The present tool, that was built to deal with specific issues concerning orthographic conventions adopted for Portuguese, marks sentence boundaries with <s>…</s>, and paragraph boundaries with <p>…</p>. Unwraps sentences split over different lines. A f-score of 99.94% was obtained when testing o...
The paragraph splitter is a web service tool which takes text as input and outputs the identified paragraphs surrounded by tags. The tool is language independent. The download for this resource only contains the narrative description in a Word file. The service has one method which can be invo...
Yake! (Campos et al. 2020) is a novel feature-based system for multi-lingual keyword extraction, which supports texts of different sizes, domain or languages. Unlike most of the systems, Yake! does not rely on dictionaries nor thesauri, neither is trained against any corpora. Instead, we follow a...
MBT is a memory-based tagger-generator and tagger in one. The tagger-generator part can generate a sequence tagger on the basis of a training set of tagged sequences; the tagger part can tag new sequences. MBT can, for instance, be used to generate part-of-speech taggers or chunkers for natural l...
Bilingual concordancer is a language independent concordancer tool for bilingual concordancing, translation revision, post-editing, etc. Note that the tool is also able to be used as a monolingual concordancer. Several corpora are also included in this resource.
Conta-me Histórias [http://contamehistorias.pt] is a temporal summarization framework of news articles that allows users to explore and revisit events in the past. To select relevant stories of different time-periods, we rely on YAKE! [http://yake.inesctec.pt] a keyword extraction algorithm devel...
Monolingual concordancer is a language independent concordancer tool. Note that the tool is also able to be used as a bilingual concordancer. Several corpora are also included in this resource.
Treat is a toolkit for natural language processing and computational linguistics in Ruby. The Treat project aims to build a language- and algorithm- agnostic NLP framework for Ruby with support for tasks such as document retrieval, text chunking, segmentation and tokenization, natural language pa...
This is a UIMA component that provides a visualization of speech based output from UIMA workflows. It has been developed at the University of Manchester, using libraries of the Java Speech Toollkit (jstk). It has been designed specifically for use with the U-Compare text mining workbench (see sep...
The purpose of the U-Compare platform is to facilitate easy and rapid development and evaluation of NLP and text mining systems. It includes utilities (including a graphical user interface, the U-Compare workbench, see separate record in META-SHARE) to create workflows from individual, interopera...