Percorrer por autor "Silva, Catarina"
A mostrar 1 - 10 de 45
Resultados por página
Opções de ordenação
- Active Manifold Learning with Twitter Big DataPublication . Silva, Catarina; Antunes, Mário; Costa, Joana; Ribeiro, BernardeteThe data produced by Internet applications have increased substantially. Big data is a flaring field that deals with this deluge of data by using storage techniques, dedicated infrastructures and development frameworks for the parallelization of defined tasks and its consequent reduction. These solutions however fall short in online and highly data demanding scenarios, since users expect swift feedback. Reduction techniques are efficiently used in big data online applications to improve classification problems. Reduction in big data usually falls in one of two main methods: (i) reduce the dimensionality by pruning or reformulating the feature set; (ii) reduce the sample size by choosing the most relevant examples. Both approaches have benefits, not only of time consumed to build a model, but eventually also performance-wise, usually by reducing overfitting and improving generalization capabilities. In this paper we investigate reduction techniques that tackle both dimensionality and size of big data. We propose a framework that combines a manifold learning approach to reduce dimensionality and an active learning SVM-based strategy to reduce the size of labeled sample. Results on Twitter data show the potential of the proposed active manifold learning approach.
- Adaptive learning for dynamic environments: A comparative approachPublication . Costa, Joana; Silva, Catarina; Antunes, Mário; Ribeiro, BernardeteNowadays most learning problems demand adaptive solutions. Current challenges include temporal data streams, drift and non-stationary scenarios, often with text data, whether in social networks or in business systems. Various efforts have been pursued in machine learning settings to learn in such environments, specially because of their non-trivial nature, since changes occur between the distribution data used to define the model and the current environment. In this work we present the Drift Adaptive Retain Knowledge (DARK) framework to tackle adaptive learning in dynamic environments based on recent and retained knowledge. DARK handles an ensemble of multiple Support Vector Machine (SVM) models that are dynamically weighted and have distinct training window sizes. A comparative study with benchmark solutions in the field, namely the Learn++.NSE algorithm, is also presented. Experimental results revealed that DARK outperforms Learn++.NSE with two different base classifiers, an SVM and a Classification and Regression Tree (CART).
- Assessing the liquidity in Portuguese hotel companiesPublication . Silva, Catarina; Lima Santos, Luís; Gomes, Conceição; Malheiros, CátiaThe hospitality companies have had substantial growth in the tourism sector which gives them a large part of the revenue generated by the sector. In this regard, its impact, whether negative or positive, is quite high and generates a response to a need felt by agents of the environment in which it operates. As a short-term sustainability indicator, the liquidity level of a company demonstrates its ability to repay its obligations, being a great management support for decision making and anticipation of financial problems that may arise. Considering the volatility of hotel companies, greater importance is given to the study of liquidity. The main liquidity ratios of Portuguese hotels in the 2010-2017 period will be analysed; data was collected on July 4, 2019, on the SABI platform and the original sample is composed of 2161 hotel companies registered with two Portuguese economic activity codes (CAE), “55111 - Hotels with restaurant” and “55121 - Hotels without restaurant”. The assessment of liquidity level will be important to decision makers understand if there are differences between hotels with or without restaurant and among the Portuguese districts were hotels are located. The results of this study are expected to be of assistance to hotel managers as decisions taken within the organization can be more deliberate and informed.
- Assistive Mobile Applications for DyslexiaPublication . Madeira, Jorge; Silva, Catarina; Ferreira, Paula Cristina; Marcelino, LuísThe ability to read is one of the main skills of a human being. However, some of us have reading difficulties, regardless of social status, level of intelligence or education. This disorder is the main characteristic of dyslexia and is maintained throughout life,requiring early and specialized intervention. Dyslexia is defined as a learning disturbance in the area of reading, writing and spelling. Although the numbers of prevalence rely heavily on the type of investigation conducted, several studies indicate that up to 17% of the world population is dyslexic, and that men have greater prevalence. In this work we will address the use of assistive mobile applications for dyslexia by analyzing possible solutions and proposing a prototype of a mobile application that can be used by dyslexic and whilst giving feedback both to the dyslexic him/herself and to the assisting technician or teacher. The implemented prototype focuses the Portuguese language and was tested with Portuguese students with ages between 10 and 12 years old. Preliminary results show that the proposed gamified set of activities, allow dyslexics to improve multisensory perception, constituting an added value facilitator of adaptiveness and learning.
- Assistive mobile software for public transportationPublication . Silva, João De Sousa E; Silva, Catarina; Marcelino, Luís; Ferreira, Rui; Pereira, AntónioThe need of mobility on public transport for persons with visual impairment is mandatory. While traveling on a public transport, the simple ability to know the current location is almost impossible for such persons. To overcome this hurdle, we developed an assistive application that can alert its user to the proximity of all public transportation stops, giving emphasis to the chosen final stop. The application is adjustable to any transportation system and is particularly relevant to use in public transports that do not have any audio system available. The developed prototype runs on an Android OS device equipped with Global Positioning System (GPS). To ensure the highest possible level of reliability and to make it predictable to users, the application's architecture is free of as much dependencies as possible. Therefore, only GPS, or other localization mechanism, is required. The interface was designed to be suitable not only for talkback (Android's inbuilt screen-reader) aimed at blind users, but also for people with low vision that can still use their sight to check the screen. Thus, it was meant to be graphically simple and unobtrusive. It was tested by visual impaired persons leading to the conclusion that it demonstrates an existing need, and opens a new perspective in public transportation's accessibility.
