Author profiling
Author profiling has gained considerable relevance due to its potential applications, for instance, in forensic linguistic studies, marketing analysis, and historic/literary text authorship verification. The goal of author profiling is to automatically extract demographic traits of the author of a text, such as gender, age, native language or sexual orientation.
Our approach focuses on structure rather than lexis. Syntactic and discourse-based features combined with character-, word-, sentence- and dictionary-based features compose our feature set (which has less than 200 features).
This demo showcases the capabilities of our author profiling approach. It allows the user to input a text and it outputs several predictions, namely the most likely gender of the author, the most similar literary author of our collection and the most stylistically similar book of our collection.
To know more about the work behind this demo, refer to:
Soler Company, Juan. Feature engineering for author profiling and identification: on the relevance of syntax and discourse. 2017
hdl.handle.net/10803/404984
To the demo is available here: /quality-taln.upf.edu/author-profiling-demo/profiling/
Online tool for PDF-to-XML conversion
You can use our freely online tool to parse yours PDF files. Our approach is based on the PDFdigest tool, a PDF textual content extraction system specially designed to extract scientific articles' headings and logical structure (title, authors, abstract and so on) and its textual content to. The result is provided in a XML file. Furthermore, PDFdigest also provides a structured HTML file as a clone of the original PDF file.
The tool is available in http://taln.upf.edu/pdfdigest
You can only process a maximun of 10 PDF Files at a time. If you have a large amount of PDF files, please contact us at horacio.saggion AT upf.edu.
References
- Saggion H, Ronzano F, Accuosto P, Ferrés D. MultiScien: a bi-lingual natural language processing system for mining and enrichment of scientific collections. InMayr P, Chandrasekaran MK, Jaidka K, editors. Proceedings of the 2nd Joint Workshop on Bibliometric-enhanced Information Retrieval and Natural Language Processing for Digital Libraries (BIRNDL 2017); 2017 Aug 11; Tokyo, Japan.[place unknown]: CEUR Workshop Proceedings; 2017. p. 26-40. 2017. CEUR Workshop Proceedings.
Web Link, BibTeX, EndNote, RefMan and RefWorks. - Ronzano F, Saggion H. Dr. inventor framework: Extracting structured information from scientific publications. InInternational Conference on Discovery Science 2015 Oct 4 (pp. 209-220). Springer International Publishing.
Web Link, BibTeX, EndNote, RefMan and RefWorks.
Praat on the Web
This system is an implementation of the widely used speech analysis tool Praat as a web application with an extended functionality for feature annotation. In particular, Praat on the Web addresses some of the central limitations of the original Praat tool and provides (i) enhanced visualization of annotations in a dedicated window for feature annotation at interval and point segments, (ii) a dynamic scripting composition exemplified with a modular prosody tagger, and (iii) portability and an operational web interface. Speech annotation tools with such a functionality are key for exploring large corpora and designing modular pipelines.
If you use this software please cite:
Domínguez M, Latorre I, Farrús M, Codina-Filbà J, Wanner L. Praat on the Web: An Upgrade of Praat for Semi-Automatic Speech Annotation. 26th International Conference on Computational Linguistics (COLING 2016) Osaka, Japan p. 218-222
Text Simplification for English
This site is a Text Simplification demo for English texts (up to 384 characters). The YATS simplifier was designed for people with intellectual disabilities. The automatic Text Simplification process aims to aid in text reading by reducing the lexical and syntactic complexity of sentecnes. Basically the text simplifier detects complex words and sentences and generates simpler vocabulary and shorter sentences. This technology has been investigated and developed by the TALN(Natural Language Processing) group of Universitat Pompeu Fabra (UPF).
Text Simplification technologies
Text simplification is the process of transforming a text into an equivalent which is more understandable for a target population. Simplified texts are appropriate for many groups of readers, such as language learners, elderly persons and people with other special reading and comprehension needs. TALN-UPFgroup develops robust natural language processing technology to produce simplified versions of documents at both syntactic and lexical levels.
Publications
Daniel Ferrés, Montserrat Marimon, Horacio Saggion, Ahmed AbuRa’ed
YATS: Yet Another Text Simplifier.
NLDB (2016) [PDF] [BibTeX]
Daniel Ferrés, Montserrat Marimon, Horacio Saggion
A Web-based Text Simplification System for English.
Procesamiento del Lenguaje Natural 55: 191-194 (2015). [PDF] [BibTeX]
Acknowledgements
Project ABLE-TO-INCLUDE
Online demo of the patent summarization system
Online demo of the patent summarization system. The system performs a linguistic analysis of the patent document, selects the most important passages and combines them in an abstractive summary.
The Demo is the result of several european projects and the experience in our group on patent processing.
To use the demo please contact joan dot codina at upf dot edu to ask for a user/password
A very easy to use e-mail web client adapted to people with IDD so that they can use it at their working activity.
Features
1. Use of the ABLE accessibility layer: Text2Picto, Text2Speech and Simplext.
2. Very easy interface.
3. Few functionalities: New message, Sent, Inbox and Contacts.
4. Accessible for all kind of dissability.
Demo page (use the unsafe mode when connecting to google)
PDFDigest
Online tool for PDF-to-XML conversion
Freely online tool to parse PDF files. Our approach is based on the PDFdigest tool, a PDF textual content extraction system specially designed to extract scientific articles' headings and logical structure (title, authors, abstract and so on) and its textual content too. The result is provided in a XML file. Furthermore, PDFdigest also provides a structured HTML file as a clone of the original PDF file.
In addition, the pre-processing step implemented in DrInventor (link) is applied to the previous XML file in order to mark off tokens and sentences. As a result, we also provide an additional GATE document.
OpenMinted Summarization Services
In the current context of scientific information overload in which new knowledge is created at a rapid pace, we propose to develop text summarization services for automatically identifying the most important information of a research article. The work will be based on an adaptation of our current scientific text mining and summarization technology at our LaSTUS/TALN lab. The summarization system will apply a natural language processing pipeline for deep analysis of scientific documents and compute a series of sentence relevance features based on the results of text analysis.
OpenMinted Babelnet extractor
UIMA component that detects BabelNet terms in text. It can be used with any of the 271 languges available in Babelnet 3.7
OpenMinted Freeling to UIMA integration
UIMA component that detects BabelNet terms in text. It can be used with any of the 271 languges available in Babelnet 3.7
Several applications from the computational analysis of language
This video is about the research that is carried out by the TALN group, Tractament Automatic del Llenguatge Natural (Natural Lenguage Processing), at Pompeu Fabra University in Barcelona.
Patent claims
Summarization and paraphrasing of patent claims.
Summa toolkit
This video shows how to create summarization applications with the SUMMA toolkit.
SUMMA is a set of language and processing resources implemented using the GATE NLP framework.