Jiwar Database

Jiwar database is an open-source neighborhood database covering 40 languages. It provides three types of neighborhood information (orthographic, phonological, phonographic) for 5k-30k words per language.

Features

• 🗣️ Supports 40 languages
• 📁 Available in:
  • TSV
  • CSV
  • XLSX
  • TXT

How to Use

1. Navigate: go to the jiwar_database directory.

2. Obtain: download the file for the language and format you need.

3. Query: we recommend using the database in Excel format as it allows for fast and robust queries.

File Structure

Each file contains 25 columns with the following information:

Column	Name	Description
1	word	The orthographic word form
2	raw_frequency	Absolute frequency in the corpus for the target word
3	freq_per_m	Frequency per million words for the target word
4	zipf	Zipf frequency scale for the target word
5	num_letters	Number of letters in the orthographic word
6	num_phonemes	Number of phonemes in the IPA transcribed word
7	IPA	The International Phonetic Alphabet transcription of the target word
8	orth_density	Number of orthographic neighbors (words differing by one letter via substitution, addition, or deletion)
9	orth_density_nbrs	List of forms of orthographic neighbors identified in 'orth_density'
10	OLD20	Orthographic Levenshtein Distance 20
11	orth_nbr_zipf_m	Mean Zipf frequency of orthographic neighbors
12	orth_nbr_zipf_SD	Standard deviation of Zipf frequency of orthographic neighbors
13	orth_C	Orthographic clustering coefficient
14	phon_density	Number of phonological neighbors (words differing by one phoneme via substitution, addition, or deletion)
15	phon_density_nbrs	List of forms of phonological neighbors identified in 'phon_density'
16	PLD20	Phonological Levenshtein Distance 20
17	phon_nbr_zipf_m	Mean Zipf frequency of phonological neighbors
18	phon_nbr_zipf_SD	Standard deviation of Zipf frequency of phonological neighbors
19	phon_C	Phonological clustering coefficient
20	pg_density	Number of phonographic neighbors (words differing by one letter and one phoneme via substitution, addition, or deletion)
21	pg_density_nbrs	List of forms of phonographic neighbors identified in 'pg_density'
22	PGLD20	Phonographic Levenshtein Distance 20
23	pg_nbr_zipf_m	Mean Zipf frequency of phonographic neighbors
24	pg_nbr_zipf_SD	Standard deviation of Zipf frequency of phonographic neighbors
25	pg_C	Phonographic clustering coefficient

Supported Languages

N	Code	Language	Supported Writing Script
1	af	Afrikaans	Latin
2	ar	Arabic	Fully diacritized Arabic
3	bg	Bulgarian	Cyrillic
4	bs	Bosnian	Cyrillic, Latin
5	ca	Catalan	Latin
6	cs	Czech	Latin
7	de	German	Latin
8	el	Greek	Greek
9	en-gb	English (GB)	Latin
10	en-us	English (US)	Latin
11	eo	Esperanto	Latin
12	es	Spanish	Latin
13	et	Estonian	Latin
14	eu	Basque	Latin
15	fa	Persian	Perso-Arabic
16	fi	Finnish	Latin
17	fr	French	Latin
18	hr	Croatian	Latin
19	hu	Hungarian	Latin
20	hy	Armenian	Armenian
21	id	Indonesian	Latin
22	it	Italian	Latin
23	kk	Kazakh	Cyrillic
24	ko	Korean	Hangul
25	lt	Lithuanian	Latin
26	lv	Latvian	Latin
27	mk	Macedonian	Cyrillic
28	ms	Malay	Latin
29	nl	Dutch	Latin
30	no	Norwegian	Latin
31	pl	Polish	Latin
32	pt	Portuguese	Latin
33	ro	Romanian	Latin
34	ru	Russian	Cyrillic
35	sk	Slovak	Latin
36	sq	Albanian	Latin
37	sr	Serbian	Cyrillic
38	sv	Swedish	Latin
39	tr	Turkish	Latin
40	uk	Ukrainian	Cyrillic
41	ur	Urdu	Perso-Arabic

File Naming

Files in the database are named using ISO language codes (e.g., 'ar' for Arabic, 'fr' for French).

Language Not Listed?

If your language isn't listed, you can use the Jiwar Calculator to generate neighborhood measures for words in 90 additional language varieties.

Citation

If you use the Jiwar database in your research, please cite:

@preprint{Alzahrani:2024:jiwar,
    title = "{Jiwar: A database and calculator for word neighborhood measures in 40 Languages}",
    author = {Alaa Alzahrani},
    year = "2024",
    note = "Preprint"
}