Unique cutting-edge linguistic technologies developed by ABBYY enable ABBYY Intelligent Search SDK to provide an unprecedentedly high relevance and recall of search results. Sophisticated syntactic and semantic analysis allows for the creation of rich semantic indexes from enterprise data repositories - as well as analysis that establishes a true "understanding" of the meaning of users' requests.
How well a computer system copes with search and text analysis largely depends on how well the system “understands” texts. Understanding requires highly sophisticated linguistic analysis, which enables computers to work with meanings rather than with character strings. Accurate interpretation of the meaning of texts is what delivers accurate translations and relevant search results.
ABBYY technologies represent a major step forward in this direction. They parse sentences syntactically and semantically to detect the true meaning of words and establish how they relate to one another. The result of this parsing is a language-independent representation that enables the computer to understand and utilize the true meaning of the text to deliver remarkably precise search results. As a result, ABBYY technologies transform strings of characters into data that "makes sense" to computers - enabling them to utilize it in a wide range of new ways.
Natural language is highly ambiguous. ABBYY technologies detect homonyms (words with the same spelling but different meanings) and select the right meaning by analyzing the surrounding context. The result is a more accurate search, information extraction, classification, translation and other text-processing tasks outcomes.
ABBYY technologies ensure accurate definition of entities (including persons, organizations, geographical objects, financial indicators any many others) and events (meetings, speech activities, occupations, ownership, etc). Their ability to analyse relations between words, ABBYY technologies deliver the most consistently accurate of entity recognition.
See right example of entity extraction from the sentence:
China Eastern bought 70 Airbus A320NEO for $6.4 billion.
Identifying links between words is paramount in search and text analysis. ABBYY technologies are capable to solve major NLP challenges that directly affect the quality of relation analysis in texts, including:
When analyzing events related to certain objects, how those events and objects are linked together has to be understood. ABBYY technologies can detect all relations among words in a text, including relations between words that stand far apart – in different sentences and even different texts.
ABBYY technologies can establish links between words and pronouns or other words that substitute them (e.g. he, she, it, the company and other entities).
Ellipsis is a complex example of an uncompleted syntactic construction when a word or phrase is omitted because human can easily infer it from the context. ABBYY technologies can tackle even these problems.