- Boosting dynamic ensemble’s performance in TwitterPublication . Costa, Joana; Silva, Catarina; Antunes, Mário; Ribeiro, BernardeteMany text classification problems in social networks, and other contexts, are also dynamic problems, where concepts drift through time, and meaningful labels are dynamic. In Twitter-based applications in particular, ensembles are often applied to problems that fit this description, for example sentiment analysis or adapting to drifting circumstances. While it can be straightforward to request different classifiers' input on such ensembles, our goal is to boost dynamic ensembles by combining performance metrics as efficiently as possible. We present a twofold performance-based framework to classify incoming tweets based on recent tweets. On the one hand, individual ensemble classifiers' performance is paramount in defining their contribution to the ensemble. On the other hand, examples are actively selected based on their ability to effectively contribute to the performance in classifying drifting concepts. The main step of the algorithm uses different performance metrics to determine both each classifier strength in the ensemble and each example importance, and hence lifetime, in the learning process. We demonstrate, on a drifted benchmark dataset, that our framework drives the classification performance considerably up for it to make a difference in a variety of applications.
- Choice of Best Samples for Building Ensembles in Dynamic EnvironmentsPublication . Costa, Joana; Silva, Catarina; Antunes, Mário; Ribeiro, BernardeteMachine learning approaches often focus on optimizing the algorithm rather than assuring that the source data is as rich as possible. However, when it is possible to enhance the input examples to construct models, one should consider it thoroughly. In this work, we propose a technique to define the best set of training examples using dynamic ensembles in text classification scenarios. In dynamic environments, where new data is constantly appearing, old data is usually disregarded, but sometimes some of those disregarded examples may carry substantial information. We propose a method that determines the most relevant examples by analysing their behaviour when defining separating planes or thresholds between classes. Those examples, deemed better than others, are kept for a longer time-window than the rest. Results on a Twitter scenario show that keeping those examples enhances the final classification performance.
- Citizens@City Mobile Application for Urban Problem ReportingPublication . Ribeiro, António Miguel; Costa, Rui Pedro; Marcelino, Luís; Silva, CatarinaUrban problems, such as holes in the pavement, poor accesses to wheelchairs or lack of public lighting, are becoming pervasive. Despite the fact that most of these problems directly affect life quality and sometimes even safety, not everyone has the readiness or initiative to report them to the proper authorities. This fact makes these “black spots” difficult to identify and the repairing process slow. Citizens@City is an Android mobile application that allows the general population to play a more active role in the identification of these problems by reporting them to the proper authorities in a simple and fast way. Moreover, citizens will have the possibility to follow the identification and repairing processes, and know at a given moment its status (e.g. identified, repairing scheduled, solved). Additionally, it will also allow the proper authorities to identify and manage the reported problems, from their identification until they are solved.
- CrowdTargeting: Making Crowds More PersonalPublication . Costa, Joana; Silva, Catarina; Ribeiro, Bernardete; Antunes, MárioCrowdsourcing is a bubbling research topic that has the potential to be applied in numerous online and social scenarios. It consists on obtaining services or information by soliciting contributions from a large group of people. However, the question of defining the appropriate scope of a crowd to tackle each scenario is still open. In this work we compare two approaches to define the scope of a crowd in a classification problem, casted as a recommendation system. We propose a similarity measure to determine the closeness of a specific user to each crowd contributor and hence to define the appropriate crowd scope. We compare different levels of customization using crowd-based information, allowing non-experts classification by crowds to be tuned to substitute the user profile definition. Results on a real recommendation data set show the potential of making crowds more personal, i.e. of tuning the crowd to the crowdtarget.
- Customized crowds and active learning to improve classificationPublication . Costa, Joana; Silva, Catarina; Antunes, Mário; Ribeiro, BernardeteTraditional classification algorithms can be limited in their performance when a specific user is targeted. User preferences, e.g. in recommendation systems, constitute a challenge for learning algorithms. Additionally, in recent years user’s interaction through crowdsourcing has drawn significant interest, although its use in learning settings is still underused. In this work we focus on an active strategy that uses crowd-based non-expert information to appropriately tackle the problem of capturing the drift between user preferences in a recommendation system. The proposed method combines two main ideas: to apply active strategies for adaptation to each user; to implement crowdsourcing to avoid excessive user feedback. A similitude technique is put forward to optimize the choice of the more appropriate similitude-wise crowd, under the guidance of basic user feedback. The proposed active learning framework allows non-experts classification performed by crowds to be used to define the user profile, mitigating the labeling effort normally requested to the user. The framework is designed to be generic and suitable to be applied to different scenarios, whilst customizable for each specific user. A case study on humor classification scenario is used to demonstrate experimentally that the approach can improve baseline active results.