Brazil

PROPOR 2022

Brasileiras em PLN panel: Computer science, linguistics and more: NLP and transdisciplinarity

EMNLP 2021 is planned to be a hybrid event in Punta Cana, Dominican Republic, with both on-site and fully virtual participation possible. The experience for on-site participants would closely approximate a normal pre-COVID *ACL conference, with 5-6 thematically organized parallel sessions and live Q/A and interactive discussion immediately after the talks. Presentations by virtual participants will be equitably interleaved with those of on-site participants, projected on the auditorium screens as if on-site, and also followed immediately by live Q/A and interactive discussion at a time during reasonable waking hours for the virtual presenter. For all participants, on-site and virtual, who are unable to attend a session due to either time-zone issues or because they are participating in another session live, talk recordings and slides will be available online at a minimum after the live presentation (and in many cases before as well), and questions may be submitted in advance on session-specific discussion boards and answered live in session with the usual visual aids if desired.

<iframe style="width:700px;height:400px" src="https://online.fliphtml5.com/ebtyf/ceby/"  seamless="seamless" scrolling="no" frameborder="0" allowtransparency="true" allowfullscreen="true" ></iframe>

Please Note: The EMNLP registration system is not currently connected to the underline site as we are still in the process of building out EMNLP 2021. You will receive access instructions from underline the week of November 1st. 

Access is given only to EMNLP upon registration, if you have not registered please do so [here](https://2021.emnlp.org/registration).

Registered attendees will receive access the week of November 1st.

EMNLP 2021

EMNLP 2021 is planned to be a hybrid event in Punta Cana, Dominican Republic, with both on-site and fully virtual participation possible.

panel

The International Conference on Computational Processing of Portuguese (PROPOR) is the main event in the area of human language processing that is focused on theoretical and technological issues of written and spoken Portuguese and Galician. The meeting has been a very rich forum for the exchange of ideas and partnerships for the research and industry communities dedicated to the automated processing of Portuguese, promoting the development of methodologies, resources and projects that can be shared among researchers and practitioners in the field.

Find more information and pay the registration fee on [event organizer’s website](https://sites.universidadedefortaleza.com/propor2022/?page_id=401)

The International Conference on Computational Processing of Portuguese (PROPOR) is the main event in the area of human language processing that is focused on theoretical and technological issues of written and spoken Portuguese and Galician.

Our contribution presents the first ever stance and sentiment annotated corpus in Portuguese. Using data from Twitter, we annotated a data set with over 15,000 tweets. In building the corpus, we made both random and supervised searches to maximize balance. Using four annotators, we have classified each tweet on four categories: support, against, neutral or inconclusive. Furthermore, we have annotated each tweet for sentiment: positive, negative, neutral, sarcasm or inconclusive. Our annotators yield strong inter-annotator agreement. In addition, we test a baseline model using a pre-trained BERT model for Portuguese. Results suggest that our data set is in line with its counterparts in other languages.

SS-PT: A Stance and Sentiment data set from Portuguese quoted tweets

This paper presents a targeted syntactic evaluation of Transformer models for Galician-Portuguese. We defined three experiments that allow to explore how these models, trained with a masked language modeling objective, encode syntactic knowledge. To do so, we created a new dataset including test instances of number (subject-verb), gender (subject-predicative adjective), and person (subject-inflected infinitive) agreement. This dataset was used to evaluate monolingual and multilingual BERT models, controlling for various aspects such as the presence of attractors or the distance between the dependent elements. The results show that Transformer models perform competently in many cases, but they are generally confounded by the presence of attractors in long-distance dependencies. Moreover, the different behavior of monolingual models trained with the same corpora reinforces the need for a deep exploration of the network architectures and their learning process.

