From IDSS Group at INESC-ID
To obtain additional information about the computational resources that we distribute, please contact email@example.com
- BIG-Corpus-PT - Portuguese corpus composed of news articles, news comments, blog posts and twitter messages crawled during a period of 7 months.
- Twitter-BrownClusters-PT - word clusters induced from Portuguese Twitter messages.
- SentiLex-PT - sentiment lexicon for Portuguese.
- NomesLex-PT - lexicon of person names from Portugal.
- POWER-PT ontology and SPARQL Endpoint - Portuguese politics.
- SentiCorpus-PT - comments in Portuguese manually annotated with sentiment and opinions about politicians.
- SentiTuites-PT - corpus of tweets posted by Portuguese users during the 2011 election campaign.
- DBpediaEntities-PT - corpus of entities(persons, places, organisations) extracted from the Portuguese DBpedia.
- DBpediaRelations-PT - corpus of semantic relationships(persons, places, organisations) extracted from the Portuguese Wikpedia and DBpedia.
Reaction-developed Online demos
- Publico's news semantified by Rembrandt
- 10 years of Publico Collection: Search People, places and organisations in 10 years of Público news articles
- 10 years of Publico politics section social network A tool for social network analysis
- TwitterEcho (open source).
- Meme browser built on the TREC BLOGS08 collection ( open source).
- Opinionizer, a tool for detecting sentiment polarity in a text sentence written in Portuguese.
- Sentibubbles A tool that shows sentiment bubbles (SentiLex-PT keywords) in a stream of tweets having a given term ("merkel" in the link above)
- Entity LInking A tool for resolving entities in news text.
- Power Profiles A tool to visualize information about politicians in POWER.
- Ngram search tool on a set of Publico news from the first decade of the century
Websites running Reaction-developed Software
- Twitómetro, a political barometer for the community of Portuguese Twitter users. The goal of the application was to identify messages targeting at least one of the 5 political candidates for the 2011 legislative elections and perform sentiment analysis. The messages were classified as either positive, neutral or negative in terms of the opinion expressed towards the targets. A more detailed report on the experiment can be found here
- O Mundo Visto daqui, an interactive application that generates a visualization of the social network of any person named in Portuguese news media. This application has been deployed live by one of our industrial partners (SAPO PTC) at http://noticias.sapo.pt/interativo/omundovistodaqui/. MVDI was developed under a collaboration between FEUP and SAPO Labs | UPorto.
- Twitteuro, a system for analysing the sentiment about the participating teams and players during the Euro2012. The TwitterEcho platform collected tweets that were analyzed with the Opinionizer Software.
- O Mundo em Pessoa, a system for automatic collection of Fernando Pessoa ciatations.
- Máquina do Tempo The Mundo Visto Daqui software incorporated in another site that seaches on a large set of media, including 25 years of the LUSA archive.
- WPT Crawls of the Portuguese Web:
- REMBRANDT - NER system
- Geo-Net-PT - Geographic Ontology
- GeoSSM Geographic Similarity Calculator