Thread structures, the reply relationships between posts, in online forums are very important for readers to understand the thread content, as well as for improving the effectiveness of automated forum information retrieval, expert findings, etc. However, most online forums only have partially labeled structures, which means that some reply relationships are known while the others are unknown. To address this problem, studies have been performed to learn and predict thread structures. However, existing work does not leverage the partially available thread structures to learn the complete thread structure. We have also observed that many online health forums are a type of person-centric forums, where persons are mentioned across posts, providing hints about the reply relationships between posts. In this paper, we first proposed to learn the complete thread structures by leveraging the partially known structures based on a statistical machine learning model: thread conditional random fields (threadCRF). Then we proposed to use person resolution, the process of identifying the same person mentioned in different contexts, together with threadCRF for thread structure learning. We have empirically verified the effectiveness of the proposed approaches.

Adverse Drug Reactions (ADRs) have become a serious health problem and even a leading cause of death in the United States. Pre‐marketing clinical trials and traditional post‐marketing surveillance using voluntary and spontaneous report systems are insufficient for ADR detection. On the other hand, online health forums provide valuable evidences in a large scale and in a timely fashion through the active participation of patients, caregivers, and doctors. In this article, we present patient‐centered and experience‐aware mining framework for effective ADR discovery using online health forum data. Our experimental evaluation with both an official ADR knowledge base and human‐annotated ground truth verifies the effectiveness of the proposed method for ADR discovery.

Online healthcare forums provide a valuable platform for people to discover knowledge. However, existing approaches rely on the syntactic information units, such as a sentence, a post, or a thread, to bind different pieces of information in a forum. In this work, we propose to connect the pieces of information by a semantic information unit, patients. Specifically, we connect the information of diseases, symptoms, treatments, and effects to the patient who experiences them, and thus enable effective knowledge discovery.

RFID (radio-frequency identification) is an emerging technology with extensive applications such as transportation and logistics, object tracking, and inventory management. How to quickly identify the missing RFID tags and thus their associated objects is a practically important problem in many large-scale RFID systems. This paper presents three novel methods to quickly identify the missing tags in a large-scale RFID system of thousands of tags. Our protocols can reduce the time for identifying all the missing tags by up to 75% in comparison to the state of art.

People-centric urban sensing is a new paradigm gaining popularity. A main obstacle to its widespread deployment and adoption are the privacy concerns of participating individuals. To tackle this open challenge, this paper presents the design and evaluation of PriSense, a novel solution to privacy-preserving data aggregation in people-centric urban sensing systems. PriSense is based on the concept of data slicing and mixing and can support a wide range of statistical additive and non-additive aggregation functions such as Sum, Average, Variance, Count, Max/Min, Median, Histogram, and Percentile with accurate aggregation results. PriSense can support strong user privacy against a tunable threshold number of colluding users and aggregation servers. The efficacy and efficiency of PriSense are confirmed by thorough analytical and simulation results.

Most large-scale sensor networks are expected to follow a two-tier architecture with resource-poor sensor nodes at the lower tier and resource-rich master nodes at the upper tier. Master nodes collect data from sensor nodes and then answer the queries from the network owner on their behalf. In hostile environments, master nodes may be compromised by the adversary and then instructed to return fake and/or incomplete data in response to data queries. Such application-level attacks are more harmful and difficult to detect than blind DoS attacks on network communications, especially when the query results are the basis for making critical decisions such as military actions. This paper presents three schemes whereby the network owner can verify the authenticity and completeness of fine-grained top-k query results in tired sensor networks, which is the first work of its kind. The proposed schemes are built upon symmetric cryptographic primitives and force compromised master nodes to return both authentic and complete top-k query results to avoid being caught. Detailed theoretical and quantitative results confirm the high efficacy and efficiency of the proposed schemes.

The open wireless medium in a mobile ad-hoc network (MANET) enables malicious traffic analysis to dynamically infer the network traffic pattern in hostile environments. The disclosure of the traffic pattern and its changes is often devastating in a mission-critical MANET. A number of anonymous routing protocols have been recently proposed as an effective countermeasure against traffic analysis in MANETs. In this paper, we propose a novel traffic inference algorithm, called TIA, which enables a passive global adversary to accurately infer the traffic pattern in an anonymous MANET without compromising any node. As the first work of its kind, TIA works on existing on-demand anonymous MANET routing protocols. Detailed simulations show that TIA can infer the traffic pattern with an accuracy as high as 95%. Our results in this paper highlight the necessity for cross-layer designs to defend a MANET against traffic analysis.