Projects
- Strategies for Planning and Performance Evaluation of IoT infrastructure for smart cities: case study in municipalities in the Brazilian Amazon.
This project proposes strategies for the provision of smart city services based on Internet of Things (IoT), considering the characteristics, limitations and peculiarities of Brazilian cities, notably municipalities in the Amazon. The implementation of a smart city needs a big apparatus in the most different areas of Information Technology. In particular, IoT currently plays an important role in promoting various possibilities to improve the functioning of cities, allowing / obtaining, for example, pattern recognition, location, tracking, monitoring and intelligent management. Among the various possible axes, three that are considered crucial for cities located in the Brazilian Amazon will be investigated: (a) monitoring of water resources, (b) provision of temporary telecommunications infrastructure, and (c) smart grid techniques for efficient management of electricity in the city. The IoT strategies applied to smart cities favor extracting knowledge and helping decision-making by city managers in general. In the end, the proposed strategies will assist in the evaluation and implementation of a smart city, taking into account the specificities of Brazilian cities.
coordinator: Carlos Renato Francês
- Research and development of computational intelligence methods for smarter knowledge discovery in integrated Social CRM systems.
Social Customer Relationship Management (SCRM) is an emerging concept that includes strategies, processes and technologies that bring together social networks on the web and CRM processes. However, transforming the large mass of data available on social networks into value adding opportunities for companies is challenging. Today, a variety of software applications based on web and text mining techniques is used for knowledge discovery in SCRM. These techniques are helpful for identifying relevant social media postings and for extracting basic information (e.g. number of likes, occurrence of key words in postings, identification of simple sentiments), but they are insufficient for identifying more complex patterns which include interconnections between actors, profiles and postings from large and unstructured databases. Computational intelligence (CI) techniques (e.g. artificial neural networks, Bayesian models, fuzzy systems and evolutionary computing) are an alternative option to address these problems. They promise a great potential to improve the capabilities in knowledge discovery and may also enable new usage scenarios in SCRM (e.g. impact simulation, network analysis, topic development). However, research on the extent to which methods can be applied in integrated SCRM systems is scarce. UFPA and IWI plan to address this gap by combining complementary research from their past and current projects in the respective fields of CI (UFPA) and SCRM (IWI). This application is thus for conducting a joint workshop that aims at identifying relevant common research topics and activities for a long-term cooperation including joint research projects, exchange of researchers and creating a laboratory.
coordinator: Reiner Alt
- Application of data mining models to study the impact of variations in income from work, social security and social programs on inequality in income distribution, poverty and extreme poverty for agglomerations.
Brazilian society is marked by several types of inequalities (social, economic, cultural, among others) that result from the poor distribution of income and the lack of investment in the social area. These inequalities have different impacts, especially on conditions of poverty and precariousness. The mechanisms for reducing inequality and eliminating poverty are not yet unanimous among the competent bodies. It is also controversial the applicability for all regions of Brazil of the same strategies, since variations in income affect differently income inequality, poverty and extreme poverty in the different Brazilian municipalities. In this sense, it is extremely important to formulate new development models and to implement public policies that enable a more equitable distribution of goods and social resources. It is important to implement actions that consider the existence of different “Brazils”, with different levels of application and scope, capable of generating permanent and effective solutions in the reduction of acute poverty scenarios, resulting in a society with greater social and economic inclusion. This paper aims to assess and identify the impact of variations in income from work, social security and social programs on inequality in income distribution, poverty and extreme poverty for homogeneous clusters of Brazilian municipalities in order to improve the formulation of development models and the definition of public policies according to the social reality of each cluster, making the application of more adequate resources more optimized. For this, it is proposed the application of data mining to discover relationships and models present in official databases, such as social security, ENEM and IBGE, for example.
coordinator: Carlos Renato Francês