isidore-en.md 40.6 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
23
- POUYLLAU, Stéphane, MINEL, Jean-Luc, CAPELLI, Laurent, SAURET, Nicolas, BUNEL, Mélanie, BAUDE, Olivier, JOUGUET, Hélène, BUSONERA, Pauline, & DESSEIGNE, Adrien. (2021). "ISIDORE celebrates its 10th anniversary". Zenodo. [10.5281/zenodo.5700008](https://doi.org/10.5281/zenodo.5700008)
- 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)
24
- 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)
25
- 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)
26

27

28
## How does ISIDORE work?
29

30
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.).
31

32
Several enrichments are performed:
33

Nicolas Larrousse's avatar
Nicolas Larrousse committed
34
- 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.
35

36
- 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.
37

38
- 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.
39

40
ISIDORE indexes, in its search engine:
41

42
43
44
45
46
- Document metadata;
- The full text (if it is available in open access) ;
- The semantic annotations ;
- Disciplinary classification;
- Author enrichment and normalization.
47

Nicolas Larrousse's avatar
Nicolas Larrousse committed
48
More information is available on [the "Vocabularies" page](https://isidore.science/vocabularies) of ISIDORE.
49

50
### Can ISIDORE index multilingual documents and data?
51

52
Yes. Since 2015, documents and datasets in English, Spanish
53
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
54
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).
55

56
### How often is ISIDORE updated?
57

58
59
ISIDORE 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 you were looking for. This requires a certain amount of processing and calculation time.
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.
60

61
62
The ISIDORE's change log is available on [https://isidore.science/releases](https://isidore.science/releases).

63
64
65
66
67
68
69
### What is the circuit for adding collections in ISIDORE?

Two scenarios:

- A research project, a team, a laboratory, a library can propose collections to be harvested by simple e-mail to <isidore-sources@huma-num.fr>. The Huma-Num team studies the request and exchanges with the requester in order to fully understand how the metadata and the data to be indexed are described. Most often, a first harvest and a first indexing and enrichment are carried out so that the requester can see and analyze how their data will be indexed in ISIDORE. Then, the exchanges potentially continue to adjust the indexing process as well as possible.

- The Huma-Num team identifies a data warehouse or a digital library and contacts the data producer or the structure that distributes this data to exchange and propose harvesting and indexing in ISIDORE. A first harvesting and a first indexing and enrichment are carried out so that the requester can see and analyze how their data will be indexed in ISIDORE. Then, the exchanges potentially continue to adjust the indexing process as well as possible.
70

71
## How to use ISIDORE?
72

73
ISIDORE offers several tools to search, discover, collect and organize the contents it indexes:
74

75
### The isidore.science portal
76

77
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.
78

79
80
81
- 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;

82

83
84
85
86
#### 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):
87

88
89
- AND: the intersection will find the terms (or set of terms) common to the query.
    For example:
90
    - consciousness AND gender
Nicolas Larrousse's avatar
Nicolas Larrousse committed
91
    - "cold war" AND migration
92
- OR: the union will find the terms belonging to both sets of terms, or to one or the other.
93
94
    For example:
    - "semantic web" OR "web 3.0"
95
- EXCEPT (NOT): the exclusion will reduce the noise by excluding terms. For example:
96
    - revolution NOT French
97
98
99
100
- 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

101

102
#### Sorting of search results
103

104
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:
105

Nicolas Larrousse's avatar
Nicolas Larrousse committed
106
- sorting by novelty
107
108
109
- 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
110
- sorting by decreasing date
111

112
VVery soon, two more options will also be available:
113

Nicolas Larrousse's avatar
Nicolas Larrousse committed
114
- sorting on the title by alphabetical order
115
- sorting on the title by reverse alphabetical order
116

117
### Advanced Search
118

119
120
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).
121

Edward Gray's avatar
Edward Gray committed
122
### Personal space for researchers
123

Nicolas Larrousse's avatar
Nicolas Larrousse committed
124
Isidore.science offers a personal space for researchers allowing them to:
125

126
- collect, classify and organize the documents found;
127
128
- gather all their scientific production in order to edit it in a personal profile page;
- follow the productions of colleagues;
129
- record and publish queries and their results for monitoring purposes;
130
- create bibliographies that can be exported to Zotero.
131

132
### The APIs of isidore.science
133

Edward Gray's avatar
Edward Gray committed
134
The [isidore.science search engine APIs](https://api.isidore.science) are available through the GET method on HTTP or HTTPS.
135
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).
136

