LSP (Lexical Semantic for Portuguese) aims to aggregate lexical and semantic information of Portuguese words on the same platform, allowing users to directly query for both lexical and semantic analysis of words in real-time. This project is being used for POS-tagging of Portuguese tags and other more specific tasks concerning text analysis and information extraction. Technical reports available here and here.

This project was developed at LIACC (artificial intelligence and computer science laboratory) and FEUP (Faculty of Engineering of the University of Porto).

This small project was fully developed on a 24h programming contest at CodeBits 2009. It was a mobile game application which purpose was to give to the user quotations without the speaker so the user could guess it based on 3 different clues.

This project was the 3rd classified on this contest.

twitterEcho is a research platform which comprises a distributed focused crawler for the twittosphere. Currently, this platform includes modules for crawling the portuguese twittosphere. Future implementations will include modules for topic-focused crawling as well as social network-focused crawling.

The platform is being developed at the Faculty of Engineering of the University of Porto, in the scope of the REACTION project and in collaboration with SAPO Labs. More details are available here.

Verbatim is a project that aims at developing modules for automatic language processing of news in Portuguese language and published on the web. The main R&D focus of this project is to create methods to automatically identify and extract quotations (direct and indirect) from news, as well as to follow news topics for later automatic news classification. This project is now a SAPO.pt product named Voxx, available at http://voxx.sapo.pt.

This project is being developed at the SAPO Labs, in collaboration with LIACC (artificial intelligence and computer science laboratory) and FEUP (Faculty of Engineering of the University of Porto) and is available here.

Verbetes project main objective is to develop method of automatic extraction of biographic data of people and organizations from news published on the web. This type of information is considered important and useful for other tool of Information Extraction (namely project Verbatim), and on the task of enhancing other media applications. The goals of this project are to create mechanisms for the extraction and aggregation of all this information in a completely automatic approach.

Verbetes ia available here, in the form of REST services, several methods to query is base of knowledge, that allows the user to have answers for queries like “Who is Barack Obama?”, “Who is the actual minister of defense of Israel” of even disambiguate a name based on the context where is occurs.

The identification of co-occurrences between people on news is also part of the Verbetes project, and is also available as a REST service here.

This project is being developed at the SAPO Labs, in collaboration with LIACC (artificial intelligence and computer science laboratory) and FEUP (Faculty of Engineering of the University of Porto).

Twitómetro is an online tool that measures the sentiment of the portuguese twitter users regarding the 5 main candidates for the Portuguese elections in 2011. A technical report is available here.

This tool was developed by in the scope of the REACTION project and is available here.

Semantic Lists is a project that aims at the construction of a linguistic resource made by lists of words grouped by its semantic category, as jobs, types of organizations, nationalities, etc. The main goal of this project is to provide lexical-semantic information and support the development of Information Extraction systems and Text Classification systems.

These lists are availiable here, as a REST service.

This project is being developed at the SAPO Labs, in collaboration with LIACC (artificial intelligence and computer science laboratory) and FEUP (Faculty of Engineering of the University of Porto).

REACTION (Retrieval, Extraction and Aggregation Computing Technology for Integrating and Organizing News) is an initiative for developing a computational journalism platform (mostly) for Portuguese. This project is mostly focused on : (i) automatic analysis of content, including news, blogs, micro-blogs and comments; (ii) automatic analysis of explicit and implicit social networks; (iii) design of rich visualization and interaction interfaces and (iv) case-study evaluation of developed computational journalism methodology.

“O mundo visto daqui” are weekly infographic series that analyze the portuguese social media and build ego-centric networks with the relations between public personalities found on news. This is a SAPO Labs project, in collaboration with SAPO Notícias and coordinated by Eduarda Mendes Rodrigues and is available here.

“2011 nas notícias” is a overview of the most relevant events that happened in 2011 and were published on the portuguese news. The analysis of the most relevant events was performed automatically based on the “key-words” found on news and the names of public personalities mentioned. This project is a collaboration between SAPO Labs, the University of Porto and SAPO Notícias, and is available here.

MVDI is an interactive news social network that identifies co-occurrences of public personalities on news. This is an interactive and enhanced version of “O Mundo Visto Daqui” project. This is project developed in collaboration with University of Porto (FEUP and LIACC), Sapo Labs and SAPO Notícias in the context of REACTION project. This project is available here.