Search

Pirá is a recently developed reading comprehension dataset focused on the ocean, the Brazilian coast, and climate change. No detailed set of baselines has been built with this dataset yet, something that certainly hinders its use by researchers. In this paper, we define five benchmarks over the Pirá dataset, covering machine reading comprehension, information retrieval, open question answering, answer triggering, and multiple choice question answering. As part of this effort, we have produced a curated version of the original dataset, where we fixed a number of grammar issues, repetitions and other shortcomings. Furthermore, the dataset, now called Pirá 2.0, has been extended in several new directions, so as to face the aforementioned benchmark tasks: translation of supporting texts into Portuguese, classification labels on answerability, multiple choice candidates, and automatic paraphrases of questions and answers. The results described in this paper provide a reference point for researchers working with Pirá 2.0. Our results show that Pirá 2.0 is indeed a very challenging dataset, particularly useful for testing the ability of current machine learning models in acquiring expert scientific knowledge.

We describe a Physics-Informed Neural Network (PINN) that simulates the flow induced by the astronomical tide in a synthetic port channel, with dimensions based on the Santos - S\~ao Vicente - Bertioga Estuarine System. PINN models aim to combine the knowledge of physical systems and data-driven machine learning models. This is done by training a neural network to minimize the residuals of the governing equations in sample points. In this work, our flow is governed by the Navier-Stokes equations with some approximations. There are two main novelties in this paper. First, we design our model to assume that the flow is periodic in time, which is not feasible in conventional simulation methods. Second, we evaluate the benefit of resampling the function evaluation points during training, which has a near zero computational cost and has been verified to improve the final model, especially for small batch sizes. Finally, we discuss some limitations of the approximations used in the Navier-Stokes equations regarding the modeling of turbulence and how it interacts with PINNs., Comment: Published at the Workshop AI: Modeling Oceans and Climate Change (AIMOCC 2022), held in conjunction with the 31st International Joint Conference on Artificial Intelligence and the 25th European Conference on Artificial Intelligence (IJCAI-ECAI 2022)

Current research in natural language processing is highly dependent on carefully produced corpora. Most existing resources focus on English; some resources focus on languages such as Chinese and French; few resources deal with more than one language. This paper presents the Pir�� dataset, a large set of questions and answers about the ocean and the Brazilian coast both in Portuguese and English. Pir�� is, to the best of our knowledge, the first QA dataset with supporting texts in Portuguese, and, perhaps more importantly, the first bilingual QA dataset that includes this language. The Pir�� dataset consists of 2261 properly curated question/answer (QA) sets in both languages. The QA sets were manually created based on two corpora: abstracts related to the Brazilian coast and excerpts of United Nation reports about the ocean. The QA sets were validated in a peer-review process with the dataset contributors. We discuss some of the advantages as well as limitations of Pir��, as this new resource can support a set of tasks in NLP such as question-answering, information retrieval, and machine translation., https://github.com/C4AI/Pira

When considering sampling models described by a distribution from an exponential family, it is possible to create two types of imprecise probability models. One is based on the corresponding conjugate distribution and the other on the corresponding predictive distribution. In this paper, we show how these types of models can be constructed for any (regular, linear, canonical) exponential family, such as the centered normal distribution. To illustrate the possible use of such models, we take a look at credal classification. We show that they are very natural and potentially promising candidates for describing the attributes of a credal classifier, also in the case of continuous attributes.