DoSTra: Discovering Common Behaviors of Objects Using the Duration of Staying on Each Location of Trajectories

AAAI Conferences

Since semantic trajectories can discover more semantic meanings of a user’s interests without geographic restrictions, research on semantic trajectories has attracted a lot of attentions in recent years. Most existing work discover the similar behavior of moving objects through analysis of their semantic trajectory pattern, that is, sequences of locations. However, this kind of trajectories without considering the duration of staying on a location limits wild applications. For example, Tom and Anne have a common pattern of Home Restaurant Company Restaurant , but they are not similar, since Tom works at Restaurant , sends snack to someone at Company and return to Restaurant while Anne has breakfast at Restaurant , works at Company and has lunch at Restaurant . If we consider duration of staying on each location we can easily to differentiate their behaviors. In this paper, we propose a novel approach for discovering common behaviors by considering the duration of staying on each location of trajectories (DoSTra). Our approach can be used to detect the group that has similar lifestyle, habit or behavior patterns and predict the future locations of moving objects. We evaluate the experiment based on synthetic dataset, which demonstrates the high effectiveness and efficiency of the proposed method.


Algebraic Expression of Subjective Spatial and Temporal Patterns

arXiv.org Machine Learning

Universal learning machine is a theory trying to study machine learning from mathematical point of view. The outside world is reflected inside an universal learning machine according to pattern of incoming data. This is subjective pattern of learning machine. In [2,4], we discussed subjective spatial pattern, and established a powerful tool -- X-form, which is an algebraic expression for subjective spatial pattern. However, as the initial stage of study, there we only discussed spatial pattern. Here, we will discuss spatial and temporal patterns, and algebraic expression for them.


Learning what matters - Sampling interesting patterns

arXiv.org Machine Learning

In the field of exploratory data mining, local structure in data can be described by patterns and discovered by mining algorithms. Although many solutions have been proposed to address the redundancy problems in pattern mining, most of them either provide succinct pattern sets or take the interests of the user into account-but not both. Consequently, the analyst has to invest substantial effort in identifying those patterns that are relevant to her specific interests and goals. To address this problem, we propose a novel approach that combines pattern sampling with interactive data mining. In particular, we introduce the LetSIP algorithm, which builds upon recent advances in 1) weighted sampling in SAT and 2) learning to rank in interactive pattern mining. Specifically, it exploits user feedback to directly learn the parameters of the sampling distribution that represents the user's interests. We compare the performance of the proposed algorithm to the state-of-the-art in interactive pattern mining by emulating the interests of a user. The resulting system allows efficient and interleaved learning and sampling, thus user-specific anytime data exploration. Finally, LetSIP demonstrates favourable trade-offs concerning both quality-diversity and exploitation-exploration when compared to existing methods.


A Statistical Learning Theory Framework for Supervised Pattern Discovery

arXiv.org Machine Learning

This paper formalizes a latent variable inference problem we call {\em supervised pattern discovery}, the goal of which is to find sets of observations that belong to a single ``pattern.'' We discuss two versions of the problem and prove uniform risk bounds for both. In the first version, collections of patterns can be generated in an arbitrary manner and the data consist of multiple labeled collections. In the second version, the patterns are assumed to be generated independently by identically distributed processes. These processes are allowed to take an arbitrary form, so observations within a pattern are not in general independent of each other. The bounds for the second version of the problem are stated in terms of a new complexity measure, the quasi-Rademacher complexity.


A Semantic Design Pattern Language for Complex Event Processing

AAAI Conferences

In recent years we have seen the rise of a new type of software called Complex Event Processing (CEP) media. These are systems to detect and process complex events from large numbers of realtime events, so called event-clouds, occurring in distributed, possibly heterogeneous systems, databases and applications. This addresses an urgent need businesses do have nowadays: detection, prediction and mastery of complex situations as a crucial prerequisite to the efficiency of a dynamic service-oriented environments and the competitiveness of networked businesses in a future Internet of Services. However, there are a number of risks and difficulties that have to be taken into account when employing CEP. First industrial experiences in using the CEP technology and setting up CEP applications have shown that the potential adopters have major problems in understanding the CEP approach and adequately designing and implementing successful CEP solutions. Best practices and discussions about frequently occurring problems and their solutions in specific CEP application domains are missing and a systematic and profound debate about CEP patterns and anti-patterns is still in its early stage. CEP engineering remains a laborious trial and error process with slow development and change cycles. Hence, necessary theoretical groundwork to reach a thoroughly understanding of the proposed CEP methodology and technology and its applications in various problem domains needs to be done. This paper should "prepare the ground" for a more structured and methodological CEP engineering approach.