What You Should Look for in a Dictionary API

Whether you are a developer in need of a reliable dictionary API for your next word game app, or a researcher seeking comprehensive data for NLP or machine learning purposes, you need to choose the most appropriate API that suits your specific requirements. A quick Google search would uncover numerous decent-looking options, so how do you make the choice?

The following summary sets out key considerations to help you navigate this crowded terrain.

Linguistic Data Characteristics

Starting with the basics, you should begin by identifying your language needs. Are these confined to a single language, a particular language pair, or any kind of multilingual combination? Do you need to appeal to native speakers of a language or to non-native users or learners? Of course, the answer follows from your business, academic, or other needs. For example, developing a word game app targeting North American users is quite different from applying a machine learning algorithm for multilingual word embedding.  

While plenty of monolingual dictionary APIs are available online, finding a quality bilingual database is rather challenging. Things get even trickier if you are looking for less-common languages or language pairs whose resources may be quite scarce.

Moving forward with our quest, after determining the relevant language needs, the next feature of interest is the linguistic data types. Dictionary APIs differ greatly by the kind of linguistic data offered, according to the designated usage. For example, a bilingual dictionary API that is meant to serve translators will probably contain definitions rather than information regarding pronunciation. On the other hand, a dictionary API designated for language learning software is more likely to include pronunciation, as well as common inflections, examples of usage and other grammatical information.

Here is a short summary of general linguistic information that can be found in quality dictionaries:

  •  Part of speech is the classification of a word’s grammatical category, such as noun, verb, adjective, etc.   
  •  Grammatical information usually specifies the gender and number of the entry, e.g. noun masculine/feminine/neutral or masculine plural, or verb transitivity/intransitivity.
  •  Pronunciation data indicates the way a word is pronounced, specifying the sequences of sounds used in speaking a given word or phrase, usually provided in IPA (International Phonetic Alphabet), or other kinds of phonetic transcriptions – including Roman characters used in Pinyin for Chinese or Romaji for Japanese.
  • ·Geographical usage features information regarding the particular dialect or region. For example, chippy is an English word meaning a carpenter, but you might not be familiar with it as it is common mainly in Australia. This also includes data about differences between American and British English, such as the meaning of the same word (e.g. football) or different words with the same meaning (e.g. sidewalk and pavement).
  •  Inflections present the headword’s modifications for different (mainly irregular) grammatical variants such as tense, personnumbergender, etc. (e.g. childchildren, eatate).
  •  Definitions provide a description of the meaning of a word, often including more than a single sense.
  •  Examples illustrate the use of a word or a multiword expression and may consist of a full sentence or a short phrase.
  •  Synonyms present words or phrases that mean exactly or nearly the same as another word or phrase in a given language. For example, destiny is a synonym of fate.
  • ·Antonyms present opposites, words lying in an inherently incompatible binary relationship. For example, new is an antonym of old.
  •  Semantic information may include semantic category, register, sentiment, domain, and other indications of meaning such as context or application.
  • ·Syntactic information offers details about the tense, aspect, subcategorization, valency, case, or mood.
  • ·Etymology provides information about the origin and history of the word.

Make sure to check all needed data types in the feature list of your chosen API before deep diving into its implementation and integration in your platform.

Technical Usage

Besides narrowing down the list of potential APIs according to your language needs, you should not neglect the technical aspects of the database user experience. It is crucial to ensure a safe, agile, and amicable connection to the dictionary resource to avoid the occurrence of malfunctions on your platform.

But what are the parameters displaying the essential qualities and where can you find them?

1.     Latency is a parameter used to evaluate the speed of response to a call by the database. It is measured in MS (milliseconds). Our open web research shows that the average latency among the leading dictionary APIs is roughly 3,500 MS. Obviously, the lower the latency, the better the API functions.

2.     Success rate is an indicator of the percentage of successful calls being replied to by the API. This can reflect the richness of the data, but also its flexibility in responding to diverse queries.

In addition, while both factors listed above are quantitative, measured by definitive means, the user experience is a practice to be gauged by your personal preferences, and you may find other elements affecting your own experience.

Last but not Least: Pricing

As of October 2020, most dictionary API owners offer use of their data on a freemium basis, charging users only when they exceed a certain number of calls. As mentioned in previous sections, it is important to conduct a comparison of the different options according to your specific requirements and budget.

Lexicala Multilingual Dictionary API

Lexicala API is based on dictionary resources developed continuously since the 1990’s, spanning 50 languages and featuring 25 monolingual sets, more than 150 language pairs and numerous multilingual combinations, suiting most types of language requirements. This is the main source of our pride in this quality product. Additionally, our dictionary API is highly compatible for both latency (scored 97MS, calculated by RapidAPI) and success rate of calls. For more technical information, visit our review page on RapidAPI.

We offer a very affordable pricing model for both heavy users and freemium seekers to use the Lexicala dictionary API. For more information visit the Lexicala dictionary API homepage.   

Leave a comment