Digital Sonata

Digital Sonata
intelligent solutions for language processing

Carabao Language Kit
The logo of Carabao Language Kit

Carabao Language Kit is a family of products whose main purpose is to understand and transform text. All products of the Carabao family can be customized by means of data entry via the desktop suites. All the server components are COM objects compatible with .NET and ASP / ASP.NET. Other programming interfaces are available upon request.

If you require a demonstration, or an evaluation license for a product not available for public download, please do not hesitate to contact us.

Components

Carabao MorphoLogic

Carabao MorphoLogic 		Analyzes single words. Returns all possible senses of the word and grammatical data. Analysis of unknown words and patterns is also supported.

List of features
  • Grammatical reference (part of speech and more)
  • Stemming
  • Lemmatization
  • Synthesis of all inflections
  • Semantic reference (hypernyms, hyponyms, pertainyms and domains)
  • Thesaurus & sense ID for every interpretation

Carabao Transliterator

Carabao Transliterator 		Converts equivalent characters between any two languages in the database.

List of features
  • Transliterates from one script to another, e.g. Arabic to Cyrillic, or Greek to Latin.
  • Overrides supported. For example, if "sh" and "h" are defined, in a word like "shoe" "sh" takes precedence.
  • Supports position-based recognition and normalization (e.g., in Hebrew a letter can transform into a different letter when positioned last).
  • A new script can be added in a couple of hours. Currently provided: Latin (English), Cyrillic (Russian), Greek, Hebrew.

Non-interactive demo

Carabao DeepAnalyzer

Carabao DeepAnalyzer 		Analyzes texts.
Extracts word sense IDs, idioms, linguistic profiling data, domains of discourse (or subjects) for every sentence, and machine readability index.

List of features
  • Domain extraction for every sentence
  • Automatic linguistic profiling
  • Sense disambiguation for every word (statistics, rule & neural network based)
  • Full context-dependent named entity extraction for over 100,000 entities
  • Idiom extraction
  • Grammatical tagging (part of speech and more)
  • Thesaurus & sense ID for every word and idiom
View demo

Carabao Machine Readability Indicator

Carabao Machine Readability Indicator 		Evaluates how difficult a text is for natural language software - which indicates the reliability of the processing results.

Carabao MeasureConvert

Carabao MeasureConvert 		Converts measures and metrics inside free flowing text as specified.

List of features
  • Supports "mixed" mode (e.g., both Fahrenheit and Centigrade are in the same text)
  • Based on the powerful Carabao sequence paradigm, which enables defining any combination of values and units (temperature, length, weight, shoe sizes, currencies).
  • Supports named entities yielding numerical values, such as fractions or sums of money.

Carabao Translation Server

Carabao Translation Server 		Analyzes, transforms or translates text aggregating the functionality of all the components of Carabao.

List of features
  • Translates from any language in the database to any other language in the database.
  • Paraphrasing / altering usage linguistic profiling styles while processing. It is also possible to process the same language, modifying styles.
  • Extracts domains and styles used in a text. Range of usage is returned as well.
  • Disambiguates words and returns sense ID and extensive grammatic information.
  • Thesaurus articles (for the actual sense only) for the words and idioms.