137
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.
138

Edward Gray's avatar
Edward Gray committed
139
The [isidore.science API web page](https://api.isidore.science) details all the commands available for the different services available.
140

141
### Enriched metadata for *Linked Open Data*.
142

143
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:
144

145
146
- 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
147

148
In the ISIDORE *TripleStore*, the main vocabularies for structuring information are:
149

150
- RDF and RDFS
151
- Dublin Core Element Set
152
- Dublin Core TERMS
153
154
155
156
157
158
159
- SIOC
- FOAF
- OWL
- SKOS
- ORE
- DBPEDIA

160
(The complete list is available at <https://isidore.science/sparql?nsdecl>)
161
162


163
### Complementarity between ISIDORE and Zotero
164

165
#### Use from ISIDORE of the Zotero connector to feed its bibliographic database
166

167
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:
168

169
- On the page listing the results of a search,
170
- On the page listing the results of a search, in the page displaying a document.
171

172
#### Using the ISIDORE search connector from Zotero
173

174
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.
175

176
177
178
179
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?"
180
181


182
These [connectors and installation documentation are available on the TGIR Huma-Num GitLab](https://gitlab.huma-num.fr/spouyllau/ISIDORtero).
183

184
### Use of RSS feeds
185

186
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.
187

188
To do so, access your personal space (login required), and click "My queries" to see your registered queries:
189

190
![My Image](media/isidore.png)
191

192
For a registered query, you have to click on the pictogram "Request
Edward Gray's avatar
Edward Gray committed
193
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"}.
194

195
The copied link is in the form: `https://isidore.science/feed/lt3913`.
196

197
If your browser is equipped with a module for reading RSS feeds, this link can be used directly in your browser.
198
For our example, we will continue with Zotero.
199

200
201
In Zotero, you have to choose: New feed > From URI:

202

203
![My Image](media/zot-001.png){: style="width:60%;margin-left:20%"}
204

205
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
206
207
the url). Then paste it in "URL" of the Zotero RSS feed creation window, example below:

208

209
![My Image](media/zot-002.png)
210

211
Then you have to give a title to your feed, for example:
212
"isidore.science - Query on ...".
213

214
## What can be found in ISIDORE?
215

216
### Organization of documents and data in ISIDORE
217

218
ISIDORE contains several million documents in SSH that are harvested, enriched with scientific references and indexed. They are organized into:
219

220
221
222
- 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
223
224


225
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
226
municipal libraries.
227

228
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.
229

Nicolas Larrousse's avatar
Nicolas Larrousse committed
230
The main publication platforms (journals and books) present in ISIDORE are:
231
232
233

- OpenEdition
- Cairn
234
- Perseus
235
236
237
238
239
- Erudit
- Oapen
- Redalyc
- Scielo Books

Edward Gray's avatar
Edward Gray committed
240
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) :
241
242
243
244
245

```
SELECT * WHERE {
 ?s rdf:type <http://isidore.science/class/Collection>.
 ?s rdf:type <http://isidore.science/class/publications>.
246
247
 ?s dcterms:title ?title
} ORDER BY ASC(?title)
248
```
Nicolas Larrousse's avatar
Nicolas Larrousse committed
249
The main digital libraries (municipal, national, etc.) present in ISIDORE are:
250
251

- Gallica
252
- Selene
253
254
255
256
257
258
259
260
261
262
- E-rara
- NuBIS
- Octaviana
- Burgerbibliothek
- Berkeley Library Digital Collections
- Argonnaute
- BNE
- Cornell
- Didόmena

Edward Gray's avatar
Edward Gray committed
263
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) :
264
265
266
267
268

```
SELECT * WHERE {
 ?s rdf:type <http://isidore.science/class/Collection>.
 ?s rdf:type <http://isidore.science/class/primaires>.
269
270
 ?s dcterms:title ?title
} ORDER BY ASC(?title)
271
272
```

273
### Indexing of the main data platforms in SHS
274

275
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.
276

277
The main data platforms (sources, archives but also publications) are harvested by ISIDORE.
278

Edward Gray's avatar
Edward Gray committed
279
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)
280
281
282
283

```
SELECT * WHERE {
 ?s rdf:type <http://isidore.science/class/Collection>.
284
285
 ?s dcterms:title ?title
} ORDER BY ASC(?title)
286
287
```

288
Please feel free to report any new source to us.
289

290
#### Can data deposited and documented in NAKALA be referenced by ISIDORE?
291

292
Yes, data deposited and documented in [NAKALA (the data repository for SSH by Huma-Num)](https://documentation.huma-num.fr/nakala/) can be
293
294
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.

