2th Brazilian Symposium in Information and Human Language Technology
Salvador, BA, Brazil, October 15-18, 2019
>>>> Deadline for paper submissions:
May 19th, 2019   May 26, 2019 (new deadline) <<<
STIL is the bi-annual Language Technology event supported by the Brazilian Computer Society (SBC - http://www.sbc.org.br) and by the Brazilian Special Interest Group on Natural Language Processing (CE-PLN - http://www.sbc.org.br/14-
In 2019, it will be held in conjunction with both BRACIS 2019 (The 8th Brazilian Conference on Intelligent Systems) and ENIAC 2019 (XVI Encontro Nacional de Inteligência Artificial e Computacional).
STIL will have the following collocated events: the VI Workshop on Portuguese Description (JDP), the VI Student Workshop on Information and Human Language Technology (TILic), and the second edition of the shared task Evaluation of Semantic Textual Similarity and Textual Inference in Portuguese (ASSIN 2).
The conference has a multidisciplinary nature and covers a broad spectrum of disciplines related to Human Language Technology, such as Linguistics, Computer Science, Psycholinguistics, Information Science, among others. It aims at bringing together both academic and industrial participants working on those areas.
STIL 2019 welcomes research work in human language technology in general (and not only Portuguese) in various fields. Topics of interest include, but are not limited to:
GUIDELINES FOR PAPER SUBMISSION
Papers can be written in Portuguese, English, or Spanish.
We accept submissions of long and short papers. Long papers should describe complete work with significant results. Short papers may report work in progress, negative results, opinion papers, or application papers.
Long papers may have up to eight (8) pages of content (including tables and pictures), with two (2) additional pages of references, and will be presented orally. Short papers should have up to four (4) pages of content, and one (1) additional page of references, and will be presented as posters. Authors should also indicate whether they accept their long paper to be reallocated as a poster should the reviewers recommend so.
Paper formatting must follow the SBC guidelines available at this address: http://bit.ly/2jFbJTa
All submitted papers will be reviewed by at least two experts in the field. The reviewing process will be double blind and therefore papers should not display any information regarding their authorship in the header or body of the text. Self-references that reveal the author’s identity, e.g., “As we previously showed (Silva, 2005) ...”, must be avoided. Instead, authors should use “Silva previously showed (Silva, 2005) ...”.
By submitting papers to STIL 2019, authors agree that in case of acceptance at least one author will register to the conference and present the paper. Furthermore, it is the conference policy that at least one of the authors of accepted papers has to register *before* the deadline for sending the camera-ready paper. Accepted papers without the respective author registration *before* the deadline will not be included in the online proceedings.
Important Dates (all deadlines are 11:59 p.m. UTC-12:00 - anywhere on earth!)
- Deadline for long and short paper submissions:
May 19th, 2019   May 26, 2019 (new deadline)
- Notification of acceptance: July 8th, 2019
- Final camera-ready versions due: August 4th, 2019
Long and short papers should only be submitted in PDF files via JEMS system https://submissoes.sbc.org.
Carlos Augusto Prolo (UFRN, Brazil)
Leandro Henrique Mendonça de Oliveira (Embrapa, Brazil)
LOCAL ORGANIZATION CHAIR
Marlo Vieira dos Santos e Souza (UFBA, Brazil)
Title: A History of Research: Textual and Terminological Accessibility in Public Utility Texts in Brazil
Bio: M.J.B. Finatto is a Professor of Linguistics at the Department of Linguistics and Philology (UFRGS), and a Brazilian Research Fellow with CNPq. She completed her first postdoctoral research in Natural Language Processing and Readability Assessment at the Institute for Mathematical and Computational Sciences –ICMC, of the Inter-institutional Center for Research and Development in Computational Linguistics – NILC, at the University of São Paulo, in 2011. She completed her second postdoctoral research at the University of Évora, in 2017, with the project Historical Corpora of Portuguese. Finatto has a PhD in Language Studies, Linguistics, Terminology and Terminography (2001). She holds a M.A. in Linguistics and Lexicography (1993) and a B.A. in Languages and Literature – Portuguese and German (1991). Her research interests involve Natural Language Processing, Corpus Linguistics, Descriptive Terminology, Lexicography, Lexicology, Translation, Text and Discourse Studies and Scientific Communication to the General Public.
Abstract: The aim of this presentation is to present an overview of the scenario of Applied Linguistics and Terminology research in Brazil. These investigations, to a certain extent, connect with the issue of accessibility of written information for the general public, especially adults with limited education. In our analysis of the national scenario, we will be looking at the history of the research group “Textual and Terminological Accessibility” (ATT). Since 2011, graduate-level researchers in our group have been working on multiple perspectives in this area, in the line of research “Lexicography, Terminology and Translation” at the Universidade Federal do Rio Grande do Sul. We have drawn insights from a number of areas, including ATT, through which we rely heavily on computer technology, as well as Corpus Linguistics and Natural Language Processing. We seek to analyze texts, discourses, terminologies, lexicon and writing conventions from different areas, under the perspective of intralingual translation. Our efforts have been carried out to support actions that may facilitate the understanding of public utility information, especially in the areas of Health care and Law. (Sponsors/Grants: CNPq / CAPES / FAPERGS / SEAD-UFRGS).
Title: Language models (Word Embeddings) - this tutorial will be presented in Portuguese
Bio: Renata Vieira holds a PhD in Computer Science from the University of Edinburgh (1998). She is a professor at PUC-RS where she works in research and teaching in the area of computational intelligence, with emphasis on natural language processing, knowledge representation, ontologies, agents and semantic web. Renata coordinates the Artificial Intelligence Center of the Polytechnic School, coordinates the Natural Language Processing Research Laboratory and is the leader of the CNPq Research Group in this area. She has experience in coordinating inter-institutional and international projects, and participates in several national and international scientific event program committees (STIL, BRACIS, PROPOR, LREC, FLAIRS, FOIS, IJCAI, ACL). She participated in the creation of the Brazilian Computer Society's NLP Special Committee, being the first chair of this committee from 2007 to 2009. With a Senior Visiting Researcher scholarship, CAPES-Fulbright, she visited the University of Texas at Austin in 2007. She participated in the Association Executive Committee. for Computational Linguistics from 2011 to 2013. In 2017 she visited the University of Toulouse. In two consecutive editions, she received the award for best paper at the Brazilian Symposium on Information Technology and Human Language. She is currently nominated by SBC as a Senior Lecturer in Natural Language Processing.
Joaquim Santos has a degree in Mathematics from the Regional University of Cariri (URCA) and is currently a MSc student in Computer Science at the Pontifical Catholic University of Rio Grande do Sul (PUCRS), under the supervision of Renata Vieira. His main area of study is Natural Language Processing with emphasis on Brazilian Portuguese. He has been conducting research on classic and newer Language Models (such as BERT and Flair Embeddings) and applying them to the problem of Named Entity Recognition for Portuguese, as in his most recent work: "Assessing the Impact of Contextual Embeddings for Portuguese Named Entity Recognition". His main topics of interest are Neural Networks, Language Models, Relation Extraction and Named Entity Recognition.
Abstract: Recent work in the area of Natural Language Processing has been impacted by sophisticated Language Models, known as Word Embeddings. Such language models capture information from the context in which the words appear and also from the characters they are composed by. In this tutorial we will talk about the evolution of these language models, how they are generated and used. We will also talk about evaluation questions of these models and their main limitations. We will also introduce some of the current Portuguese language models that are available for free use.