Meaning Representation (AMR) is a meaning representation (MR) designed to abstract away from syntax, allowing syntactically different sentences to share the same AMR graph.Unlike other MRs, existing AMR corpora typically link one AMR graph to a single reference.This paper investigates the value of paraphrase generation in low-resource AMR-to-Text generation by testing various paraphrase generation strategies and evaluating their impact.The findings show that paraphrase generation significantly outperforms the baseline and traditional data augmentation methods, even with fewer training instances.Human evaluations indicate that this strategy often produces syntactic-based paraphrases and can exceed the performance of previous approaches.Additionally, the paper releases a paraphrase-extended version of the AMR corpus.

Developing new applied Artificial Intelligence solutions can be complex and time-consuming. Automated Machine Learning (AutoML) can speed up the solutions' development, but may require hundreds of datasets to make the learning process viable, which may be a challenge for Natural Language Processing (NLP) tasks, as many of them need annotated datasets that are expensive to produce. It may also be necessary to artificially recombine classes to generate new datasets for Meta-Learning (MtL), which can be considered as a particular AutoML approach. This paper addresses such challenge and proposes an AutoML benchmark for sentiment analysis, which comprises relevant tasks in NLP, resulting in 49 preprocessed datasets, more than 650 binary datasets, and 204 meta-features. More than this, a proof of concept is carried out for the hate speech detection task.

This paper presents PortiLexicon-UD, a large and freely available lexicon for Portuguese delivering morphosyntactic information according to the Universal Dependencies model. This lexical resource includes part of speech tags, lemmas, and morphological information for words, with 1,221,218 entries (considering word duplication due to different combination of PoS tag, lemma, and morphological features). We report the lexicon creation process, its computational data structure, and its evaluation over an annotated corpus, showing that it has a high language coverage and good quality data.

One of the challenges of aspect-based sentiment analysis is the implicit mention of aspects. These are more difficult to identify and may require world knowledge to do so. In this work, we evaluate frequency-based, hybrid, and machine learning methods, including the use of the pre-trained BERT language model, in the task of extracting aspect terms in opinionated texts in Portuguese, emphasizing the analysis of implicit aspects. Besides the comparative evaluation of methods, the differential of this work lies in the analysis’s novelty using a typology of implicit aspects that shows the knowledge needed to identify each implicit aspect term, thus allowing a mapping of the strengths and weaknesses of each method.

Due to the severity of the social media offensive and hateful comments in Brazil, and the lack of research in Portuguese, this paper provides the first large-scale expert annotated corpus of Brazilian Instagram comments for hate speech and offensive language detection. The HateBR corpus was collected from the comment section of Brazilian politicians’ accounts on Instagram and manually annotated by specialists, reaching a high inter-annotator agreement. The corpus consists of 7,000 documents annotated according to three different layers: a binary classification (offensive versus non-offensive comments), offensiveness-level classification (highly, moderately, and slightly offensive), and nine hate speech groups (xenophobia, racism, homophobia, sexism, religious intolerance, partyism, apology for the dictatorship, antisemitism, and fatphobia). We also implemented baseline experiments for offensive language and hate speech detection and compared them with a literature baseline. Results show that the baseline experiments on our corpus outperform the current state-of-the-art for the Portuguese language.

Most information is passed on in the form of language. Therefore, research on how people use language to inform and misinform, and how this knowledge may be automatically extracted from large amounts of text is surely relevant. This survey provides first-hand experiences and a comprehensive review of rhetorical-level structure analysis for online deception detection. We systematically analyze how discourse structure, aligned or not with other approaches, is applied to automatic fake news and fake reviews detection on the web and social media. Moreover, we categorize discourse-tagged corpora along with results, hence offering a summary and accessible introductions to new researchers.