295

296
297
However, referencing by OAI-PMH harvesting is not
automatic for the moment, in particular to allow users to prepare and organize their data and
298
data and metadata. To be referenced, simply request by email to be indexed in ISIDORE via <isidore-sources@huma-num.fr>.
299

300
#### How will scientific articles and images deposited in the HAL, HAL-SHS and MédiHAL open archive be accessible in ISIDORE?
301

302
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.
303

304
#### Can the data deposited in the Didómena (EHESS) warehouse be referenced by ISIDORE?
305

306
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>.
307

308
#### Can data deposited in Calames (ABES) be referenced by ISIDORE?
309

310
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.
311

312
#### Can the data deposited in the Data.sciencespo warehouse be referenced by ISIDORE?
313

314
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.
315

316
#### Can the data deposited in the COCOON platform be referenced by ISIDORE?
317

318
319
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.

320

321
#### Can files and documents deposited in the European Zenodo platform be referenced by ISIDORE?
322

323
324
Yes, it is possible for ISIDORE to reference the files and
documents deposited and documented on the platform
325
[Zenodo](https://zenodo.org).
326

327
328
329
330
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
331
ISIDORE collection, allowing the depositors of several corpora
332
333
deposited in Zenodo to group them in ISIDORE to give them more
visibility.
334

335

336
To add your Zenodo repositories in ISIDORE, [please send us the URL OAI-PMH](mailto:isidore-sources@huma-num.fr?subject=%22Je%20souhaiterai%20faire%20moissonner%20mes%20dépôts%20Zenodo%22)
337
of your repository (see <https://developers.zenodo.org/#oai-pmh>).
338

339
340
341
#### Can files and documents deposited in *Gallica Marque Blanche* platform be referenced by ISIDORE?

Yes, the data deposited and documented in [*Gallica Marque Blanche*](https://www.bnf.fr/fr/gallica-marque-blanche) offer interoperability in OAI-PMH with a dedicated "Set".
342

343
344
345
346
347
#### Can Omeka farm powered by INIST-CNRS be referenced by ISIDORE?

Yes, it is possible for ISIDORE to reference the files and
documents deposited and documented on the [Omeka farm powrered by INIST-CNRS](https://www.inist.fr/realisations/omeka-pour-des-bases-de-donnees-valorisees/).

348
## How do I get data referenced by ISIDORE?
349

350
351
There are several ways to get data and documents referenced by
ISIDORE:
352

353
- 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).
354
355
This method is adapted to research program websites presenting document or data corpora, scientific blogs (except Hypotheses.org), and web pages in general.

356

357
These two methods are also often implemented by data publication tools (CMS, etc.), for example:
358

359
### Can a web site using Drupal be indexed by ISIDORE?
360

361
Yes, it is possible to have web pages generated by the Drupal CMS indexed by ISIDORE.
362
There are two ways to do this, depending on the nature of the
363
content of your pages:
364

365
- Either via the OAI-PMH protocol and in this case there are several
Edward Gray's avatar
Edward Gray committed
366
    modules for Drupal, see [OAI-PMH for Drupal](https://www.drupal.org/search/site/OAI-PMH?f%5B0%5D=ss_meta_type%3Amodule).
367
368
369
370
- 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.
371

372
### Can a website using Omeka Classic and Omeka-S be referenced by ISIDORE?
373

374
Yes, Omeka *Classic* and Omeka S offer modules to expose metadata according to the OAI-PMH protocol:
375

376
377
- Module for [Omeka S](https://omeka.org/s/modules/OaiPmhRepository/)
- Module for [Omeka Classic](https://omeka.org/classic/docs/Plugins/OaiPmhRepository/)
378
379


380
### How to report data in ISIDORE with metadata and OAI-PMH protocol?
381

382
To report your data in ISIDORE using the
383
OAI-PMH protocol, you just have to:
384

385
386
387
- 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
388
    make them accessible via [the OAI-PMH protocol](https://en.wikipedia.org/wiki/Open_Archives_Initiative_Protocol_for_Metadata_Harvesting);
389
- To organize and document the *Sets* in its OAI-PMH repository.
390
- To write to <isidore-sources@huma-num.fr> and give the address of the repository to Huma-Num.
391

392
#### Document sets in OAI-PMH: *Sets*
393

394
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.
395

396
It also makes it possible to define a hierarchy in the *Sets* with an inheritance mechanism by specifying
397
in the set name the name of the parent *Set* and the child *Set*,
398
399
separated by the `:` character. ISIDORE is able to use these
*Sets* to limit harvesting to a set of records or to differentiate between different
400
data sources within the same warehouse.
401
402
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.
403
404
405
406
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.

