ULAPI: Linguistic Intelligence for Developers
Add linguistic smarts to your applications with the Ultralingua Application Programming Interface (ULAPI). Multi-lingual dictionary lookup, verb conjugation, root word discovery, and more. Add ULAPI’s features to your software via an HTTP-based API or compile it right into your own applications (in C++, Swift, Objective-C, Python,…).
Questions? Contact us at email@example.com.
ULAPI via HTTP
Try out a few of the most popular ULAPI features here. Contact us at firstname.lastname@example.org to register for an account and to get access to the full documentation.
Word lookup / translation
Lemmatization (finding root words)
ULAPI via HTTP
Single Dataset Plan: $1 per 1000 queries, starting at $10 per month
This plan gives you access to one bilingual dictionary of your choice. For example, if you choose our Spanish-English dictionary, you get Spanish-to-English dictionary lookup, English-to-Spanish dictionary lookup, and all our Spanish and English monolingual features (e.g. both Spanish verb conjugation and English verb conjugation).
All Datasets Plan: $1 per 250 queries, starting at $25 per month
This plan gives you access to all our bilingual and monolingual dictionaries.
Get an API key for a one-month trial. Contact us at email@example.com.
ULAPI for iOS, ANDROID, C++, Python, etc.
Want to build ULAPI’s features into your mobile, desktop, or server application? Contact us at firstname.lastname@example.org so we can learn more about your needs.
ULAPI Features and Case Studies
ULAPI is written in C++, and has been incorporated into applications on iOS, macOS, Windows, Android, and Linux, using Swift, Objective C, Java, C++, and Python. It is also available as an HTTP-based API.
Monolingual ULAPI features (e.g. verb conjugation and lemmatization/root-word finding) are available in English, French, Spanish, German, Italian, Portuguese, Dutch, Russian, and Latin.
Our bilingual dictionary datasets include the following. All datasets support lookup in both directions (e.g. both Spanish-to-English and English-to-Spanish for our Spanish-English dataset).
We also have a monolingual dictionary and a thesaurus for both English and French.
+ Translation dictionary look-up
Specify a word in a source language, and obtain translations and definitions in a target language. Each translation comes with detailed part-of-speech information. This works for multi-word phrases as well (e.g. "take over" or "put down") if the phrase has a corresponding entry in the dictionary.
Example: Look up dog in English-to-French, and get chien [noun, masculine], chienne [noun, feminine], and pitou [noun, masculine, dialect=Quebec, tone=Colloquial].
+ Lemmatization / stemming / root-word discovery
Specify a word and obtain a complete list of its root forms.
Examples: Lemmatize rakes in English to get information about both rake [noun] and rake [verb]. ULAPI stemming works for compound forms as well as single-word forms, so for example, will have thrown will lemmatize to throw.
+ Verb conjugation
Specify a verb in a language, and obtain a complete list of its conjugated forms.
Example: Conjugate dormir ("to sleep") in Spanish and get a list of forms like (tense=present, person=first, number=singular: duermo), etc.
+ Pluralization and singularization
Specify a word and obtain its singular and/or plural form(s).
Example: Pluralize the English word forum to get both forums and fora. Singularize patches to get patch.
+ Number translation
Specify a number to get the word form of that number.
Example: 123 in German gives you hundertdreiundzwanzig.
+ Word structure analysis (also known as morphological analysis)
Specify a word and obtain an analysis of the word's structure based on each of its root forms.
Example: the word bakers is the plural form of baker [noun], which is in turn the "one who does something" form of the bake [verb].
+ Word inflection (also known as morphological synthesis)
Specify a word and obtain a list of all its derived forms, including detailed part of speech information.
Example: the complete inflection of the French adjective bleu includes bleu, bleus, bleue, and bleues, along with superlative and comparative forms like la plus bleue ("the most blue", feminine, singular).
+ Part-of-speech tagging
Provide a sentence or sentence fragment to obtain a list of the most likely parts of speech for each word, in context.
Example: the sentence William drinks two drinks at night gives you:
William [noun, singular, proper]
drinks [verb, third-person, singular, present tense]
drinks [noun, plural]
night [noun, singular]
We have tons of ideas of ways to put ULAPI to work. Text mining support, natural language database queries or game input, numbers written out in word form for a fancy countdown, help predict the next word in the user's sentence, etc., etc.
If you have language feature ideas for your apps and you want to know whether ULAPI can help you realize them, let's talk!. In the meantime, here are a few ideas to get you thinking.
Case Study 1: An image library with user-generated keywords
Imagine your application involves a database of images uploaded by your customers, who tag their own images with keywords so other people can search for them by concept. You might, for example, end up with many photographs of cattle in the snow, with keywords assigned like so:
Problems: people use different forms of the same words as keywords (cow and cows, snow and snowing), and words for the same ideas in different languages (cow and vache, snow and neige).
Solution: use ULAPI's stemming and dictionary look-up features to support keyword search that can find the right images even though the form and language of the keywords vary.
Case Study 2: In-app dictionary
In a lot of apps, especially ones that are for reading or writing, it's handy for people to have quick access to a dictionary. Though there are built-in monolingual dictionary tools on some operating systems, there are seldom bilingual or multilingual dictionaries available.
Problem: Your customers need quick access to a (possibly multilingual) dictionary.
Solution: Use ULAPI's dictionary lookup and rendering tools, and any of its more than 30 dictionary datasets.
Here's an example of the iOS popup dictionary that we have available for licensing:
Need to add your own specialized vocabulary for lookup? We can work with you to build special datasets for just such a purpose.
Case Study 3: Word games
When we were doing play tests of our word association game Psychobabble, we discovered that people would often type one form of a word when the puzzle contained another form of the same word. In this game, penalizing people for typing cats when the puzzle contained cat was just frustrating, and made the game less fun.
Problem: How can we treat different forms of a word as identical for the purposes of the game?
Solution: Use ULAPI's pluralizer, singularizer, and alternate spelling detector to detect matches between words that should be considered the same.
The result? See here how our player has typed colour in the first image, and the game has granted the player a match with color in the second image.