An incremental FP-growth web content mining and its application in preference identification

This paper presents a real application of Web-content mining using an incremental FP-Growth approach. We firstly restructure the semi-structured data retrieved from the web pages of Chinese car market to fit into the local database, and then employ an incremental algorithm to discover the association rules for the identification of car preference. To find more general regularities, a method of attribute-oriented induction is also utilized to find customer’s consumption preferences. Experimental results show some interesting consumption preference patterns that may be beneficial for the government in making policy to encourage and guide car consumption.

An web content based data mining for car consumption in China

This paper introduces an incremental FP-Growth approach for Web content based data mining and its application in solving a real world problem The problem is solved in the following ways. Firstly, we obtain the semi-structured data from the Web pages of Chinese car market and structure them and save them in local database. Secondly, we use an incremental FP-Growth algorithm for mining association rules to discover Chinese consumers' car consumption preference. To find more general regularities, an attribute-oriented induction method is also utilized to find customer's consumption preference among a range of car categories. Experimental results have revealed some interesting consumption preferences that are useful for the decision makers to make the policy to encourage and guide car consumption. Although the current data we used may not be the best representative of the actual market in practice, it is still good enough for the decision making purpose in terms of reflecting the real situation of car consumption preference under the two assumptions in the context.

Mining sequential causal patterns with user-specified skeletons in multi-sequence of event data

The approaches proposed in the past for discovering sequential patterns mainly focused on single sequential data. In the real world, however, some sequential patterns hide their essences among multi-sequential event data. It has been noted that knowledge discovery with either user-specified constraints, or templates, or skeletons is receiving wide attention because it is more efficient and avoids the tedious selection of useful patterns from the mass-produced results. In this paper, a novel pattern in multi-sequential event data that are correlated and its mining approach are presented. We call this pattern sequential causal pattern. A group of skeletons of sequential causal patterns, which may be specified by the user or generated by the program, are verified or mined by embedding them into the mining engine. Experiments show that this method, when applied to discovering the occurring regularities of a crop pest in a region, is successful in mining sequential causal patterns with user-specified skeletons in multi-sequential event data.

Entropy-based collaborative detection of DDOS attacks on community networks

A community network often operates with the same Internet service provider domain or the virtual network of different entities who are cooperating with each other. In such a federated network environment, routers can work closely to raise early warning of DDoS attacks to void catastrophic damages. However, the attackers simulate the normal network behaviors, e.g. pumping the attack packages as poisson distribution, to disable detection algorithms. It is an open question: how to discriminate DDoS attacks from surge legitimate accessing. We noticed that the attackers use the same mathematical functions to control the speed of attack package pumping to the victim. Based on this observation, the different attack flows of a DDoS attack share the same regularities, which is different from the real surging accessing in a short time period. We apply information theory parameter, entropy rate, to discriminate the DDoS attack from the surge legitimate accessing. We proved the effectiveness of our method in theory, and the simulations are the work in the near future. We also point out the future directions that worth to explore in the future.

Constructing the monster : notions of the monstrous in classical antiquity

Focussing on humaniod monsters, this thesis uses insights from Foucault's theory about the "archaeology" of discourses and Derrida's practice of deconstruction to examine how monstrosity was spoken of in antiquity, and how the various "sciences" dealt with anomalous monsters without jeopardising their epistemological credibility. Discussion begins with a survey of the semantic field of teras and monstrum. Since portentousness was central to both terms, the signification of monstrous portents in divinatory practice is next aalysed in the historiography of Herodotus, Livy, and others. Cicero's De divinatione reveals the theory and the problem for that science posed by accidental monstrosities. Chance and novelty are also issues in mythical and scientific cosmogonies < of Hesiod, and Orphism, Empedo-cles, and Lucretius> , where monsters arise and are dealt with while cosmic regularities, reproductive and ethical, are being established. Teleology and the stability of species'forms emerge as important concerns. These issues are further considered in Aristotle's bioogy and in medical writings from Hippocrates to Galen. There, theories are produced about monstrous embryology which attempt to answer the question of how deformities occur if species' forms are perpetuated through repro-duction. Biological and taxonomic--as well as ethical--boundaries are violated also by mythic human-beast hybrids. Narratives about such anomalies clarify the nature of monstrous deviance and enact solutions to the problem. Their strategies have much in common with other modes of discourse. Ethnography is posed similar questions about monstrous races' physical and ethical deviations from the civilised norm; it speaks of those issues in terms of invariance of form through generations, geographical remoteness and the codes which situate those races ethically. Finally, Augustine’s discourse on monstrous individuals and races is examined as Christianity’s belief in God’s governance reformulates the ancient’s discussions of chane or novelty and the invariance of species. In all these discourses founded on determinate meaning, the persistant paradox of monstrosity need offer no challenge to rationality provided its indefinable diversity is unacknowledged and the notion is constructed in such a way as to reaffirm the certainties.

Portfolio diversification and share price behavior : some evidence from the Taiwan Stock Exchange

This dissertation consists of four separate but closely related studies which investigate different aspects of share price behavior on the Taiwan Stock Exchange over the period 1980-89: 1.The benefits of diversification available to investors using the Markowitz model and the Single Model Index. 2. The applicability of the CAPM to the TSE over the decade. 3. Regularities in proce sequences. 4. Market reaction to the announcements of stock dividends, right issues and combinations of both.

Learning Boltzmann distance metric for face recognition

We introduce a new method for face recognition using a versatile probabilistic model known as Restricted Boltzmann Machine (RBM). In particular, we propose to regularise the standard data likelihood learning with an information-theoretic distance metric defined on intra-personal images. This results in an effective face representation which captures the regularities in the face space and minimises the intra-personal variations. In addition, our method allows easy incorporation of multiple feature sets with controllable level of sparsity. Our experiments on a high variation dataset show that the proposed method is competitive against other metric learning rivals. We also investigated the RBM method under a variety of settings, including fusing facial parts and utilising localised feature detectors under varying resolutions. In particular, the accuracy is boosted from 71.8% with the standard whole-face pixels to 99.2% with combination of facial parts, localised feature extractors and appropriate resolutions.

Learning sparse latent representation and distance metric for image retrieval

The performance of image retrieval depends critically on the semantic representation and the distance function used to estimate the similarity of two images. A good representation should integrate multiple visual and textual (e.g., tag) features and offer a step closer to the true semantics of interest (e.g., concepts). As the distance function operates on the representation, they are interdependent, and thus should be addressed at the same time. We propose a probabilistic solution to learn both the representation from multiple feature types and modalities and the distance metric from data. The learning is regularised so that the learned representation and information-theoretic metric will (i) preserve the regularities of the visual/textual spaces, (ii) enhance structured sparsity, (iii) encourage small intra-concept distances, and (iv) keep inter-concept images separated. We demonstrate the capacity of our method on the NUS-WIDE data. For the well-studied 13 animal subset, our method outperforms state-of-the-art rivals. On the subset of single-concept images, we gain 79:5% improvement over the standard nearest neighbours approach on the MAP score, and 45.7% on the NDCG.