NeuroGram is a system used for morphosyntactic analysis (defining the lemma, the part of speech and a detailed grammatical description) and for surface syntactic analysis of texts in the Polish language. The software may be used as an element of the system for indexing and searching any full-text database in which there is a need for a precise term search taking into account Polish inflection as well as many other relations defining word origins.
NeuroGram is a great tool supporting the expansion or analysis of queries (through morphosyntactic analysis, acronym interpretation, identification of particular phrases and accounting for relations of origin).
The features of NeuroGram include:
- normalization of a given text, ie. lemmatization – changing words into their base forms (e.g. in the case of nouns – into nominative singular, in the case of verbs – into an infinitive, etc.)
- highlighting keywords in a text (e.g. in order to mark them a different color in a larger document)
- generating all inflectional variants of a given basic term
- identifying and expanding abbreviations in a text, and in version 3 – identifying many kinds of phrases (such as person, company or organization, location, name of product, temporal phrases and other)
The application software is targeted at integrators and programmers who want to take full advantage of the NeuroGram system’s possibilities in creating new solutions (software, websites, services). NeuroGram is a REST server which only requires the knowledge of XML or JSON. Additionally, it only requires Java Virtual Machine to run.