Big Data and Visual Analytics

Big data and visual analytics can help organizations harness their data and use it to promote faster, better decision making and create new products and services. They imply examining large amounts of data to uncover hidden patterns, correlations and other insights to understand what happened in the past and to identify new opportunities. Besides the huge data that streams into most organizations’ businesses, new data streams (e.g. IoT) pose interesting challenges for applied research in this area, such as (i) the use of parallelization architectures for in-memory analytics (e.g. Hadoop/Spark) or GPGPU visualization frameworks (e.g. OpenCL/WebCL) and machine learning techniques supported by highly interactive visual interfaces, (ii) the discovery, interpretation and communication of meaningful data patterns, (iii) dealing with time-dependent data, even time-critical and, (iv) detection of data anomalies that, although corresponding to a tiny part of data, may be critical and costly.

Application Areas

Healthcare
Detection of Telomeres (nucleoprotein structures composed of protein, DNA and RNA molecules) from electron microscope images (in cooperation with Claus Azzalin research team at IMM). Use of graph-based approaches to store and visualize social networks in the context of healthcare systems (in cooperation with João Louro, CIO at Hospital Sta. Maria).

Intelligent transportation systems
Identification of drivers’ main fuel consumption parameters from big data generated from driving monitor process. Project performed in Lisbon City with Carris buses, savings of 2-3 liters per 100km were achieved through monitor process and the identification of main consumption parameters. Project in collaboration with the company Tecmic and Carris.

Smart Grids
Renewable production forecast based on historical data and weather conditions. Smart Electric Vehicle charging using OpenADR standard to turn on/off remotely the charging process, taking into account driving profile, energy prices based on renewable production excess and electrical distribution network limitations. Research in collaboration with Centro Algoritmi at Minho University.

Fisheries Control Activity
Fishing activity needs to be controlled to avoid massive fishing. Vessel Monitor System were implemented and fishing reporting systems were created. Huge data is generated and knowledge extraction and data manipulation is needed. Our work is based on the identification fishing patterns (GPS coordinates, time period, weight and fish type) and identification of fishing activity type based on boat speed and the change of direction. This research is being carried out in the context of the PT2020 project SeaITall in cooperation with the company Xsealance, Inov and IPMA at Portugal.

Smart Cities
Track user movements passively in a city mobility process based on mobile device sensor information. Identification of transportation mode used, time, distance and patterns. Research in collaboration with Centro Algoritmi at Minho University