#how-to-signal-data-in-isidore-with-metadata-and-the-oai-pmh-protocol) associated with metadata in Dublin core format. This method is adapted for documentary databases, corpora, scientific archives and document/data libraries. As an example, [a tool such as Omeka (Classic or S) offers the OAI-PMH protocol via modules](#a-website-using-omeka-classic-and-omeka-s-can-be-referenced-by-isidore). 
This method is adapted to research program websites presenting document or data corpora, scientific blogs (except Hypotheses.org), and web pages in general.

345

346
These two methods are also often implemented by data publication tools (CMS, etc.), for example:
347

348
### Can a web site using Drupal be indexed by ISIDORE?
349

350
Yes, it is possible to have web pages generated by the Drupal CMS indexed by ISIDORE.
351
There are two ways to do this, depending on the nature of the
352
content of your pages:
353

354
- Either via the OAI-PMH protocol and in this case there are several
Edward Gray's avatar
Edward Gray committed
355
    modules for Drupal, see [OAI-PMH for Drupal](https://www.drupal.org/search/site/OAI-PMH?f%5B0%5D=ss_meta_type%3Amodule).
356
357
358
359
- Or via the use of a Dublin
    Core metadata structure in the web pages generated by Drupal using RDFa and a
    sitemap.xml. An article dedicated to this way of proceeding is
    available at the above address.
360

361
### Can a website using Omeka Classic and Omeka-S be referenced by ISIDORE?
362

363
Yes, Omeka *Classic* and Omeka S offer modules to expose metadata according to the OAI-PMH protocol:
364

365
366
- Module for [Omeka S](https://omeka.org/s/modules/OaiPmhRepository/)
- Module for [Omeka Classic](https://omeka.org/classic/docs/Plugins/OaiPmhRepository/)
367
368


369
### How to report data in ISIDORE with metadata and OAI-PMH protocol?
370

371
To report your data in ISIDORE using the
372
OAI-PMH protocol, you just have to:
373

374
375
376
- Prepare your data and metadata using the
    Documentary vocabulary Dublin Core Element Set or Dublin Core
    Terms, depending on the level of precision you want, and to
377
    make them accessible via [the OAI-PMH protocol](https://en.wikipedia.org/wiki/Open_Archives_Initiative_Protocol_for_Metadata_Harvesting);
378
- To organize and document the *Sets* in its OAI-PMH repository.
379
- To write to <isidore-sources@huma-num.fr> and give the address of the repository to Huma-Num.
380

381
#### Document sets in OAI-PMH: *Sets*
382

383
The OAI-PMH protocol makes it possible, through the creation of *Sets*, to bring together a coherent set of records whose perimeter makes sense from a scientific or editorial point of view and which is left to the discretion of the producer of the data.
384

385
386
It also makes it possible to define a hierarchy in the *Sets* with an inheritance mechanism by specifying
in the set name the name of the parent *Set* and the child *Set*, 
387
388
separated by the `:` character. ISIDORE is able to use these
*Sets* to limit harvesting to a set of records or to differentiate between different
389
data sources within the same warehouse.
390
391
The producer will therefore have to specify the harvesting methods that seem to be
appropriate in order to make the most of their resources within ISIDORE.
392
393
394
395
To do this, he must indicate the *Set* or *Sets*
concerned or a rule enabling the *Sets* to be taken into
account to be distinguished.

396

397
The *Sets* can present metadata, in the Dublin Core Element Set, which are specific to them. For example:
398

Edward Gray's avatar
Edward Gray committed
399
```xml
400
401
<set>
 <setSpec>OuvColl</setSpec>
402
 <setName>OuvColl</setName>
403
404
 <setDescription>
  <oai_dc:dc xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:dcterms="http://purl.org/dc/terms/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
405
   <dc:description>Research works distributed on Cairn.info</dc:description>
406
407
408
409
410
  </oai_dc:dc>
 </setDescription>
</set>
```

411
#### Records in OAI-PMH or *Records*:
412

413
414
415
In the ISIDORE framework, each OAI-PMH "record" corresponds to a document.
The ISIDORE harvester thus exploits the metadata described according to the
application profile defined by the Open Archive Initiative for the
416
Dublin Core Element Set (also known as Dublin Core "simple"). 
417
In addition, the harvester also collects the full-text document(s) whose URLs
418
419
420
421
422
423
 (beginning with `https://` or `http://`) are specified in the
`<dc:identifier>` element.

