Nowadays, autonomous driving cars have become commercially available. However, the safety of a self-driving car is still a challenging problem that has not been well studied. Motion prediction is one of the core functions of an autonomous driving car. In this paper, we propose a novel scheme called GRIP which is designed to predict trajectories for traffic agents around an autonomous car efficiently. GRIP uses a graph to represent the interactions of close objects, applies several graph convolutional blocks to extract features, and subsequently uses an encoder-decoder long short-term memory (LSTM) model to make predictions. The experimental results on two wellknown public datasets show that our proposed model improves the prediction accuracy of the state-of-the-art solution by 30%. The prediction error of GRIP is one meter shorter than existing schemes. Such an improvement can help autonomous driving cars avoid many traffic accidents. In addition, the proposed GRIP runs 5x faster than the state-of-the-art schemes.

Social networking sites have become very popular in recent years. Users use them to find new friends, updates their existing friends with their latest thoughts and activities. Among these sites, Twitter is the fastest growing site. Its popularity also attracts many spammers to infiltrate legitimate users' accounts with a large amount of spam messages. In this paper, we discuss some userbased and content-based features that are different between spammers and legitimate users. Then, we use these features to facilitate spam detection. Using the API methods provided by Twitter, we crawled active Twitter users, their followers/following information and their most recent 100 tweets. Then, we analyzed the collected dataset and evaluated our detection scheme based on the suggested user and content-based features. Our results show that among the four classifiers we evaluated, the Random Forest classifier produces the best results. Our results based on the 100 most recent tweets also show that spam detection based on our suggested features can achieve 95.7% precision and 95.7% Fmeasure using the Random Forest classifier.

No abstract

Abstract-As Cloud Computing technology becomes more mature, many organizations and individuals are interested in storing more sensitive data e.g. personal health records, customers related information in the cloud. Such sensitive data needs to be encrypted before it is outsourced to the cloud. Typically, the cloud servers also need to support a keyword search feature for these encrypted files. Traditional searchable encryption schemes typically only support exact keyword matches. However, users sometimes have typos or use slightly different formats e.g. "data-mining" versus "data mining". Thus, fuzzy keyword search is a useful feature to have. Recently, some researchers propose using wildcard based approach to provide fuzzy keyword search. They also propose a solution for multikeyword search. Their approaches have some limitations, namely (a) their fuzzy keyword search solution consumes large storage size since it inserts every fuzzy keyword as a leaf node in the index tree, (b) their fuzzy single-keyword search solution does not support multi-keyword search, (c) the existing multi-keyword search scheme does not provide efficient incremental updates. In this paper, we propose a privacy-aware bedtree based approach to support fuzzy multi-keyword feature. Incremental updates can be easily done using our solution. We have implemented our solution. Our evaluation results show that our approach is more cost-effective in terms of storage size and construction time. Our search time is usually better than the wildcard approach for multi-keyword queries where many encrypted files are returned using single-word queries for approaches that do not support multi-keyword queries.

The rapid deployment of sensing technology in smartphones and the explosion of their usage in people's daily lives provide users with the ability to collectively sense the world. This leads to a growing trend of mobile healthcare systems utilizing sensing data collected from smartphones with/without additional external sensors to analyze and understand people's physical and mental states. However, such healthcare systems are vulnerable to user spoofing, in which an adversary distributes his registered device to other users such that data collected from these users can be claimed as his own to obtain more healthcare benefits and undermine the successful operation of mobile healthcare systems. Existing mitigation approaches either only rely on a secret PIN number (which can not deal with colluded attacks) or require an explicit user action for verification. In this paper, we propose a user verification system leveraging unique gait patterns derived from acceleration readings to detect possible user spoofing in mobile healthcare systems. Our framework exploits the readily available accelerometers embedded within smartphones for user verification. Specifically, our user spoofing mitigation framework (which consists of three components, namely Step Cycle Identification, Step Cycle Interpolation, and Similarity Comparison) is used to extract gait patterns from run-time accelerometer measurements to perform robust user verification under various walking speeds. We show that our framework can be implemented in two ways: user-centric and server-centric, and it is robust to not only random but also mimic attacks. Our extensive experiments using over 3,000 smartphone-based traces with mobile phones placed on different body positions confirm the effectiveness of the proposed framework with users walking at various speeds. This strongly indicates the feasibility of using smartphone based low grade accelerometer to conduct gait recognition and facilitate effective user verification without active user cooperation.