A Targeted Assessment of the Syntactic Abilities of Transformer Models for Galician-Portuguese

This paper describes the development of the first test suite for the language direction Portuguese-English. Designed for fine-grained linguistic analysis, the test suite comprises 330 test sentences for 66 linguistic phenomena and 14 linguistic categories. Eight different MT systems were compared using quantitative and qualitative methods via the test suite: DeepL, Google Sheets, Google Translator, Microsoft Translator, Reverso, Systran, Yandex and an internally built NMT system
trained over 30 hours on 2,5M sentences. It was found that ambiguity, named entity & terminology and verb valency are the categories where MT systems struggle most. Negation, pronouns, subordination, verb tense/aspect/mood and false friends are the categories where MT systems perform best.

A test-suite for the evaluation of Portuguese-English machine translations

This work focuses on the enrichment of existing Portuguese word embeddings with visual information. The combined text-image embeddings were tested against their text-only counterparts in common NLP tasks. The new embeddings were tested in two different domains - general news and a geosciences. The results show an increase in performance for several tasks, which indicates that visual information fusion for word embeddings can be useful for certain tasks.

Enriching Portuguese Word Embeddings with Visual Information

Abstract Meaning Representation (AMR) is a semantic formalism that has been widely adopted in the area of semantic parsing. We present in this paper our contribution to the task for Portuguese. We investigated semantic parsing methods of different paradigms, producing state-of-the-art results for this language. We also introduced the first AMR-annotated corpus for Portuguese, a novel and better semantic parsing evaluation measure, and a new AMR-based alignment method.

Abstract Meaning Representation Parsing for the Brazilian Portuguese Language

Languages evolve internally and externally in relation to themselves and to others. It is for this reason that the automatic calculation of the historical and current distance between languages from corpora is an interesting challenge. In this thesis we have developed a methodology for calculating this distance from perplexity, a metric that has also proved to be suitable for automatic identification between languages. We have also tested the role that orthography plays in the distance between languages. Finally, we have created an adhoc corpus called Carvalho in orthographies as close as possible to the original ones, where we have carried out our experiments.

Corpus-based measures of distance between languages and variants. Application to the historical linguistics of Galician, Portuguese, Spanish and English

In the legal domain, documents have peculiarities related to their length, specialized use, and vocabulary. Recent neural network-based approaches can generate high-quality summaries. However, these approaches have been used mostly for creating concise abstracts for news articles in English and suffer from the "hallucination", i.e. the inclusion of extraneous material that is not supported by the original text. 
In this work, we propose LegalSumm, a method for summarizing legal texts in Portuguese. Our solution relies on Transformers and textual entailment is used to reduce hallucinations. To cope with long documents, we propose different chunking strategies. We compared LegalSumm to strong extractive and abstractive summarization baselines and collected the opinion of legal experts on the automatically generated summaries. The results show that LegalSumm outperforms existing summarization methods.

Summarizing Legal Rulings

In the hospital environment, the incidence of adverse events (unforeseen incidents that cause harm to patients) is the primary concern of risk management teams. The use of machine learning techniques may help healthcare professional to identify and mitigate adverse events. This thesis develops two experiments to evaluate machine learning approaches to identify two major adverse events in electronic health records. In this extended abstract we cover the algorithm created to identify fall events in clinical notes, as they are closely related to the processing of portuguese language. We annotated 1,402 clinical sentences with fall events to train a Token Classifier (TkC) to detect words within the context of falls. The TkC was able to correctly identify 85% of the sentences with fall events.

Applying Machine Learning to Electronic Health Records: a Study on Two Adverse Events

Tackling Multilingual Media Monitoring with Natural Language Processing

From (Our) Past to (Our) Future with Natural Language Processing (NLP): Linguistic and Terminological Approaches to Digital Historical Epidemiology in Portuguese (Maria José Finatto)

Brasileiras em PLN panel: Computer science, linguistics and more: NLP and transdisciplinarity

Downloads

Next from PROPOR 2022

SS-PT: A Stance and Sentiment data set from Portuguese quoted tweets

Similar lecture

Analyzing the Surprising Variability in Word Embedding Stability Across Languages

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES