by Cesar Teixeira
by Ruhan Bidart, Adriano C. M. Pereira and Jussara Almeida
Currently, it has been observed an increasing popularization of e-commerce. Thus, in areas such as tourism, online retailers are dealing with different challenges such as performance, scalability and personalization. In this context, it is fundamental to understand the characteristics of the user requests, and the access patterns of users on these systems. In this work we analyze the workload of a tourism Web system, which has on average six thousand unique daily access. We present a characterization of users sessions, their requests, and their navigation profile. As results we identify the distinct user profiles and we understand the navigation and access patterns of these users and we define a methodology of profiles and workload characterization that can be used in any other type of system, not only in tourism applications.
by Anna Carolina Medeiros and Tatiana Tavares
The use of our body language to communicate with computer systems is an increasingly possible and applicable feature in the real world. This fact is potentiated by the evolution of gesture recognition based commercial solutions. A gesture interface complements or replaces navigation in a conventional interface, it is up to each developer to choose the most appropriate option for their application. Therefore when opting for gesture usage, the gestures will be responsible to activate the systems functions. This work presents an interface development process centered on the selection of gestures. The proposed process should help interface designers to incorporate gesture based natural interaction into their applications in a more systematic way.
by Maria Da Graca Pimentel
by Paulo Artur de Sousa Duarte, Felipe Mota Barreto, Francisco Anderson de Almada Gomes, Fernando Trinta and Windson Viana de Carvalho
With the advances in mobile devices and ubiquitous computing, mobile and context-aware applications is becoming increasingly emerging. However, the development process of these applications is still facing some challenges (e.g, heterogeneity of devices, complexity of the sensors access code, etc). The adoption of middleware platforms for context-aware systems is a well-known solution used to overcome such problems. An example of middleware developed for this purpose is LoCCAM, which allows adaptive acquisition of contextual information on Android based devices. But, LoCCAM still presents issues concerning middleware configuration of contextual information. One approach, to mitigate these existing problems is the usage of to combine them with the MDE (Model-Driven Engineering) paradigm. This paper shows a Visual Domain-Specific Language for modeling contextual information that will be used in the development of applications using the LoCCAM. This language aims at generating skeletons of Android projects. This project is properly configured for the use of middleware, including a transparent access to the contextual information. Thus, use of this language allows the developer a greater abstraction in the access to middleware
by José Dias
by Alexsandro Bastos, Iwens Sene Júnior and Renato De Freitas Bulcão-Neto
This paper contributes with a knowledge representation model of the human vital sign monitoring activity based on Semantic Web specifications for ontologies and Horn-like rules. The proposed model is founded in interviews with intensive care units professionals, medical and nursing literature, and the reuse of existing ontologies. As a result, this work contributes with an ontology describing vital signs, actors and temporal information involved in monitoring, a set of requirements underlying to that model, and dozens of Horn-like rules describing alarms and the respective abnormal values for each vital sign.
by Thiago Prota, Douglas Véras and Carlos André Guimarães Ferraz
This work aims to present a thesis proposal in the context of Ginga-NCL, proposes a solution that allows analyze the interactive TV applications in this context from two perspectives: structural and behavioral. Thus, seeks to present the theoretical context, emphasizing the problem identification and the proposed objectives. Moreover, presents research planning and development, detailing the current stage of work, the proposed solution and a proposed evaluation. Finally, details the state of the art, assessing the related works.
by Luiz Fernando Gomes Soares
by Guilherme de Souza Alves and Carlos Eduardo C. F. Batista
by valdecir becker, Bruno Pacine and Guido Souza Filho
by Arthur Fortes Da Costa and Marcelo Garcia Manzato
In this paper, we propose a framework that uses multimodal interactions of users to generate a more accurate list of recommendations optimized for the user. Our approach is a response to the actual scenario on the Web which allows users to interact with the content in different ways, and thus, more information about his preferences can be obtained to improve recommendation. The proposal consists of an ensemble learning technique that combines rankings generated by unimodal recommenders based on particular interaction types. By using a combination of different types of feedback from users, we are able to provide better recommendations, as shown by our experimental evaluation.
by Virginia P. Campos, Tiago Maritan and Guido Lemos de Souza Filho
Audio description (AD) is a fundamental accessibility feature for people who are blind or have low vision. This paper presents a system for automatic generation of audio description, called CineAD. The system detects gaps in speech and generates these descriptions from the analysis of the original script and subtitles. To evaluate the solution, we automatically generate the audio description script of two movies. The preliminary results indicate that the solution is able to generate descriptions of the most important events of the film and the size of the descriptions is a major challenge related to the automatic creation of AD scripts.
by Carlos Pernisa Jr
by Marcelo F. Moreno and Eduardo Barrére
by Bruna Neuenschwander, Adriano C. M. Pereira, Wagner Meira Jr. and Denilson Barbosa
With the rise of Web 2.0 applications, most people started consuming information and sharing opinions and ideas about most aspects of their lives on a variety of social media platforms, creating massive and continuous streams of valuable data. While this opened the door for information extraction and mining techniques that can help us understand different aspects of society, extracting useful information from such streams of Web data is far from trivial. In this setting, sentiment analysis techniques can be convenient as they are capable of summarizing general feeling about entities people care about, such as products and companies. Therefore, they can be quite applicable in scenarios like the stock market, which also has tremendous impact on society. This paper describes and evaluates two different techniques for sentiment analysis applied to the Brazilian stock market data: lexicon-based and machine learning based, considering a wide range of text pre-processing and feature selection approaches.
by Carleandro Noleto, Windson Viana de Carvalho and Fernando Trinta
Dispositivos móveis possuem sensores integrados capazes de capturar informações contextuais, como a localização e o perfil de seus usuários, e repassar tais informações a servidores remotos. Sistemas ditos sensíveis ao contexto exploram essas informações contextuais para adaptar suas interfaces e conteúdos, e/ou melhorar o seu desempenho. Os jogos pervasivos são exemplos de sistemas sensíveis ao contexto que utilizam informações capturadas pelos sensores dos dispositivos, como elemento dentro da narrativa do jogo, onde desafios e objetos virtuais podem ser colocados, de acordo com o contexto do jogador, por exemplo, sua posição em uma determinada área do mundo real. Outra evolução tecnológica que permite trazer os jogos virtuais para o espaço físico do usuário é a realidade aumentada (RA). O principal objetivo deste trabalho é propor uma ferramenta de autoria para tais jogos pervasivos, na qual usuários sem conhecimentos de programação possam montar seus jogos, de acordo com mecânicas, personagens e elementos pré-estabelecidos na ferramenta, que por sua vez, é responsável por gerar uma versão implantável do jogo para sua plataforma Android.
by Valter Roesler
by Josivan Silva, Geiza Caruline Costa, João Melo, Iann Vicari Mascarenhas and Celso S. Kurashima
A recent kind of Web application incorporates three-dimensional
image content, which provides interactivity and new visual
experiences to the users. However, rendering 3D video content
has been a research challenge due to modeling complexity and
visual content definition. This paper proposes an interactive Web
application that shows 3D modeled human faces that were
resources for 3D reconstruction and interactivity. The 3D
modeling process is realized with a triangulation algorithm called
Convex Hull. We propose a method that divides the model into
four parts in order to achieve improvements on visual details
around the concave areas. The final model mesh is saved into an
XML file, which is readable by any Web browser. The
implemented viewer resulted in image visual enhancements,
simple synchronized integration of video and audio content, and
friendly interactivity capabilities.
by Paulo Roberto de Lima Lopes, Thiago D. Lima Verde Brito and Luiz Ary Messina
by Ernesto Veiga, Guilherme Maranhão and Renato De Freitas Bulcão-Neto
Este artigo propõe a infraestrutura Hermes para apoio ao desenvolvimento de aplicações sensíveis a contexto. Hermes fornece componentes que manipulam a semântica de informações de contexto e o gerenciamento de eventos por meio de um middleware de comunicação distribuída em tempo real. Com base na arquitetura de referência Context Toolkit, Hermes visa o desenvolvimento rápido de aplicações, nas quais a semântica das informações é mantida de forma desacoplada e as alterações nessas informações são notificadas em tempo real, requisitos esses demandas na literatura de Computação Sensível a Contexto. Um estudo de caso em monitoramento de sinais vitais humanos tem sido desenvolvido para fins de validação da arquitetura da infraestrutura Hermes.
by Jane de Almeida, Denise Stringhini, Cicero Inacio da Silva, Maria Amélia Eliseo and Fernanda Maria Oliveira Araújo
by Luiz Eduardo Cunha Leite and Luiz Marcos Garcia Gonçalves
by Luiz Fernando Gomes Soares
In this talk, inter-thing synchronization is considered as a QoS problem in its phases of initialization, negotiation, provisioning, and managing. Intermedia synchronization is taken as example. For the several QOS phases, we will discuss topics such as: infrastructure performance parameters, inter-source timeline synchronization, hypermedia specification languages, authoring tools, intermedia-service admission control, compile-time adaptations, intelligent prefetching, resource reservation in advance, pushed-data carousel management, runtime adaptations, elastic time computation, inter-destination synchronization, ambient intelligence, etc. In all these topics, the state of the art is presented as well as open research issues. The talk will be given in Portuguese.
by Zé Dias
José Dias diretor de pesquisa e desenvolvimento da Rede Globo apresentaré as contribuições que a divisão de P&D da Globo vem desenvolvendo para gerar Surpresas Estratégicas para o espectador.
18th–21st November 2014