407

408
The *Sets* can present metadata, in the Dublin Core Element Set, which are specific to them. For example:
409

Edward Gray's avatar
Edward Gray committed
410
```xml
411
412
<set>
 <setSpec>OuvColl</setSpec>
413
 <setName>OuvColl</setName>
414
415
 <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">
416
   <dc:description>Research works distributed on Cairn.info</dc:description>
417
418
419
420
421
  </oai_dc:dc>
 </setDescription>
</set>
```

422
#### Records in OAI-PMH or *Records*:
423

424
425
426
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
427
Dublin Core Element Set (also known as Dublin Core "simple").
428
In addition, the harvester also collects the full-text document(s) whose URLs
429
430
431
432
433
434
 (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:
435
436


Edward Gray's avatar
Edward Gray committed
437
```xml
438
439
440
441
442
<dcterms:description>
<dcterms:creator>
<dcterms:date>
```

443
are essential.
444

445
##### Example of a complete record according to the OAI-PMH protocol:
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461

```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>
462
  <oai_dc:dc xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
463
464
465
466
467
  <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>
468
  <dc:subject>[SHS:ECO] Humanities and Social Sciences/Economy and finances</dc:subject>
469
470
  <dc:subject>[SDV:BIO] Life Sciences/Biotechnology</dc:subject>
  <dc:subject>[INFO:INFO_BT] Computer Science/Biotechnology</dc:subject>
471
  <dc:subject>[SDV:SA:AEP] Life Sciences/Agricultural sciences/Agriculture, economy and politics</dc:subject>
472
473
474
475
476
477
478
479
480
481
  <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>
482
  <dc:coverage>Montpelier</dc:coverage>
483
484
485
486
487
488
489
490
491
492
493
  <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>
```

494
495
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
496
the administrator of the OAI-PMH warehouse.
497

498

499
500
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.
501
502
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:

503

504
505
506
- 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).
507

508
We advise producers to regularly validate the compliance of their repository using, for example, the [tools of the Open archive initiative](https://www.openarchives.org/pmh/tools/). Finally, we advise data producers to contact the Huma-Num team for any information requests.
509

510
### How to report data in ISIDORE with RDFa metadata?
511

512
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
513
attributes", i.e. within the HTML code).
514

515
How to express metadata of a web page very simply by
516
517
518
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/),
519
the obsolescence of the latter can make it difficult to maintain them
520
521
over time. Another solution is to implement RDFa in the
HTML code of the WordPress theme you have chosen. For this to be easy
522
523
524
and manageable over time, the simplest way is to use the HTML header
in order to place `<meta>` tags that will contain some metadata.

525
526

Expressing metadata according to the RDF model via the RDFa syntax allows
527
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
528
529
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
530
531
for example with the Dublin Core Element Set.

532
533
534
535
536

#### 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
537
DOCTYPE will be:
538

Edward Gray's avatar
Edward Gray committed
539
```xml
540
541
542
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML+RDFa 1.0//EN" "http://www.w3.org/MarkUp/DTD/xhtml-rdfa-1.dtd">
```

543
The `<html>` tag must contain the addresses of the
544
ontology (via their *NameSpace XML*) which are used
545
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):
546

547

Edward Gray's avatar
Edward Gray committed
548
```xml
549
550
551
552
553
554
555
<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/">
```

556
To encode more information, It is possible to use more
557
document ontologies:
558

559

Edward Gray's avatar
Edward Gray committed
560
```xml
561
562
563
564
565
566
567
568
569
570
571
<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#">
```

572
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.
573

574
The RDFa structure through tags
575
576
577
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:
578

Edward Gray's avatar
Edward Gray committed
579
```xml
580
581
582
<link rel="dc:identifier" href="http://monblog.com/monbillet.html" />
```

583
584
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
585
page on the web.
586
587


588
The `<meta>` tags then define a set of metadata, which in our case is descriptive information about the blog post's web page:
589

Edward Gray's avatar
Edward Gray committed
590
```xml
591
592
593
<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" />
594
<meta property="dcterms:created" content="2011-01-27" />
595
<meta property="dcterms:abstract" content="A descriptive summary of my page's content" xml:lang="en" />
596
<meta property="dcterms:abstract" content="A summary in english" xml:lang="en" />
597
598
<meta property="dc:subject" content="keyword 3" />
<meta property="dc:type" content="ticket" />
599
<meta property="dc:format" content="text/html" />
600
<meta property="dc:relation" content="A link to a complementary web page" />
601
602
```

603
Depending on the nature of the content of the web page, it is of course possible
604
to be more precise, more refined and more complete in the
605
606
encoded information. For example, it would be wise to use the DC Terms vocabulary.