We recommend data producers to provide records that are as metadata-rich as possible
 since relevance in
ISIDORE favors the richest possible metadata. Fields such as:
424
425


Edward Gray's avatar
Edward Gray committed
426
```xml
427
428
429
430
431
<dcterms:description>
<dcterms:creator>
<dcterms:date>
```

432
are essential.
433

434
##### Example of a complete record according to the OAI-PMH protocol:
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450

```xml
<record>
 <header>
  <identifier>oai:halshs.archives-ouvertes.fr:halshs-00514304</identifier>
  <datestamp>2010-09-02T11:06:50Z</datestamp>
  <setSpec>halshs</setSpec>
  <setSpec>SHS:ECO</setSpec>
  <setSpec>SDV:BIO</setSpec>
  <setSpec>INFO:INFO_BT</setSpec>
  <setSpec>SDV:SA:AEP</setSpec>
  <setSpec>SDV:SA:STA</setSpec>
  <setSpec>CIRAD</setSpec>
  <setSpec>SHS</setSpec>
 </header>
 <metadata>
451
  <oai_dc:dc xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
452
453
454
455
456
  <dc:identifier>http://halshs.archives-ouvertes.fr/halshs-00514304/en/ </dc:identifier>
  <dc:identifier>http://halshs.archives-ouvertes.fr/docs/00/51/43/98/PDF/Regulation_GMO_pprint.pdf</dc:identifier>
  <dc:identifier>http://halshs.archives-ouvertes.fr/docs/00/51/43/98/PDF/ppt_nocmt_broader_regulation.pdf </dc:identifier>
  <dc:title>Broadening the scope of regulation: a prerequisite for a positive contribution of transgenic crop useto sustainable development</dc:title>
  <dc:creator>Fok, Michel</dc:creator>
457
  <dc:subject>[SHS:ECO] Humanities and Social Sciences/Economy and finances</dc:subject>
458
459
  <dc:subject>[SDV:BIO] Life Sciences/Biotechnology</dc:subject>
  <dc:subject>[INFO:INFO_BT] Computer Science/Biotechnology</dc:subject>
460
  <dc:subject>[SDV:SA:AEP] Life Sciences/Agricultural sciences/Agriculture, economy and politics</dc:subject>
461
462
463
464
465
466
467
468
469
470
  <dc:subject>[SDV:SA:STA] Life Sciences/Agricultural sciences/Sciences and technics of agriculture</dc:subject>
  <dc:subject>regulation</dc:subject>
  <dc:subject>coordination</dc:subject>
  <dc:subject>GMO</dc:subject>
  <dc:subject>biotechnology</dc:subject>
  <dc:subject>seed price</dc:subject>
  <dc:subject>research</dc:subject>
  <dc:subject>weed resistance</dc:subject>
  <dc:subject>pest complex shift</dc:subject>
  <dc:description>Ex-ante regulation of transgenic crop use generally prevails, before the authorization of commercial release.This kind of regulation addresses the concerns of biosafety and coexistence, under pressure of pros and/or cons of GMO. After fifteen years of large scale use of transgenic crops (notablysoybean and cotton) in various countries (USA, China, Brasil, India...), ecological and economic phenomena are observed and which could threaten the sustainable use of transgenic varieties. I advocate that the regulation scope must be extended so as to a) promote a systemic and coordinatedapproach of transgenic crop use, b) ensure seed purity with regard to the transgenic trait, c) maintain research on non-transgenic varieties, and d) warrant fair pricing of transgenic seeds.</dc:description>
471
  <dc:coverage>Montpelier</dc:coverage>
472
473
474
475
476
477
478
479
480
481
482
  <dc:coverage>France</dc:coverage>
  <dc:date>2010-08-29</dc:date>
  <dc:language>English</dc:language>
  <dc:type>proceeding with peer review</dc:type>
  <dc:source>Proceedings of Agro2010, the XIth ESA Congress</dc:source>
  <dc:source>Agro2010, the XIth ESA Congress</dc:source>
 </oai_dc:dc>
</metadata>
</record>
```

