Skip to main content
Join the official 2019 Python Developers SurveyStart the survey!

Labeled contexts of Russian polysemous words

Project description

Contexts sampled from RuTenTen and RNC. Sense definitions from Active Dictionary. Some words have two annotators. Number of contexts is 100 for most words and 500 for 7 words.

Annotators (words):

  • Анастасия Лопухина (47)
  • Константин Лопухин (11)
  • Александра Удальцова (2)
  • Анастасия К. (2)
  • Анна Кот (2)
  • Анна Татаренко (2)
  • Борис Иомдин (2)
  • Иван Самойленко (1)

Contexts are stored in rl_wsd_labeled/:

├── adjectives
│   └── RuTenTen
├── nouns
│   ├── RNC
│   └── RuTenTen
└── verbs
    └── RuTenTen

A python interface is provided. Intall the package first:

pip install rl_wsd_labeled

and then in order to get labeled contexts:

>>> import rl_wsd_labeled
>>> f = rl_wsd_labeled.contexts_filename('nouns', 'RuTenTen', 'горшок')
>>> rl_wsd_labeled.get_contexts(f)

({'1': 'Округлый глиняный сосуд для приготовления пищи (печной горшок)',
  '2': 'Расширяющийся кверху сосуд с отверстием в дне (цветочный горшок)',
  '3': 'Ночной горшок'},
 [(('телевизор, - ковер, , - музыкальный центр, - стол, - аквариум, - 3 шкафа, - цветы в',
    ' горшках',
    ', - мелкие аксессуары.'),
  (('ибо настанет срок и оно будет разрушено течением времени либо войною, будто старый',
    ' горшок',
    ' с вином в трюме торгового корабля, попавшего в бурю и разбившегося о скалы.'),

Apart from senses, there are two special annotations: “0” means “I don’t know/the context is unclear/the contexts is invalid”, and “max sense + 1” mean “other sense, not listed among given senses”. Contexts marked as “0” or “other” are not returned, unless with_skipped=True is passed. If there was more then one annotator, contexts where annotators did not agree are also not included. There is a function rl_wsd_labeled.get_agreement that returns the ratio of senses where where both annotators gave either the same concrete sense, or both skipped the senses (so “0” and “other” are considered equal).

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for rl_wsd_labeled, version 0.1.1
Filename, size File type Python version Upload date Hashes
Filename, size rl_wsd_labeled-0.1.1.tar.gz (1.3 MB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page