607

608
The DC Terms allow, for example, a precise form for a bibliographic reference of the content to be included:
609
610


Edward Gray's avatar
Edward Gray committed
611
```xml
612
<meta property="dcterms:bibliographicCitation" content="Put a bibliographic reference here" />
613
614
```

615
It would be possible to describe the entire text of a web page using the SIOC vocabulary [using the
616
property](http://www.lespetitescases.net/rdfaiser-votre-blog-2-la-pratique).
617

618
619
620
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`.
621
622

```xml
623
<meta property="dcterms:isPartOf" content="URL of another web page" />
624
625
```

626
#### Creating the Sitemap
627

628
629
630
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:

631

Edward Gray's avatar
Edward Gray committed
632
```xml
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
<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>
```

649
650
651
652
653
654
655
656
657
658
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
659
explanations. It may be that your articles are published in
660
journals that are not electronic or that do not make their articles available even
661
 long after they have been published. Since its
662
creation, [ISIDORE](https://isidore.science) favors open
663
access since indexing is better for articles available in
664
open access. Many electronic journals have made this choice through
665
portals such as Open Edition Journal (formerly Revues.org)
666
Érudit, Persée, and Cairn.info, Redalyc, OApen and
667
and articles from these journals are therefore collected and indexed by
668
669
[ISIDORE](https://isidore.science).

670
671
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
672
that does not allow indexing via the standard protocol
673
(see the question and answer on OAI-PMH).
674

675
Other journals make their articles available, but only after an
676
embargo period. In this case,
677
678
[ISIDORE](https://isidore.science) indexes only the metadata
of the article. If you connect via your university library
679
, documentation center or via BibCNRS,
680
you may still have access to these articles.
681

682
683
The collections indexed by
[ISIDORE](https://isidore.science) can be searched by using the engine itself and by
684
indicating that you want to search the collections.
685

686
It is also possible that your article is published as a PDF image,
687
in which case only the indexing by
Edward Gray's avatar
Edward Gray committed
688
[ISIDORE](https://isidore.science) will be allowed, but not its
689
full text indexing.
690

691
Lastly, it is possible that some of your articles are published in
Nicolas Larrousse's avatar
Nicolas Larrousse committed
692
journals that are not classified in SSH.
693

694
In all these cases, you can deposit your articles in an
695
open archive such as HAL (HAL-SHS in particular) which is also indexed by
Edward Gray's avatar
Edward Gray committed
696
[ISIDORE](https://isidore.science) or contact your
697
bu/documentation center.
698

699
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.
700

701
### Why are some books/chapters of books not reported in ISIDORE?
702

703
704
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
705
706
ISIDORE.

707
708
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:
709

710
711

- [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
712
- [Scielo Books](https://isidore.science/search/?collection=10670/3.7oraz1) (Brazil);
713
- [OApen](https://isidore.science/search/?collection=10670/3.pwofj8) (Netherlands);
714
- [Erudit](https://isidore.science/s/collection?q=erudit) (Canada);
715
- ...
716

717
In addition, you can, in agreement with your publisher, deposit your work or
718
book or book chapters in the open archive
Edward Gray's avatar
Edward Gray committed
719
[HAL-SHS](https://halshs.archives-ouvertes.fr). It will then be indexed by
720
721
ISIDORE within the framework of the indexing of HAL-SHS and recognized as a book chapter.

722

723
### Why are some databases are not reported in ISIDORE?
724

725
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).
726

727
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.
728

729
## ISIDORE training courses
730

731
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:
732

733
- The *Urfist Méditerranée* proposes a new e-learning training on Isidore (only in french)](https://urfist.univ-cotedazur.fr/nouvelle-formation-en-ligne-une-initiation-a-isidore/) (March 2021)
734
- ["Isidore, my personal research assistant"](https://ig.hypotheses.org/2215) by Johanna Daniel (April 2020)