483
484
In addition to this description in the *Dublin Core Element Set*, each
record can be described in one or more metadata formats, the choice of which is left to the
485
the administrator of the OAI-PMH warehouse.
486

487

488
489
The ISIDORE harvester is able to use the *Dublin Core Terms* format and any XML schema allowing
full-text exposure (including TEI or EAD) thus improving its indexing.
490
491
The data producer will have to take care to scrupulously respect the specifications of the OAI-PMH protocol version 2.0, in particular as regards:

492

493
494
495
- The strict respect of the "datestamp" values in the *records* in order to synchronize the updates between the producer and ISIDORE;
- The good management of deleted data ([detail on the OAI-PMH protocol documentation](http://www.openarchives.org/OAI/openarchivesprotocol.html#DeletedRecords));
- In the case of a publisher's data warehouse or one of significant size, access to its OAI-PMH warehouse via the IP addresses of ISIDORE's OAI-PMH harvesters (harvesting reported by ISIDORE to its IT department).
496

497
We advise producers to regularly validate the compliance of their repository using, for example, the [tools of the Open archive
498
initiative](https://www.openarchives.org/pmh/tools/). Finally, we advise data producers to contact the Huma-Num team for any information requests.
499

500
### How to report data in ISIDORE with RDFa metadata?
501

502
RDFa can express a metadata structure according to the principles of the Semantic Web (RDF for *[Resource Description Framework](https://en.wikipedia.org/wiki/Resource_Description_Framework)*) in the HTML code of Web pages. The "a" in RDFa stands for "in
503
attributes", i.e. within the HTML code).
504

505
How to express metadata of a web page very simply by
506
507
508
509
using the [RDFa syntax](https://tcuvelier.developpez.com/tutoriels/web-semantique/rdfa/introduction/)
? For example, in a blog post published with WordPress. While there
exist [plugins to do this](https://wordpress.org/plugins/search/RDFa/),
the obsolescence of the latter can make it difficult to maintain them 
510
511
over time. Another solution is to implement RDFa in the
HTML code of the WordPress theme you have chosen. For this to be easy
512
513
514
and manageable over time, the simplest way is to use the HTML header
in order to place `<meta>` tags that will contain some metadata.

515
516

Expressing metadata according to the RDF model via the RDFa syntax allows
517
machines (mainly search engines and indexers) to better process information because it becomes more explicit: for a machine, a string can be a title or a summary; if you don't tell the machine that it's a title or a summary it
518
519
will not guess it. So, at the very least, it is possible to use the
tags to define an RDF structure that allows you to structure the minimal metadata
520
521
for example with the Dublin Core Element Set.

522
523
524
525
526

#### How to do it practically?

First of all, it is necessary to indicate in the DOCTYPE of the web page, that it will
contain information that will use the RDF model, so the
527
DOCTYPE will be:
528

Edward Gray's avatar
Edward Gray committed
529
```xml
530
531
532
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML+RDFa 1.0//EN" "http://www.w3.org/MarkUp/DTD/xhtml-rdfa-1.dtd">
```

533
534
The `<html>` tag must contain the addresses of the
ontology (via their *NameSpace XML*) which are used 
535
to "type" the information. RDFa - which places metadata in the Semantic Web, requires at least the use of RDF and RDF Schema ontologies and the Dublin Core Element Set (dc). It is possible to use in addition - in order to refine the metadata - the Dublin Core Terms (dcterms):
536

537

Edward Gray's avatar
Edward Gray committed
538
```xml
539
540
541
542
543
544
545
<html xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
xmlns:xsd="http://www.w3.org/2001/XMLSchema#"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:dcterms="http://purl.org/dc/terms/">
```

