Daisuke Mashima

Evaluating electricity theft detectors in smart grid networks

Electricity theft is estimated to cost billions of dollars per year in many countries. To reduce electricity theft, electric utilities are leveraging data collected by the new Advanced Metering Infrastructure (AMI) and using data analytics to identify abnormal consumption trends and possible fraud. In this paper, we propose the first threat model for the use of data analytics in detecting electricity theft, and a new metric that leverages this threat model in order to evaluate and compare anomaly detectors. We use real data from an AMI system to validate our approach.

Visualizing Dynamic Data with Maps

Maps offer a familiar way to present geographic data (continents, countries), and additional information (topography, geology), can be displayed with the help of contours and heat-map overlays. In this paper, we consider visualizing large-scale dynamic relational data by taking advantage of the geographic map metaphor. We describe a map-based visualization system which uses animation to convey dynamics in large data sets, and which aims to preserve the viewers mental map while also offering readable views at all times. Our system is fully functional and has been used to visualize user traffic on the Internet radio station last.fm, as well as TV-viewing patterns from an IPTV service. All map images in this paper are available in high-resolution at [CHECK END OF SENTENCE] as are several movies illustrating the dynamic visualization.

Enhancing accountability of electronic health record usage via patient-centric monitoring

Electronic Health Record (EHR) and Personal Health Record (PHR) systems could allow patients to better manage their health information and share it to enhance the quality and efficiency of their healthcare. Unfortunately, misuse of information stored in EHR and PHR systems will create new risks for patients, and we need to empower them to safeguard their health information to avoid problems such as medical identity theft. In this paper, we introduce the notion of accountable use and update of electronic health records and design a patient-centric monitoring system based on it. We develop a system architecture and associated protocols that enable either explicit or implicit patient control over when and how health information is accessed. Our approach provides a reasonable solution rather than addressing the more general information flow control problem in distributed systems. We also implement and evaluate a prototype system motivated by a health record sharing scenario based on NHIN Direct to demonstrate that enhanced accountability can be supported with acceptable performance and integration overheads.

Towards a User-Centric Identity-Usage Monitoring System

The misuse of identity-related information in cyberspace is one of the biggest concerns among all Internet users. So far several technologies have been proposed and implemented to prevent and detect identity theft and misuse, but none of them are completely successful in terms of privacy, user-centricity and generality. In this paper, we identify the requirements for a user-centric identity-usage monitoring system to solve such problems and propose a monitoring system that runs on a trusted third party. This system can transparently use context information of a request to detect anomalous use of online identity. Finally, we provide a prototypical implementation in an OpenID setting and evaluate it in terms of scalability, performance, user-centricity, and security.

Visualizing dynamic data with maps

Maps offer a familiar way to present geographic data (continents, countries), and additional information (topography, geology), can be displayed with the help of contours and heat-map overlays. In this paper we consider visualizing large-scale dynamic relational data by taking advantage of the geographic map metaphor. We describe a system that visualizes user traffic on the Internet radio station last.fm and address challenges in mental map preservation, as well as issues in animated map-based visualization.12

User-centric handling of identity agent compromise

Digital identity credentials are a key enabler for important online services, but widespread theft and misuse of such credentials poses serious risks for users. We believe that an identity management system (IdMS) that empowers users to become aware of how and when their identity credentials are used is critical for the success of such online services. Furthermore, rapid revocation and recovery of potentially compromised credentials is desirable. By following a user-centric identity-usage monitoring concept, we propose a way to enhance a user-centric IdMS by introducing an online monitoring agent and an inexpensive storage token that allow users to flexibly choose transactions to be monitored and thereby to balance security, privacy and usability. In addition, by utilizing a threshold signature scheme, our system enables users to revoke and recover credentials without communicating with identity providers. Our contributions include a system architecture, associated protocols and an actual implementation of an IdMS that achieves these goals.

Using identity credential usage logs to detect anomalous service accesses

For e-commerce companies providing online services, fraudulent access resulting from theft of identity credentials is a serious concern. Such online service providers deploy a variety of defenses and invest significant time and effort to the analysis of a large amount of log data to detect malicious activities and their impact. To reduce this burden, we explore the effectiveness of an anomaly detection based approach that relies on identity credential usage log records. More specifically, we use an anomaly-based metric to score the risk of each identity credential usage, e.g., a login request. Scores are determined based on categorical attribute values extracted from log records, such as timestamps. We utilize actual log data of login attempts to a university portal to evaluate the effectiveness of this approach. Our approach can work in conjunction with intrusion or fraud detection systems. It is also possible that stronger authentication can be required only when the risk score is high, which can help balance security and usability demands.

Privacy settings from contextual attributes: A case study using Google Buzz

Social networks provide users with privacy settings to control what information is shared with connections and other users. In this paper, we analyze factors influencing changes in privacy-related settings in the Google Buzz social network. Specifically, we show statistics on contextual data related to privacy settings that are derived from crawled datasets and analyze the characteristics of users who changed their privacy settings. We also investigate potential neighboring effects among such users.