Matches in ScholarlyData for { ?s ?p ?o. }
- authorList hasItem item-3.
- authorList hasItem item-4.
- authorList hasLastItem item-4.
- item-1 type ListItem.
- item-1 hasContent yeliz-yesilada.
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent caroline-jay.
- item-2 next item-3.
- item-2 previous item-1.
- item-3 type ListItem.
- item-3 hasContent robert-stevens.
- item-3 next item-4.
- item-3 previous item-2.
- item-4 type ListItem.
- item-4 hasContent simon-harper.
- item-4 previous item-3.
- 569 creator anirban-dasgupta.
- 569 creator jure-leskovec.
- 569 creator kevin-lang.
- 569 creator michael-mahoney.
- 569 type InProceedings.
- 569 label "Community structure in real graphs: the ''negative dimensionality'' paradox".
- 569 sameAs 569.
- 569 abstract "Identifying communities in networks is one of the fundamental interests of network analysis, where communities are usually thought of as sets of nodes that interact more strongly with each other than with the remainder of the network. In this work we ask questions about statistical properties of such communities in large real-world networks: To what extent is community structure expressed in large real-world networks? What are typical community sizes and quality? How do we measure and characterize this? We define the {\em network community profile plot} that characterizes network community structure at various scales. We study almost 70 different large real-world networks and find that tightest communities tend to be small and mostly trivially separable from the graph. We also observe a counterintuitive ``negative dimensionality'' of the community profile, an inverse relationship between cluster quality and cluster size, which is exactly the opposite of what we would expect based on experience with both large graphs and also small social networks. Intuitively, for large real-world networks, we have tight communities at smaller scales, but as the community grows it vanishes, blending with the network. This behavior is not explained by most of the popular network models, even at a qualitative level. The recently-developed Forest Fire model, however, tends to produce graphs exhibiting similar community structure as found in real networks.".
- 569 hasAuthorList authorList.
- 569 hasTopic World_Wide_Web.
- 569 isPartOf proceedings.
- 569 keyword "conductance".
- 569 keyword "graph partitioning".
- 569 keyword "on-line communities".
- 569 keyword "random walks".
- 569 keyword "social networks".
- 569 title "Community structure in real graphs: the ''negative dimensionality'' paradox".
- authorList type List.
- authorList hasFirstItem item-1.
- authorList hasItem item-1.
- authorList hasItem item-2.
- authorList hasItem item-3.
- authorList hasItem item-4.
- authorList hasLastItem item-4.
- item-1 type ListItem.
- item-1 hasContent jure-leskovec.
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent kevin-lang.
- item-2 next item-3.
- item-2 previous item-1.
- item-3 type ListItem.
- item-3 hasContent anirban-dasgupta.
- item-3 next item-4.
- item-3 previous item-2.
- item-4 type ListItem.
- item-4 hasContent michael-mahoney.
- item-4 previous item-3.
- 590 creator honglei-guo.
- 590 creator xian-wu.
- 590 creator xiaoxun-zhang.
- 590 creator xueying-wang.
- 590 creator zhili-guo.
- 590 creator zhong-su.
- 590 type InProceedings.
- 590 label "FloatCascade Learning for Fast Imbalanced Web Mining".
- 590 sameAs 590.
- 590 abstract "This paper is concerned with the problem of Imbalanced Classification (IC) in web mining, which often arises on the web due to the “Matthew Effect”. As web IC applications usually need to provide online service for user and deal with large volume of data, classification speed emerges as an important issue to be addressed. In face detection, Asymmetric Cascade is used to speed up imbalanced classification by building a cascade structure of simple classifiers, but it often causes a loss of classification accuracy due to the iterative feature addition in its learning procedure. In this paper, we adopt the idea of cascade classifier in imbalanced web mining for fast classification and propose a novel asymmetric cascade learning method called FloatCascade to improve the accuracy. To the end, FloatCascade selects fewer but more effective features at each stage of the cascade classifier. In addition, a decision-tree scheme is adopted to enhance feature diversity and discrimination capability for FloatCascade learning. We evaluate FloatCascade through two typical IC applications in web mining: web page categorization and citation matching. Experimental results demonstrate the effectiveness and efficiency of FloatCascade comparing to the state-of-the-art IC methods like Asymmetric Cascade, Asymmetric AdaBoost and Weighted SVM.".
- 590 hasAuthorList authorList.
- 590 hasTopic World_Wide_Web.
- 590 isPartOf proceedings.
- 590 keyword "Cascade learning".
- 590 keyword "Citation matching".
- 590 keyword "Fast classfication".
- 590 keyword "Imbalanced web mining".
- 590 keyword "Web page categorization".
- 590 title "FloatCascade Learning for Fast Imbalanced Web Mining".
- authorList type List.
- authorList hasFirstItem item-1.
- authorList hasItem item-1.
- authorList hasItem item-2.
- authorList hasItem item-3.
- authorList hasItem item-4.
- authorList hasItem item-5.
- authorList hasItem item-6.
- authorList hasLastItem item-6.
- item-1 type ListItem.
- item-1 hasContent xiaoxun-zhang.
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent xueying-wang.
- item-2 next item-3.
- item-2 previous item-1.
- item-3 type ListItem.
- item-3 hasContent honglei-guo.
- item-3 next item-4.
- item-3 previous item-2.
- item-4 type ListItem.
- item-4 hasContent zhili-guo.
- item-4 next item-5.
- item-4 previous item-3.
- item-5 type ListItem.
- item-5 hasContent xian-wu.