Regionalized models for Spanish language variations based on Twitter.

Spanish is one of the most spoken languages in the world. Its proliferation comes with variations in written and spoken communication among different regions. Understanding language variations can help improve model performances on regional tasks, such as those involving figurative language and local context information. This manuscript presents and describes a set of regionalized resources for the Spanish language built on 4-year Twitter public messages geotagged in 26 Spanish-speaking countries. We introduce word embeddings based on FastText, language models based on BERT, and per-region sample corpora. We also provide a broad comparison among regions covering lexical and semantical similarities and examples of using regional resources on message classification tasks.

Contesting views on mobility restrictions in urban green spaces amid COVID-19-Insights from Twitter in Latin America and Spain.

Positive sentiments towards urban green spaces (UGS) unequivocally increased worldwide amid COVID-19. In contrast, this paper documents that views on mobility restrictions applicable to UGS are of a contested nature. That is, while residents unambiguously report positive sentiments towards UGS, they do not share views on how to administer access to UGS-which is a matter of public policy. These contesting views reflect opposite demands that managers of UGS had to balance during the pandemic as they faced the challenge of reducing risk of spread while providing services that support physical and mental health of residents. The empirical analysis in this paper relies on views inferred through a text classification algorithm implemented on Twitter messages posted from January to October 2020, by urban residents in three Latin American countries-Argentina, Colombia, and Mexico-and Spain. The focus on Latin America is motivated by the documented lack of compliance with mobility restrictions; Spain works as a comparison point to learn differences with respect to other regions. Understanding and following in real-time the evolution of contesting views amid a pandemic is useful for managers and city planners to inform adaptation measures-e.g. communication strategies can be tailored to residents with specific views.

Selection Heuristics on Semantic Genetic Programming for Classification Problems.

Individual semantics have been used for guiding the learning process of Genetic Programming. Novel genetic operators and different ways of performing parent selection have been proposed with the use of semantics. The latter is the focus of this contribution by proposing three heuristics for parent selection that measure the similarity among individuals' semantics for choosing parents that enhance the addition, Naive Bayes, and Nearest Centroid. To the best of our knowledge, this is the first time that functions' properties are used for guiding the learning process. As the heuristics were created based on the properties of these functions, we apply them only when they are used to create offspring. The similarity functions considered are the cosine similarity, Pearson's correlation, and agreement. We analyze these heuristics' performance against random selection, state-of-the-art selection schemes, and 18 classifiers, including auto-machine-learning techniques, on 30 classification problems with a variable number of samples, variables, and classes. The result indicated that the combination of parent selection based on agreement and random selection to replace an individual in the population produces statistically better results than the classical selection and state-of-the-art schemes, and it is competitive with state-of-the-art classifiers. Finally, the code is released as open-source software.

Liking Product Landscape: Going Deeper into Understanding Consumers' Hedonic Evaluations.

The use of graphical mapping for understanding the comparison of products based on consumers' perceptions is beneficial and easy to interpret. Internal preference mapping (IPM) and landscape segmentation analysis (LSA) have successfully been used for this propose. However, including all the consumers' evaluations in one map, with products' overall liking and attributes' perceptions, is complicated; because data is in a high dimensional space some information can be lost. To provide as much information as possible, we propose the liking product landscape (LPL) methodology where several maps are used for representing the consumers' distribution and evaluations. LPL shows the consumers' distribution, like LSA, and also it superimposes the consumers' evaluations. However, instead of superimposing the average overall liking in one map, this methodology uses different maps for each consumer's evaluation. Two experiments were performed where LPL was used for understanding the consumers' perceptions and compared with classic methodologies, IPM and cluster analysis, in order to validate the results. LPL can be successfully used for identifying consumers' segments, consumers' preferences, recognizing perception of product attributes by consumers' segments and identifying the attributes that need to be optimized.

Models of performance of evolutionary program induction algorithms based on indicators of problem difficulty.

Modeling the behavior of algorithms is the realm of evolutionary algorithm theory. From a practitioner's point of view, theory must provide some guidelines regarding which algorithm/parameters to use in order to solve a particular problem. Unfortunately, most theoretical models of evolutionary algorithms are difficult to apply to realistic situations. However, in recent work (Graff and Poli, 2008, 2010), where we developed a method to practically estimate the performance of evolutionary program-induction algorithms (EPAs), we started addressing this issue. The method was quite general; however, it suffered from some limitations: it required the identification of a set of reference problems, it required hand picking a distance measure in each particular domain, and the resulting models were opaque, typically being linear combinations of 100 features or more. In this paper, we propose a significant improvement of this technique that overcomes the three limitations of our previous method. We achieve this through the use of a novel set of features for assessing problem difficulty for EPAs which are very general, essentially based on the notion of finite difference. To show the capabilities or our technique and to compare it with our previous performance models, we create models for the same two important classes of problems-symbolic regression on rational functions and Boolean function induction-used in our previous work. We model a variety of EPAs. The comparison showed that for the majority of the algorithms and problem classes, the new method produced much simpler and more accurate models than before. To further illustrate the practicality of the technique and its generality (beyond EPAs), we have also used it to predict the performance of both autoregressive models and EPAs on the problem of wind speed forecasting, obtaining simpler and more accurate models that outperform in all cases our previous performance models.