Paolo Cappellari

Linking and using social media data for enhancing public health analytics

There is a large amount of health information available for any patient to address his/her health concerns. The freely available health datasets include community health data at the national, state, and community level, readily accessible and downloadable. These datasets can help to assess and improve healthcare performance, as well as help to modify health-related policies. There are also patient-generated datasets, accessible through social media, on the conditions, treatments, or side effects that individual patients experience. Clinicians and healthcare providers may benefit from being aware of national health trends and individual healthcare experiences that are relevant to their current patients. The available open health datasets vary from structured to highly unstructured. Due to this variability, an information seeker has to spend time visiting many, possibly irrelevant, Websites, and has to select information from each and integrate it into a coherent mental model. In this paper, we discuss an approach to integrating these openly available health data sources and presenting them to be easily understandable by physicians, healthcare staff, and patients. Through linked data principles and Semantic Web technologies we construct a generic model that integrates diverse open health data sources. The integration model is then used as the basis for developing a set of analytics as part of a system called ‘Social InfoButtons’, providing awareness of both community and patient health issues as well as healthcare trends that may shed light on a specific patient care situation. The prototype system provides patients, public health officials, and healthcare specialists with a unified view of health-related information from both official scientific sources and social networks, and provides the capability of exploring the current data along multiple dimensions, such as time and geographical location.

A Tool for Automatic Assessment and Awareness of Privacy Disclosure

With increasing frequency, the communication between citizens and institutions occurs via some type of e-mechanism, such as websites, emails, and social media. In particular, social media platforms are widely being adopted because of their simplicity of use, the large user base, and their high pervasiveness. One concern is that users may disclose sensitive information beyond the scope of the interaction with the institutions, not realizing that such data remains on these platforms. While awareness about basic data (e.g. address, date of birth) protection has risen in the past few years, many users still neglect or fail to realize the amount and significance of the personal information deliberately or involuntarily disclosed on these communication platforms. Determining private from non-private data is difficult. The goal of this work is to devise a method to detect messages carrying sensitive information from those that not. Specifically, we employ machine learning methods to build a privacy decision making tool. This work will contribute to develop a privacy protection framework where a client-side privacy awareness mechanism can alert users of the potential private information leakages in their communications.

ISE: A High Performance System for Processing Data Streams

Many organizations require the ability to manage high-volume high-speed streaming data to perform analysis and other tasks in real-time. In this work, we present the Information Streaming Engine, a high-performance data stream processing system capable of scaling to high data volumes while maintaining very low-latency. The Information Streaming Engine adopts a declarative approach which enables processing and manipulation of data streams in a simple manner. Our evaluation demonstrates the high levels of performance achieved when compared to existing systems.

Leveraging Social Data for Health Care Behavior Analytics

User-generated social health data can provide valuable insights into the health care behavior and practices of patients for clinicians, policy makers and other patients. Social InfoButtons is a tool for integrating and analyzing social health data, using a semantic data integration framework that is flexible and extensible with respect to a variety of health data sources. The integration overlay provides a framework that semantically links data enabling cross dataset exploration and analysis. The unified data model for both scientific and social sources allows multi-dimensional analysis and the ability to explore and compare the expert knowledge with the actual health care practices of patients.

Detecting and Analyzing Privacy Leaks in Tweets.

Social network platforms are changing the way people interact not just with each other but also with companies and institutions. In sharing information on these platforms, users often underestimate potential consequences, especially when such information discloses personal information. For such reason, actionable privacy awareness and protection mechanisms are becoming of paramount importance. In this paper we propose an approach to assess the privacy content of the social posts with the goal of: protecting the users from inadvertently disclosing sensitive information, and rising awareness about privacy in online behavior. We adopt a machine learning approach based on a crowd-sourced definition of privacy that can assess whether messages are disclosing sensitive information. Our approach can automatically detect messages carrying sensitive information, so to warn users before sharing a post, and provides a set of analysis to rise users awareness about online behavior related to privacy disclosure.

Identifying Electromyography Sensor Placement using Dense Neural Networks.

Neural networks are increasingly being used in medical settings to support medical practitioners and researchers in performing their work. In the field of prosthetics for amputees, sensors can be used to monitor the activity of remaining muscle and ultimately control prosthetic limbs. In this work, we present an approach to identify the location of intramuscular electromyograph sensors percutaneously implanted in extrinsic muscles of the forearm controlling the fingers and wrist during single digit movements. A major challenge is to confirm whether each sensor is placed in the targeted muscle, as this information can be critical in developing and implementing control systems for prosthetic limbs. We propose an automated approach, based on artificial neural networks, to identify the correct placement of an individual sensor. Our approach can provide feedback on each placed sensor, so researchers can validate the source of each signal before performing their data

Optimizing data stream processing for large-scale applications: Optimizing Data Stream Processing

Stream processing systems are designed to analyze data arriving in real time and using continuous queries and respond when a specific event or sequence of events are detected. An important aspect of these systems is Streaming Analytics, which facilitates statistical calculations on continuous data within the stream. These systems must be designed to handle high volumes of data, be scalable, and accommodate a multitude of long‐lived concurrently running analytics. The challenges involved in the development of stream processing include on‐the‐fly transformation of data streams to match the query needs of users and the ability to model stream transformations to detect overlaps and possibilities for optimizations and to specify a methodology to deliver optimizations. In particular, this work focuses on exposing data stream application internals in order to detect reusable parts and then consolidate applications to optimize computational resource usage. The Streaming Data Analytics Model presented in this paper adopts a declarative approach that enables processing and manipulation of data streams in a simple manner while facilitating powerful optimizations necessary for managing high volumes of streaming data in real time. An evaluation is provided to demonstrate in both theoretical and quantitative aspects the high performance offered by our approach.

Discovering and Analyzing Alternative Treatments Hypothesis via Social Health Data

User-generated social health data can provide valuable information to extend the status of the medical knowledge. We present a tool geared towards social health data exploration and reasoning. Starting from a repository of semantically linked social health data, we enable researchers to discover alternative treatments as well as similar conditions by exploring the semantic repository via potentially compatible concepts. Researchers are prompted with the features of the concepts under investigation to analyze similarities and contradictions, when present. Concepts are enriched with confidence values that help researchers in assessing the reliability of the information they are analyzing.

A Scalable Platform for Low-Latency Real-Time Analytics of Streaming Data

The ability to process high-volume high-speed streaming data from different data sources is critical for modern organizations to gain insights for business decisions. In this research, we present the streaming analytics platform (SDAP), which provides a set of operators to specify the process of stream data transformations and analytics. SDAP adopts a declarative approach to model and design, delivering analytics capabilities through the combination of a set of primitive operators in a simple manner. The model includes a topology to design streaming analytics specifications using a set of atomic data manipulation operators. Our evaluation demonstrates that SDAP is capable of maintaining low-latency while scaling to a cloud of distributed computing nodes, and providing easier process design and execution of streaming analytics.