OpenLexicon

Lexique-Dicollecte is a French lexicon containing a list of inflected forms (version 6.4.1). It is generated from the Dicollecte database and includes both old and new spellings resulting from the 1990 French orthographic reforms proposed by the Conseil supérieur à la langue française and the Académie française.

Corpus Statistics

The lexicon is built upon a massive corpus of recognized words:

• Google 1-grams: ~145.6 billion words
• Wikipédia: ~1.09 billion words
• Wikisource: ~569.5 million words
• Littérature: ~143.1 million words
• Total: ~147.4 billion recognized words

Licensing and Attribution

• Author: Olivier R. (Grammalecte)
• Website: https://grammalecte.net/
• License: Mozilla Public License version 2.0 (MPL 2.0)

CHACQFAM is a French lexical database containing estimated acquisition age and familiarity scores for 1,225 monosyllabic and bisyllabic words. The dataset was developed by Christian Lachaud at the University of Geneva in the 2000s.

Dataset Details:

• Language: French
• Content: 1,225 monosyllabic and bisyllabic words.
• Metrics: Estimated age of acquisition and word familiarity.
• Source Table: CHACQFAM.xls
• Website: HAL Open Archive

Citation: Christian Lachaud. (2007) CHACQFAM : une base de données renseignant l’âge d’acquisition estimé et la familiarité pour 1225 mots monosyllabiques et bisyllabiques du Français. (2007) L'Année Psychologique, 107 (1), pp.39-63.

The Voisins 1.02 dataset provides descriptors for orthographic neighbors (neighbors created by changing a single letter without changing the position of other letters) for the 130,000 entries in the Graphemes database of Lexique262.

Data is provided in the voisins.txt file, which includes the following fields:

• Graph: The orthographic entry from Lexique2.
• NbVoisOrth: The number of orthographic neighbors.
• VoisOrth: The list of different orthographic neighbors.
• FreqVoisOrth: The frequencies (from Frantext and per million) for each neighbor.
• FreqCum: The cumulative frequency of all neighbors.

The French Lexicon Project (FLP) provides lexical decision data for 38,840 French words and 38,840 nonwords (pseudowords). The dataset contains approximately 1,942,000 reaction times collected from 975 participants.

Available Data Tables:

• FLP.words.csv: Lexical decision data for real French words.
• FLP.pseudowords.csv: Lexical decision data for pseudowords.

Reference Information:

• Website: https://sites.google.com/site/frenchlexicon/
• Publication: Ferrand et al. (2010). The French Lexicon Project: Lexical Decision Data for 38,840 French Words and 38,840 Pseudowords. Behavior Research Methods 42 (2): 488--96. DOI: 10.3758/BRM.42.2.488
• License: CC BY-SA 4.0

The British Lexicon Project (Keuleers et al, 2012) is a dataset containing lexical decision data for over 28,000 monosyllabic and disyllabic English words. It is useful for research involving lexical decision tasks and English word properties.

Key Resources:

• Web site: http://crr.ugent.be/programs-data/lexicon-projects
• Data Tables: http://crr.ugent.be/blp/txt/
• Publication: Keuleers, E., Lacey, P., Rastle, K., & Brysbaert, M. (2012). "The British Lexicon Project: Lexical Decision Data for 28,730 Monosyllabic and Disyllabic English Words." Behavior Research Methods 44 (1): 287–304. https://doi.org/10.3758/s13428-011-0118-4

The English Lexicon Project (ELP) provides a standardized behavioral and descriptive dataset containing 40,481 words and 40,481 nonwords. The dataset is derived from two primary types of experimental tasks:

1. Lexical Decision Task: Data collected from 816 participants (approx. 3,400 responses per participant).
2. Speeded Naming Task: Data collected from 444 participants (approx. 2,500 responses per participant).

For detailed methodology and research context, refer to the official website or the primary publication.

SUBTLEX-CH (Cai & Brysbaert 2010) is a lexical database containing Chinese word and character frequencies. The frequencies are derived from a large corpus of film and television subtitles, totaling approximately 46.8 million characters and 33.5 million words.

Use this dataset when you need frequency measures for Chinese linguistic research, specifically focusing on word and character distributions in naturalistic subtitle data.

SUBTLEX-US is a lexical database containing word frequencies derived from English-US movie and TV series subtitles, totaling approximately 51 million words. It provides two primary frequency measures for each word:

1. SUBTLEXWF (Subtitle frequency: word form frequency): The frequency of the word per million words.
2. SUBTLEXCD (Subtitle frequency: contextual diversity): The percentage of films in which a word occurs. This measure represents how widely distributed a word is across different contexts.

The dataset is available in a TSV format containing 74,286 words.

SILEX is a unified French lexical database containing 119,609 word entries. It is designed for researchers to select words based on specific statistics and characteristics, enabling the construction of frequency and probability tables for phonological and orthographic units. It is particularly useful for studying spelling and silent-letter endings.

SILEX is derived from two primary sources:

• Manulex-infra: Based on 54 children's schoolbooks (divided into grade-specific corpora).
• Lexique 3.80: Based on 218 adult novels.

When using SILEX, you must credit its parent databases (Manulex-infra and/or Lexique 3.80) in your research.

Lexique382 is a French lexical database containing approximately 140,000 words. For each entry, it provides:

• Orthographic and phonemic representations
• Associated lemmas
• Syllabification
• Grammatical category (part of speech)
• Gender and number
• Frequency data (from both a book corpus and a film subtitle corpus)

You can download the dataset directly via the following link: Lexique382.zip

The Gougenheim 1.00 dataset provides frequency and distribution data for 1,064 words. The data is derived from an oral language corpus consisting of interviews with 275 people (163 texts, 312,135 words, and 7,995 unique lemmas).

Note on limitations: This dataset does not provide frequency data for words with a frequency lower than 20.

Morpholex-en is a lexical database containing approximately 70,000 English words. It is designed to provide morphological variables for these words, supporting research and applications involving derivational morphology.

For more detailed information or to access the original source, visit the project website: https://github.com/hugomailhot/MorphoLex-en