Meaning Representation (AMR) is a graph-based semantic formalism where the nodes are concepts and edges are relations among them.Most of AMR parsing methods require alignment between the nodes of the graph and the words of the sentence.However, this alignment is not provided by manual annotations and available automatic aligners focus only on the English language, not performing well for other languages.Aiming to fulfill this gap, we developed an alignment method for the Portuguese language based on a more semantically matched word-concept pair.We performed both intrinsic and extrinsic evaluations and showed that our alignment approach outperforms the alignment strategies developed for English, improving AMR parsers, and achieving competitive results with a parser designed for the Portuguese language.

The proliferation of fake news is a current issue that influences a number of important areas of society, such as politics, economy and health. In the Natural Language Processing area, recent initiatives tried to detect fake news in different ways, ranging from language-based approaches to content-based verification. In such approaches, the choice of the features for the classification of fake and true news is one of the most important parts of the process. This paper presents a study on the impact of readability features to detect fake news for the Brazilian Portuguese language. The results show that such features are relevant to the task (achieving, alone, up to 92% classification accuracy) and may improve previous classification results.

In the last years, the popularity of smartphones and social networks has been contributing to the spread of fake news. Through these electronic media, this type of news can deceive thousands of people in a short time and cause great harm to individuals, companies, or society. Fake news has the potential to change a political scenario, to contribute to the spread of diseases, and even to cause deaths. Despite the efforts of several studies on fake news detection, most of them only cover English language news. There is a lack of labeled datasets of fake news in other languages and, moreover, important questions still remain open. For example, there is no consensus on what are the best classification strategies and sets of features to be used for automatic fake news detection. To answer this and other important open questions, we present a new public and real dataset of labeled true and fake news in Portuguese, and we perform a comprehensive analysis of machine learning methods for fake news detection. The experiments were performed using different sets of features and employing different types of classification methods. A careful analysis of the results provided sufficient evidence to respond appropriately to the open questions. The various evaluated scenarios and the drawn conclusions from the results shed light on the potentiality of the methods and on the challenges that fake news detection presents.

In the last years, the opinion summarization task has gained much importance because of the large amount of online information and the increasing interest in learning the user evaluation about products, services, companies, and people. Although there are many works in this area, there is room for improvement, as the results are far from ideal. In this paper, we present our investigations to generate extractive and abstractive summaries of opinions. We study some well-known methods in the area and compare them. Besides using these methods, we also develop new methods that consider the main advantages of the ones before. We evaluate them according to three traditional summarization evaluation measures: informativeness, linguistic quality, and utility of the summary. We show that we produce interesting results and that our methods outperform some methods from literature.

Multi-document summarization is a very important area of Natural Language Processing (NLP) nowadays because of the huge amount of data in the web.People want more and more information and this information must be coherently organized and summarized.The main focus of this paper is to deal with the coherence of multi-document summaries.Therefore, a model that uses discursive information to automatically evaluate local coherence in multi-document summaries has been developed.This model obtains 92.69% of accuracy in distinguishing coherent from incoherent summaries, outperforming the state of the art in the area.

In this paper we investigate the main linguistic phenomena that can make texts complex and how they could be simplified. We focus on a corpus analysis of simple account texts available on the web for Brazilian Portuguese and propose simplification strategies for this language. This study illustrates the need for text simplification to facilitate accessibility to information by poor literacy readers and potentially by people with other cognitive disabilities. It also highlights characteristics of simplification for Portuguese, which may differ from other languages. Such study consists of the first step towards building Brazilian Portuguese text simplification systems. One of the scenarios in which these systems could be used is that of reading electronic texts produced, e.g., by the Brazilian government or by relevant news agencies.

In this paper we present experiments on scientific text summarization. From a complete text, we produce a shorter version containing all the main parts of the research. Having in mind the sophisticated structure of such texts, we show that good results can be achieved using simple extractive summarizers with some obvious improvements that consider the specificity of the text genre. Specifically, we enhance the summarization process with the ability to detect and appropriately treat the text structure.