546
To encode more information, It is possible to use more
547
document ontologies:
548

549

Edward Gray's avatar
Edward Gray committed
550
```xml
551
552
553
554
555
556
557
558
559
560
561
<html
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:dcterms="http://purl.org/dc/terms/"
xmlns:skos="http://www.w3.org/2004/02/skos/core#"
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
xmlns:xsd="http://www.w3.org/2001/XMLSchema#"
xmlns:foaf="http://xmlns.com/foaf/0.1/"
xmlns:cc="http://creativecommons.org/ns#">
```

562
IIn the example above, [foaf](http://www.foaf-project.org/) is used to encode information about a person or object described by the metadata. The [CC](https://creativecommons.org) ontology is used to indicate which license, from the *Creative Commons*, applies to this content.
563

564
The RDFa structure through tags
565
566
567
in the `<head>` header of the HTML page. In a first step
, using a `<link>` tag, we will define the digital object to which the
RDF encoded information will be attached:
568

Edward Gray's avatar
Edward Gray committed
569
```xml
570
571
572
<link rel="dc:identifier" href="http://monblog.com/monbillet.html" />
```

573
574
This tag defines a container for the information that we are going to indicate using the `<meta>` tags. This container is
identified by a URI which is a URL, i.e. the address of the
575
page on the web.
576
577


578
The `<meta>` tags then define a set of metadata, which in our case is descriptive information about the blog post's web page:
579

Edward Gray's avatar
Edward Gray committed
580
```xml
581
582
583
<meta property="dc:title" content="The title of my post" />
<meta property="dc:creator" content="First name Last name of author 1" />
<meta property="dc:creator" content="First name Last name of author 2" />
584
<meta property="dcterms:created" content="2011-01-27" />
585
<meta property="dcterms:abstract" content="A descriptive summary of my page's content" xml:lang="en" />
586
<meta property="dcterms:abstract" content="A summary in english" xml:lang="en" />
587
588
<meta property="dc:subject" content="keyword 3" />
<meta property="dc:type" content="ticket" />
589
<meta property="dc:format" content="text/html" />
590
<meta property="dc:relation" content="A link to a complementary web page" />
591
592
```

593
Depending on the nature of the content of the web page, it is of course possible
594
595
596
to be more precise, more refined and more complete in the 
encoded information. For example, it would be wise to use the DC Terms vocabulary.

597

598
The DC Terms allow, for example, a precise form for a bibliographic reference of the content to be included:
599
600


Edward Gray's avatar
Edward Gray committed
601
```xml
602
<meta property="dcterms:bibliographicCitation" content="Put a bibliographic reference here" />
603
604
```

605
It would be possible to describe the entire text of a web page using the SIOC vocabulary [using the 
606
property](http://www.lespetitescases.net/rdfaiser-votre-blog-2-la-pratique).
607

608
609
610
It is also possible to link web pages together (to
define a corpus of authors for example) by using in the
DC Terms vocabulary the DC Terms property: `dcterms:isPartOf`.
611
612

```xml
613
<meta property="dcterms:isPartOf" content="URL of another web page" />
614
615
```

616
#### Creating the Sitemap
617

618
619
620
Once the RDFa encoding has been done in the HTML pages, you still need to create
a Sitemap XML file listing the pages you want ISIDORE to harvest and to submit the URL of this sitemap:

621

Edward Gray's avatar
Edward Gray committed
622
```xml
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
	<url>
		<loc>http://monsiteweb.com/</loc>
		<lastmod>2018-01-01</lastmod>
		<changefreq>monthly</changefreq>
		<priority>1.0</priority>
	</url>
	<url>
		<loc>http://monsiteweb.com/page1/</loc>
		<lastmod>2018-03-05</lastmod>
		<changefreq>weekly</changefreq>
		<priority>0.5</priority>
  </url>
</urlset>
```

639
640
641
642
643
644
645
646
647
648
It is possible to test the extraction that ISIDORE will do of your
RDFa metadata using the "ISIDORE on demand" application
available at <https://rd.isidore.science/ondemand/fr/rdfa.html>

## ISIDORE perimeter

### Why are some items not found in ISIDORE?

If you do not find all of your scientific production
in [ISIDORE](https://isidore.science), there may be several
649
explanations. It may be that your articles are published in 
650
journals that are not electronic or that do not make their articles available even
651
 long after they have been published. Since its
652
creation, [ISIDORE](https://isidore.science) favors open
653
access since indexing is better for articles available in
654
open access. Many electronic journals have made this choice through
655
656
portals such as Open Edition Journal (formerly Revues.org)
Érudit, Persée, and Cairn.info, Redalyc, OApen and 
657
and articles from these journals are therefore collected and indexed by
658
659
[ISIDORE](https://isidore.science).

