
| Home \ Research Areas \ Research Stream | Login |
Multimodal Systems researches methods and tools to build interactive systems using information processing techniques. The research topics are defined in a context where multiple devices are used for interaction and an increasing amount of multimedia information is produced everyday. The potential of multimedia information can only be harnessed if the semantic content is properly represented. Additionally, the interface between the information space and the humans has to be carefully designed.
Alexandre Agustini, Pontifícia Universidade Católica do Rio Grande do Sul (Brazil)Visit to overview the work that could be done together in future
Gael Dias, Université de Caën Basse Normandie (France)Gael Dias was invited to argue the McS Thesis by Luis Teixeira named "Automatic extraction of document Topics (Extracção Automática de Tópicos de Documentos)". Meanwhile we discussed about on-going work in the framework of VIP-ACCESS project as well as on future coperation.
Irene Pimenta Rodrigues, CENTRIA (Portugal)Irene Rodrigues was invited to argue the MSc dissertation by Gonçalo Nuno Ramos Gomes on\"Support to translation extraction of simpe and multi-word terms in a multilingual environment (Suporte à extracção de traduções de termos simples ou compostos em ambiente multilingue)\"
Gael Dias, Université de Caën Basse Normandie (France)Gael Dias was invited to argue the MSc dissertation by Luís Gomes on "parallel Texts Alignment". It was also dicussed the on-going project VIP-ACCESS
Lucia Specia, Xerox Research Centre Europe (France)Lucia Specia has been invited as she has been selected as a possible post-doc to be contracted by FCT/UNL under the Programme Compromisso coma Ciência 2008. A presentation was made of ongoing work at Xerox. The post was not latter accepted as Xerox invited her to stay at Xerox research Centre Europe, where she was working at the time.
Seminars @ CITI: Predicting the Sentence-Level Quality of Machine Translation Systems
Alexandre Agustini, Pontifícia Universidade Católica do Rio Grande do Sul (Brazil)This visit was organized in order to analyse the possibility to continue cooperation in the area of word meaning discrimination using the knowledge acquired in the framework of Agustini's PhD Thesis and the full text indexing machinery implemented meanwhile by José Aires. It was also intended for programming future cooperation in bilateral projects including Universidade Federal do rio Grande do Sul.
Aline Villavicencio, Universidade Federal do Rio Grande do Sul (Brazil)Main objectives of Alene Villavicencio visit were: 1) participating the jury for discussion of Alexandre Agustini PhD Thesis that occurred on 23 November 2006; 2) discussing on future reserach collaboration with Universidade Federal do Rio Grande do Sul.
Pablo Gamallo, Universidade de Santiago de Compostela (Spain)Pablo Gamallo came mainly for participating in the jury for the PhD proofs of Alexandre Agustini that took place on 23 November 2006. Meanwhile we have discussed a lot on the ultimate research results obtained at both sides in Lisbon and In Santiago de Compostela.
Pavel Brazdil, LIACC - Laboratório de Inteligência Artificial e Ciência de Computadores (Portugal)Pavel Brazdil was one of the jury members for examining the PhD Thesis of Alexandre Agustini who, together with Irene Rodrigues, was in charge of careful examination of the thesis and the candidate.
Irene Pimenta Rodrigues, CENTRIA (Portugal)Irene Rodrigues, together with Pavel Brazdil, was one of the jury members who was in charge of careful examination of the PhD thesis of Alexandre Agustini and of interrogating the candidate.
António Branco, Faculdade de Ciências da Universidade de Lisboa (Portugal)António Branco was one of the jury members for the exam of the PhD thesis of Alexandre Agustini.
Juan Otero Pombo, Universidade de Vigo (Spain)The work developed in this visit focused on the extraction of agreement information from partial parses and its use for spelling correction purposes.
Juan Otero Pombo, Universidade de Vigo (Spain)The main target of his work was the integration of spelling correction and parsing tasks to improve the performance of both. Finding a good correction for a misspelled word is one of the currend challenges of the spelling correction tools but it is quite difficult without taking into acount some contextual infomation. Moreover, robust incomplete parsing performs better when mispelling erros are corrected. Juan Otero research has been centered arround the following topics: * Exporation of DyAlog tool, which alows the development of parsers in natural language processing. * Study of Defined Clauses Gramars (DCG's) used to implement the portuguese grammar used by my parser and its possible adaptation to spanish or galician. * Adaptation of regional spelling corrector to correct misspelled words before the parser was run. For each possible candidate to be a correction of a misspelled word, the spelling corrector obtains a prolog fact which can be processed by the parser.
Gregory Grefenstette, LIC2M, CEA-LIST (France)Gegory was invited as an invited lecturer to the workshop on Text Mining and Applications (TEMA05) that Gabriel Lopes, Joaquim Ferreira da Silva, Vitor Rocio and Paulo Quaresma have organized in the framework of the Portuguese Artifitial Intelligence Biannual Meating at Covilhã.
Víctor Bilbao, Universidade de Vigo (Spain)Victor Bilbao came to work on parallel text alignment and on evaluation of alignments.
Seminars @ CITI: Evaluation of Machine Translation Systems and of Parallel Text Alignement
Evandro Manara Miletto, Universidade Federal do Rio Grande do Sul (Brazil)Evandro came to Lisbon in the framework of WE-LEARN cooperation project to work on the ergonomics of the graphical interface for foreign Language training.
Ravi Kishore Singh, Indian Institute of Technology Guwahati, Assam (India)Ravi came to work on alignment of parallel texts written in English and Hindi.
Rosa Maria Viccari, Universidade Federal do Rio Grande do Sul (Brazil)This visit was done in the framework of WE-LEARN project.
Éric Villemont de la Clergerie, INRIA-Rocquencourt (France)Eric came in the framework of KLING cooperation project. We discussed on the ways how our cooperation could continue.
Víctor Bilbao, Universidade de Vigo (Spain)For working on parallel text alignment and evaluation.