Open Source Mac Linguistics Software - Page 4

Using rules written in a Drools decision table, BuckTagger determines the correct Buckwalter Tag based on morphological properties of the input, automatically extracted or given by the user. At the moment, BuckTagger is not complete; it can only handle input that is: - Uninflected - In lexical form, i.e., no clitics or affixes. - A Perfect or Imperfect Verb - Preferably the first and before-last letters are diacritized/vocalized. The interface is in Arabic. See the README for more details. There is much room for development. Feel free to comment.

C4 is a C++ class library for analyzing sound files, particularly spoken and sung phonations. C4 provides features such as frequency analysis, pitch extraction, or calculation of voice quality parameters (e.g. alpha ratio, HNR, jitter, etc.).

Connecting Historical Authorities with Links, Contexts and Entities. CHALICE is a historic placename gazetteer for the UK, published as Linked Data and linked to other widely-used sources of placename reference information on the semantic web.

CLEiM (Cross Lingual Education in Medicine) is an opensource version of an Intelligent System which extract concepts from medical texts and provides qualified information. It integrates information from various sources. This system has been developed by the Intelligent System Group GSI (http://www.esi.uem.es/gsi/) at UEM University. We do NER (Named Entity Recognition) based on GATE platform. The installation is simple, you can use it as a Web application. It has been tested under apache-tomcat. The original system has been successfully used to carry out active learning activities with medical students. However, it could be interesting in much more knowledge fields.

CRFSharp(aka CRF#) is a .NET(C#) implementation of Conditional Random Fields, an machine learning algorithm for learning from labeled sequences of examples. It is widely used in Natural Language Process (NLP) tasks, for example: word breaker, postagging, named entity recognized, query chunking and so on. CRF#'s mainly algorithm is the same as CRF++ written by Taku Kudo. It encodes model parameters by L-BFGS. Moreover, it has many significant improvement than CRF++, such as totally parallel encoding, optimizing memory usage and so on. Currently, when training corpus, compared with CRF++, CRF# can make full use of multi-core CPUs and only uses very low memory, and memory grow is very smoothly and slowly while amount of training corpus, tags increase. with multi-threads process, CRF# is more suitable for large data and tags training than CRF++ now. For example, in machine with 64GB, CRF# encodes model with more than 4.5 hundred million features quickly.

This application was developed by the NIHR Biomedical Research Centre at the Institute of Psychiatry and South London and Maudsley NHS Foundation Trust, in collaboration with the University of Sheffield. Its purpose is to identify the smoking status of a individual, based on text evidence in clinical notes. Currently, it classifies patients as 'current', 'past' or 'never'. It runs on the GATE infrastructure, available at http://gate.ac.uk/. Please contact richard.g.jackson@slam.nhs.uk for support/queries.

Distributed phrase-based machine translation training tool based on Hadoop.

Colloquium QDA is a tool for custom coding and analyzing qualitative ethnographic interviews. To run, make sure you first have JRE 8 or later installed (http://www.oracle.com/technetwork/java/javase/downloads/). Colloquium QDA is an open source cross-platform Java Swing app utilizing an embedded Java DB with Lucene integrated search.

ConTextKit is a Java-based implementation of Wendy Chapman's ConText algorithm for annotating the context of medical documents, specifically the negation, temporality, and experiencer.

A Demo is available at: http://coocviewer.sourceforge.net/coocviewer/index.php

CORPSE (CORPus SEarch) is a powerful search engine written in Java. The aim is to provide an efficient implementation of a word level inverted index search with various cool functions that can be used on very large corpora.

Redundancy due to cut-paste operations in text creates bias in machine learning for NLP. This module takes a directory and produces a subset of the files in that directory (in a list) with an upper bound on similarity between two files.

AFEWC corpus is a multilingual comparable text articles in Arabic, French, and English languages. Each triple article is related to the same topic (aligned at article level). AFEWC corpus is collected from Wikipedia. The corpus is available for free for research purposes only. It is composed of 40K aligned articles, 91.3M English words, 57.8M French words, 22M Arabic words, 2.8M English unique words, 1.9M French unique words, and 1.5M Arabic unique words. Wikipedia text is available under Creative Commons Attribution-ShareAlike 3.0 License. https://en.wikipedia.org/wiki/Wikipedia:About To cite the corpora: M. Saad, D. Langlois, and K. Smaïli. Extracting Comparable Articles from Wikipedia and Measuring their Comparabilities. Procedia - Social and Behavioral Sciences, 95(0):40 – 47, 2013. ISSN 1877-0428.

Cunei is a data-driven machine translation system that builds dynamic, statistical models based on instances of known translations found in a corpus.

Web pages do not offer reliable metadata concerning their creation date and time. However, getting the document creation time is a necessary step for allowing to apply temporal normalization systems to web pages. DCTFinder is a system that parses a web page and extracts from its content the title and the creation date of this web page. DCTFinder combines heuristic title detection, supervised learning with Conditional Random Fields (CRFs) for document date extraction, and rule-based creation time recognition. DCTFinder is released under CeCILL free software license agreement. The system is described in the following paper (see 'Files' section): Xavier Tannier. "Extracting News Web Page Creation Time with DCTFinder". Proceedings of the 9th Language Resources and Evaluation Conference. Reykjavik, Iceland.

DawNLITE is a Natural-Language-based Image Transmoding Engine. The software transforms an image to a video as recorded by a virtual camera panning and zooming over the image, following a natural language text description of the image.

Dendrarium służy do wybierania i weryfikacji składnikowych drzew składniowych generowanych przez parser Świgra. System jest użytkowany w Instytucie Podstaw Informatyki PAN do tworzenia banku drzew składniowych dla języka polskiego Składnica.

Dictionary Additions Management System (DAMS), a collection of open source translation dictionaries. These files are compatible with the Open Translation Engine (OTE). For more info, see http://sourceforge.net/projects/ote/

Text editor in Java that is able to detect discriminative expressions while the user is typing. When the internal ontology-based analyzer detects a potential discriminative expression the user is advised by underscoring the related words in the text. A descriptive message about the issue is also shown to the user when the cursor is placed over the potential discriminative expression.

Donatus is an on-going project consisting of Python, NLTK-based tools and grammars for deep parsing and syntactical annotation of Brazilian Portuguese corpora. It includes a user-friendly graphical user interface for building syntactic parsers with the NLTK, providing some additional functionalities.

Drug name recognition and normalisation/grounding to DrugBank ids and standard names. Package provides 2 taggers: 1. DrugTagger - CRF-based with DrugBank presence feature (see feature set for details). 2. DrugnameGazetteer - gazetteer/dictionary-based. Dictionary created from DrugBank.ca database. Both taggers include grounding/normalisation to DrugBank ids and standard names. Feature set: Word, Word-1, Word+1, Word-1_Word, Word_Word+1, DrugBankPresence, POS DrugBankPresence feature indicates the presence of the drug name in the DrugBank. Using CONLL-Evaluation: processed 32065 tokens with 3656 phrases; found: 3251 phrases; correct: 2786. accuracy: 95.25%; precision: 85.70%; recall: 76.20%; FB1: 80.67 Using GATE Corpus Benchmark: Strict: P: 0.65 R: 0.73 F1: 0.69 Lenient: P: 0.74 R: 0.84 F1: 0.78 The details of how to reproduce evaluation, see README. To use standalone version for tagging download DrugExtractionStandalone.tar.gz from Files.

PubMed Central browser. Source code: http://github.com/dualword/dualword-pmc/