660
661
It is also possible that your articles are published online, but not
on an electronic publishing platform (but a website), or on an electronic publishing platform
662
that does not allow indexing via the standard protocol 
663
(see the question and answer on OAI-PMH).
664

665
Other journals make their articles available, but only after an
666
embargo period. In this case,
667
668
[ISIDORE](https://isidore.science) indexes only the metadata
of the article. If you connect via your university library
669
, documentation center or via BibCNRS, 
670
you may still have access to these articles.
671

672
673
The collections indexed by
[ISIDORE](https://isidore.science) can be searched by using the engine itself and by
674
indicating that you want to search the collections.
675

676
It is also possible that your article is published as a PDF image,
677
in which case only the indexing by
Edward Gray's avatar
Edward Gray committed
678
[ISIDORE](https://isidore.science) will be allowed, but not its
679
full text indexing.
680

681
Lastly, it is possible that some of your articles are published in
Nicolas Larrousse's avatar
Nicolas Larrousse committed
682
journals that are not classified in SSH.
683

684
In all these cases, you can deposit your articles in an
685
open archive such as HAL (HAL-SHS in particular) which is also indexed by
Edward Gray's avatar
Edward Gray committed
686
[ISIDORE](https://isidore.science) or contact your
687
bu/documentation center.
688

689
If none of these cases correspond to your problem and you therefore think that there may be an error, you can send us an e-mail to isidore@huma-num.fr.
690

691
### Why are some books/chapters of books not reported in ISIDORE?
692

693
694
ISIDORE knows how to identify that a document is of the type "book", thus, there are
more than 500,000 books and book chapters reported in
695
696
ISIDORE.

697
698
It should be noted that there are relatively few platforms that publish online books in open access. ISIDORE indexes in SSH, for example, the
contents of book platforms such as:
699

700
701

- [OpenEdition Books](https://isidore.science/search/?collection=10670/3.szxq6s) (at the chapter level, and to flag them);
Edward Gray's avatar
Edward Gray committed
702
- [Scielo Books](https://isidore.science/search/?collection=10670/3.7oraz1) (Brazil);
703
- [OApen](https://isidore.science/search/?collection=10670/3.pwofj8) (Netherlands);
704
- [Erudit](https://isidore.science/s/collection?q=erudit) (Canada);
705
- ...
706

707
In addition, you can, in agreement with your publisher, deposit your work or
708
book or book chapters in the open archive
Edward Gray's avatar
Edward Gray committed
709
[HAL-SHS](https://halshs.archives-ouvertes.fr). It will then be indexed by
710
711
ISIDORE within the framework of the indexing of HAL-SHS and recognized as a book chapter.

712

713
### Why are some databases are not reported in ISIDORE?
714

715
Harvesting by ISIDORE requires standardized and normalized metadata exposure (documentary, scientific, etc.) (either using the OAI-PMH protocol or using an XML Sitemap and RDFa metadata, see above).
716

717
If you know of any databases that are not present in ISIDORE, please inform us so that we can check with their publishers/data producers.
718

719
## ISIDORE training courses
720

721
Here we list training courses, functional presentations and online self-training courses on the use of ISIDORE. Do not hesitate to let us know about any training session you would like to organize:
722

723
724
- The Urfist Méditerranée proposes a new e-learning training on Isidore](https://urfist.univ-cotedazur.fr/nouvelle-formation-en-ligne-une-initiation-a-isidore/) (March 2021)
- ["Isidore, my personal research assistant"](https://ig.hypotheses.org/2215) by Johanna Daniel (April 2020)