Onterm dictionary: reinforcing specialized terminology knowledge
Bareun supplements the specialized terms that Urimalsaem (Open Korean Dictionary, NIKL's collaborative dictionary) does not cover with the Onterm dictionary,
onterm-dict.fb. It recognizes words that rarely make it into general Korean dictionaries, such as academic, technical, and industrial terms,
so that morphological analysis and spell checking can keep up with specialized terminology.
What is 'Onterm'
Onterm (온용어) is a specialized-terminology knowledge platform run by the National Institute of Korean Language (NIKL).
So that anyone can easily find and use terminology knowledge, it gathers term data that was scattered across institutions and fields,
classifies it, and integrates it by linking terms with the same form and concept. It is also adding the essential information needed to
understand and use terms (definitions, examples, original-language forms, and so on).
Bareun uses this terminology knowledge in morphological analysis and correction.
This dictionary is included in builds that include the spell check feature.
Filling in the specialized terms Urimalsaem does not cover
Urimalsaem covers a broad Korean vocabulary, but it cannot include every specialized term and neologism that appear quickly. The Onterm dictionary fills exactly this gap.
| Type of vocabulary | Examples |
|---|---|
| Information technology terms | 클라우드 (cloud), 머신러닝 (machine learning), 오픈소스 (open source) |
| Academic and industrial terms | Technical terms for each field |
| Neologisms and compounds | Newly established expressions |
How the two dictionaries share the work
If Urimalsaem is the standard for general Korean vocabulary, the Onterm dictionary is a supplement for specialized and recent vocabulary. The two dictionaries work together, so everything from everyday words to specialized terms is recognized broadly.
Stored as a FlatBuffer and looked up quickly by headword
Like Urimalsaem, the Onterm dictionary is stored in FlatBuffer format (.fb)
and looked up through a dictionary lookup structure keyed by headword. The structure is designed for fast lookup in a read-only dictionary.
graph LR
W[Input word] --> U{In Urimalsaem?};
U -- Yes --> R1[IN_URIMALSAEM];
U -- No --> O{In Onterm?};
O -- Yes --> R2[IN_ON_DICT];
O -- No --> R3[OUT_OF_VOCAB guess];
Source marking: IN_ON_DICT
Morphemes recognized by the Onterm dictionary get IN_ON_DICT as their out_of_vocab value.
When you review results, this is evidence that "this specialized term was confirmed in the Onterm dictionary."
Because it is distinct from words that came from Urimalsaem (IN_URIMALSAEM), you can clearly trace where a word came from.
Frequently asked questions
Q. How is the Onterm dictionary different from the Urimalsaem dictionary?
Urimalsaem is the general Korean vocabulary of the National Institute of Korean Language, and the Onterm dictionary supplements the
specialized terms and neologisms that dictionary does not cover. The two dictionaries are distinguished by their source values (IN_URIMALSAEM, IN_ON_DICT).
Q. Which builds include the Onterm dictionary?
onterm-dict.fb is shipped with builds that include the spell checker.
Q. What should I do if a specialized term comes out as an out-of-vocabulary word?
If a domain term you use often is analyzed as an out-of-vocabulary word (OOV, OUT_OF_VOCAB),
register it in the custom dictionary so it is recognized as a proper noun or compound noun.
You manage the custom dictionary through the CustomDictionaryService API.
Q. Does Onterm mean "online terms"?
No. Onterm is the proper name of a specialized-terminology knowledge platform. It is specialized terminology knowledge in which term data scattered across institutions and fields is gathered, classified, and integrated, with information such as definitions, examples, and original-language forms added. Bareun uses it in morphological analysis and the spell checker.