Semantic trajectory based event detection and event pattern mining

Video event detection is an effective way to automatically understand the semantic content of the video. However, due to the mismatch between low-level visual features and high-level semantics, the research of video event detection encounters a number of challenges, such as how to extract the suitable information from video, how to represent the event, how to build up reasoning mechanism to infer the event according to video information. In this paper, we propose a novel event detection method. The method detects the video event based on the semantic trajectory, which is a high-level semantic description of the moving object’s trajectory in the video. The proposed method consists of three phases to transform low-level visual features to middle-level raw trajectory information and then to high-level semantic trajectory information. Event reasoning is then carried out with the assistance of semantic trajectory information and background knowledge. Additionally, to release the users’ burden in manual event definition, a method is further proposed to automatically discover the event-related semantic trajectory pattern from the sample semantic trajectories. Furthermore, in order to effectively use the discovered semantic trajectory patterns, the associative classification-based event detection framework is adopted to discover the possibly occurred event. Empirical studies show our methods can effectively and efficiently detect video events.

On the incremental learning and recognition of the pattern of movement of multiple labelled objects in dynamic scenes

In this paper we discuss combining incremental learning and incremental recognition to classify patterns consisting of multiple objects, each represented by multiple spatio-temporal features. Importantly the technique allows for ambiguity in terms of the positions of the start and finish of the pattern. This involves a progressive classification which considers the data at each time instance in the query and thus provides a probable answer before all the query information becomes available. We present two methods that combine incremental learning and incremental recognition: a time instance method and an overall best match method.

Infrequent item mining in multiple data streams

The problem of extracting infrequent patterns from streams and building associations between these patterns is becoming increasingly relevant today as many events of interest such as attacks in network data or unusual stories in news data occur rarely. The complexity of the problem is compounded when a system is required to deal with data from multiple streams. To address these problems, we present a framework that combines the time based association mining with a pyramidal structure that allows a rolling analysis of the stream and maintains a synopsis of the data without requiring increasing memory resources. We apply the algorithms and show the usefulness of the techniques. © 2007 Crown Copyright.

A multi-agent incremental negotiation scheme for meetings scheduling

We present an agent-oriented approach to the meeting scheduling problem and propose an incremental negotiation scheme that makes use of a hierarchical structure of an individual agent's working knowledge. First, we formalise the meeting scheduling problem in a multi-agent context, then elaborate on the design of a common agent architecture of all agents in the system. As a result, each agent becomes a modularised computing unit yet possesses high autonomy and robust interface with other agents. The system reserves the meeting participants' privacy since there are no agents with dominant roles, and agents can communicate at an abstract level in their hierarchical structures

Face recognition via incremental 2DPCA

Recently, the Two-Dimensional Principal Component Analysis (2DPCA) model is proposed and proved to be an efficient approach for face recognition. In this paper, we will investigate the incremental 2DPCA and develop a new constructive method for incrementally adding observation to the existing eigen-space model. An explicit formula for incremental learning is derived. In order to illustrate the effectiveness of the proposed approach, we performed some typical experiments and show that we can only keep the eigen-space of previous images and discard the raw images in the face recognition process. Furthermore, this proposed incremental approach is faster when compared to the batch method (2DPCD) and the recognition rate and reconstruction accuracy are as good as those obtained by the batch method.

Mining condensed sets of frequent episodes with more accurate frequencies from complex sequences

Many previous approaches to frequent episode discovery only accept simple sequences. Although a recent approach has been able to nd frequent episodes from complex sequences, the discovered sets are neither condensed nor accurate. This paper investigates the discovery of condensed sets of frequent episodes from complex sequences. We adopt a novel anti-monotonic frequency measure based on non-redundant occurrences, and dene a condensed set, nDaCF (the set of non-derivable approximately closed frequent episodes) within a given maximal error bound of support. We then introduce a series of effective pruning strategies, and develop a method, nDaCF-Miner, for discovering nDaCF sets. Experimental results show that, when the error bound is somewhat high, the discovered nDaCF sets are two orders of magnitude smaller than complete sets, and nDaCF-miner is more efficient than previous mining approaches. In addition, the nDaCF sets are more accurate than the sets found by previous approaches.

A comparative analysis of decision trees vis-à-vis other computational data mining techniques in automotive insurance fraud detection

The development and application of computational data mining techniques in financial fraud detection and business failure prediction has become a popular cross-disciplinary research area in recent times involving financial economists, forensic accountants and computational modellers. Some of the computational techniques popularly used in the context of - financial fraud detection and business failure prediction can also be effectively applied in the detection of fraudulent insurance claims and therefore, can be of immense practical value to the insurance industry. We provide a comparative analysis of prediction performance of a battery of data mining techniques using real-life automotive insurance fraud data. While the data we have used in our paper is US-based, the computational techniques we have tested can be adapted and generally applied to detect similar insurance frauds in other countries as well where an organized automotive insurance industry exists.