NooJ_Association

NooJ

A Corpus Processor - A Linguistic Development Environment -  A Linguistic Engine for developing Natural Language Processing software Applications.

logo%20UFC

A Corpus Processor

placeholder image

Import text files in any format

Construct your own corpus by importing text files in 150+ file formats. Texts can be structured (e.g. XML).

placeholder image

Apply a query

Enter a regular expression to produce the corresponding concordance. The regular expression here:
    (will | shall) <ADV>* <V+INF>
recognizes sentences that contain a verb in the Future tense. 

placeholder image

Various Statistical Analyses

NooJ offers half a dozen statistical tools to measure the corpus vocabulary and concordances' results (interesting spike in future tense frequency in part #6 of the corpus).

placeholder image

A simple grammar

It takes a few minutes to draw a NooJ graph and apply it to a corpus.

placeholder image

A lexical field

Define and accumulate lexical fields via NooJ grammars (here: "death" theme) and apply them automatically to index and color texts, construct concordances, perform statistical analyses, etc.

placeholder image

Text Annotation Structure

All linguistic analyses' results are stored in the TAS. Linguistic resources are used to add and/or remove annotations to the TAS. TAS can be imported from or exported to XML files.

placeholder image

Automatic Disambiguation

Local grammars can be used to remove ambiguities. The Spanish word "la" is disambiguated as a pronoun or a determiner, depending on its right context.

C8%20Contract

Linguistic Engineering

NooJ contains a dozen tools to help develop and edit, test, debug, check consistency of all linguistic resources. Here: the grammar's contract garantees the grammar's consistency.

Develop Linguistic Resources

placeholder image

Dictionaries

NooJ's dictionaries manage any type of Atomic Linguistic Units (ALUs): simple words, multiword units, intra-word units as well as discontiguous expressions.

placeholder image

Inflection and Derivation

Inflectional and Derivational paradigms can be formalized with enhanced regular or Context-Free rules.

placeholder image

Term Derivation

Morphological grammars can be used to describe agglutinations, neologisms as well as sets of morphological variants (here: France).

placeholder image

Context-Free Grammars

Users can construct CFGs, either as sets of recursive rules or as recursive graphs.

placeholder image

Syntactic Trees

NooJ can display Derivation, Constituent and Dependency Trees, taking into account all types of ALUs. The text here contains two discontiguous ALUs: "ne ... pas" (French negation) and "baisser ... le ton" (French idiomatic expression).

placeholder image

Context-Sensitive Grammars

NooJ grammars can include various types of linguistic constraints existence, distributional restrictions, agreements, etc. The grammar here verifies agreement in Gender and in Number among all components of a French Noun Phrase.

placeholder image

Unrestricted Grammars

Unrestricted grammars allow NooJ to link syntactic and semantic variants, such as a noun and its pronoun, a verb and its nominalized form, etc. The grammar here represents all variants of a simple transitive sentence such as "Joe loves Lea', including aspects, modality, nominalizations, focus, pronouns, tense, etc.

placeholder image

Automatic Transformation Recognition & Generation

Transformational grammars can be used either to recognize or to generate texts automatically. Here, NooJ generates over one million transformed sentences.

Over a dozen Natural Language Processing Software Applications

09%20Disamb

Named Entity Recognition

NooJ has been used to develop NERs in over 20 languages. The graph here is part of a 20-graph grammar used to recognize and annotate English dates automatically.

10%20Grammar%20Checker

Grammatical Checker

Local grammars can be developed to recognize and annotate specific grammatical mistakes.

11%20LVF

Discourse Analysis

NooJ is being used for sentiment and opinion analysis. Here: a debate during the French 2017 Presidential Campaign shows the degree of assertiveness of the two participants: red participant clearly lost the debate.

12%20LVF%20Concordance

Automatic Annotations

NooJ is being used in various Business Intelligence software. Here, NooJ automatically annotates calls for tenders and links all legal references to the French database "Legifrance".

11%20Simple%20CFG

Semantic Disambiguation

NooJ manages lexicon-grammars as well as semantic dictionaries. Here, the LVF dictionary (Dubois & Dubois-Charlier 1997) contains 25,000 different verb meaning, associated with their characteristic syntactic and semantic contexts.

13%20Transformations

Intelligent Search Engine

By applying the LVF dictionary to one's corpus of texts, it becomes possible to pinpoint  a specific meaning of a verb. Here NooJ has found in the newspaper Le Monde Diplomatique, 2002, all occurrences of the verb abriter, sense #4: to hide oneself behind a pretext in order not to act.

14%20Lab

Semantic Analysis: Text to RDF to Text

NooJ grammars can be used to parse texts in Natural Languages and produce a semantic representation (e.g. in RDF), and reciprocally: produce all the sentences that can express a predicate represented by an RDF statement. 

placeholder image

Machine Translation

NooJ is being used to develop automatic translation systems: Arabic to French, Arabic to English, Portuguese to English, Greek to Spanish, etc. Here, a simple grammar recognizes some French NPs and translate them in English, reordering the adjectives and taking Number agreement into account.