Submissions/OpenCorpora.org: wiki-principles + Wikimedians + Wikimedia projects
|
After careful consideration, the programme committee has decided not to accept the below submission at this time. Thank you to the author(s) for participating in the Wikimania 2013 programme submission, we hope to still see you at Wikimania this August. |
- Submission no.
- 2022
- Subject no.
- B10
- Title of the submission
OpenCorpora.org: wiki-principles + Wikimedians + Wikimedia projects
- Type of submission
presentation
- Author of the submission
- Country of origin
Russian Federation
- Affiliation
OpenCorpora.org
- E-mail address
lvova
wikimedia.ru
- Username
Lvova
- Personal homepage or blog
- Abstract
OpenCorpora is a community-driven project that aims at creating a corpus of Russian (for beginning) texts with linguistic annotation, which is fully accessible to researchers under the terms of the Creative Commons Attribution/Share-Alike License. OpenCorpora includes texts taken from Wikimedia projects and uses wiki approach to the annotation process.
- Detailed proposal
Annotated corpus is a database that includes natural language text with its linguistic annotation: part-of-speech tags, syntax, semantics and so on. Open access annotated corpora are important tools for the researchers working in the field of natural language processing since they can be used in a wide range of computational experiments. There is a number of such open access corpora for English, Spanish and other languages but there have been no open access annotated corpora available for the Russian language. This is due to both high cost of annotation and poor understanding of open licenses in Russia. OpenCorpora improves this: all our data can be downloaded and used under the terms of the CC-BY-SA and annotation works are performed by volunteers. It’s important, that Wikimedians are really useful source of texts (from Wikipedia. Wiktinary, Wikibooks, Wikisource, from own blogs) and working with the annotations of texts.
OpenCorpora uses a lot core principles of wiki-style content development and applies them to corpus annotation:
- participants can edit annotation at any moment
- all changes are recorded in the history and can be cancelled
- low entry threshold for volunteers
This wiki-style is combined with pipeline of both human and automatic data verification in order to maintain high data quality suitable for linguistic software assessment purposes.
Within the scope of OpenCorpora project we are developing infrastructure that makes it possible to involve any Russian language native speaker into corpus annotation works. This works because a large number of annotation tasks really doesn't require deep understanding of linguistic theory and only native speakers' intuition is required.
In this lecture we will explain the way corpus annotation process works, methods we use to cope with mistakes and the results we are delivering.
- Track
- Cultural and Educational Outreach
- Length of presentation/talk
- 25 Minutes
- Language of presentation/talk
- English
- Will you attend Wikimania if your submission is not accepted?
- probably yes
- Slides or further information (optional)
Some publications in Russian are there: http://www.opencorpora.org/?page=publications
- Special requests
Interested attendees
If you are interested in attending this session, please sign with your username below. This will help reviewers to decide which sessions are of high interest. Sign with four tildes. (~~~~).
- Yes, i'm interested. --Rave (talk) 12:05, 19 March 2013 (UTC)
- It would be also great to learn if the project will get new languages. Wizardist (talk) 17:21, 27 March 2013 (UTC)
- Yep, we planned to reach this goal too :) Lvova (talk) 17:27, 27 March 2013 (UTC)
- Slashme (talk) 18:08, 7 April 2013 (UTC)
- អមីរ ឯ. អហរោណិ 06:52, 12 April 2013 (UTC)
- Jtmorgan (talk) 23:38, 17 April 2013 (UTC)
- CT Cooper · talk 00:06, 29 April 2013 (UTC)
- Amakuha (talk) 19:14, 13 May 2013 (UTC)
- Ijon (talk) 05:36, 16 May 2013 (UTC)
- Your name here