FR-MIGR-TWIT Corpus 2.0
Authors/Creators
Description
The MIGR-TWIT Corpus is a multilingual corpus of tweets about the topic of migration in Europe, developed within the framework of the OLiNDiNUM (Observatoire LINguistique du DIscours NUMérique, Linguistic Observatory of Online Debate), with the aim of documenting and analyzing online public discourse on (im)migration in contemporary European politics. Considering the global issue of migration over the last decade (2011–2022), and in order to observe discursive evolution accros the political spectrum and in two national contexts (France and the UK), the MIGR-TWIT Corpus was published:
- Tweets of right and far-right politics in Europe (Battaglia, Blandino, Jeon & Pietrandrea, 2022)
- UK-R-MIGR-RA-TWIT-2012-2022
- FR-R-MIGR-TWIT-2011-2022
- Tweets of French left-wing politics (Pietrandrea & Jeon, 2023)
- FR-L-MIGR-TWIT-2011-2022
The FR-MIGR-TWIT Corpus 1.0, compiled from the FR-R and FR-L modules, comprises 17,395 tweets posted by 39 French political figures and parties (16 right-wing and 23 left-wing) between 2011 and 2022. Tweets containing migr- derivatives were retrieved via the Twitter API v2 Academic Research, and truncated retweets (>140 characters) were restored through targeted verification (for detailed information on each module see the links above).
This second version provides multilayer linguistic annotations of all occurrences of forms derived from the Latin root migr-.
The FR-MIGR-TWIT Corpus 2.0 offers:
- multilayer linguistic annotations associated with each occurrence of a migr- derivative (MIGR-LEXICON), including semantic roles (ROLE_SEM), syntactic functions (FUNC_SYN), lemmatised forms (LEMMA), as well as features and collocational items related to modification (MODIFICATION, LEMMA_MODIF_*, LEMMA_NOUN-1) and list/parallelism constructions (LIST_PAR, LENGTH-1, #forme#_MIGR-LIST_PAR) (Non-exhaustive list);
- tweet URLs (tweet_url) and 44 types of data retrieved through the Full Archive Search endpoints of the Twitter API v2, such as the textual content of tweets (data__text), posting date (data__created_at), user ID (data__author_id), number of retweets (data__public_metrics__retweet_count) likes (data__public_metrics__like_count), replies (data__public_metrics__reply_count), quotes (data__public_metrics__quote_count). (Non-exhaustive list)
Notes
Technical info
The corpus is available in CSV, XML, and TEI-XML formats. The CSV and XML files provide stand-off linguistic annotations and metadata. The TEI-XML files encode the canonical textual layer.
Series information
The FR-MIGR-TWIT Corpus 1.0 (© 2025 Jeon, Battaglia & Pietrandrea) was previously released via Zenodo.
Files
FR-MIGR-TWIT_2.0.zip
Files
(11.7 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:b14595998e34134a7d3aa6cb4e0d567d
|
11.7 MB | Preview Download |
Additional details
Related works
- Is compiled by
- Dataset: 10.5281/zenodo.7347479 (DOI)
- Dataset: 10.5281/zenodo.7871602 (DOI)
- Is part of
- Thesis: https://theses.fr/2025ULILH003 (URL)
- Conference proceeding: 10.14618/1z5k-pb25 (DOI)
- Presentation: https://hal.science/hal-05240728v1 (URL)
Funding
- Institut Universitaire de France
- Université Lille Nord de France
- Projet d'Internalisation 2021
- Campus France
- Hubert Curien Partnerships PHC Galilée 2018-19
- Campus France
- Hubert Curien Partnerships PHC Van Gogh 2018-19