Big data analytics


The usage of data in the process of decision making is nowadays interesting several fields, from business to education. However, the cost of collecting data is very low with respect to the value of the massive and heterogeneous amount of collected information itself (Big-Data). The process of retrieving information out of these data, to support organisations in making decisions, is called Big-Data Analytics.

The usage of data in the process of decision making is nowadays interesting several fields, from business to education. Data is used in general to define needs, to set goals, to plan interventions, and to evaluate progress. Datasets are growing rapidly in part because they are increasingly gathered by cheap and numerous information-sensing mobile devices (through cameras, microphones, wireless sensor networks), and at a very irrelevant cost by web servers, Internet clickstream, social media, social networks, mobile-phone call detail records, and machines data captured by sensors connected to the Internet of Things.
This massive and heterogeneous amount of collected information is called Big-Data, which differently from classical datasets might include structured, semi-structured and unstructured data. The process of collecting, organizing and analyzing these large sets of data, in order to discover patterns and other useful information, is called Big-Data analytics. This process supports organisations when they have to make decisions, in fact it helps to better understand the collected information and to identify the most relevant data.

Researchers working at ISIN have in the last years developed relevant skills in order to work with BigData. These skills can be classified in three main fields.

  1. Collection of BigData: Internet of Things (especially applied to smart homes and health-care/health-monitoring) and mobile sensing.
  2. Managing BigData: data warehouse based on relational databased for structured data, and new technologies to handle semi-structured and unstructured data, which include Hadhoop and NoSQL databases.
  3. Analysing BigData: machine learning, high-performance data mining, predictive analytics and optimisation.

Further Information

Dr. Michela Papandrea, Researcher

st.wwwsupsi@supsi.ch