Visual matrix explorer for collaborative seriation

Liiv, Innar; Opik, Rain; Ubi, Jaan; Stasko, John

doi:10.1002/wics.193

Cited by 8 publications

(6 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We regard seriation as "sequencing objects along a continuum that rely upon a symmetric proximity measure defined between the objects to be seriated" (Hubert, 1974). Seriation is widely used in the visualisation of binary matrices (Chen, 2002;Liiv et al, 2011).…”

Section: Mapping Evolving Semantic Structuresmentioning

confidence: 99%

“…Two-way seriation assumes a matrix representation, or a two-dimensional layout of the data, where rows correspond to instances and columns correspond to features that describe the individual instances (this assignment of rows and columns is arbitrary and can happen the other way). This approach has a history of over a hundred years (see (Liiv, 2010) for an overview), and it is commonly used in a wide variety of information visualisation methods, including microarray data (Eisen et al, 1998;Caraux and Pinloche, 2005), binary matrices (Liiv et al, 2011), and others (Chen, 2002). Thus two-way seriation introduces a kind of regionality: a local neighbourhood is guaranteed to have related elements.…”

Section: Mapping Evolving Semantic Structuresmentioning

confidence: 99%

See 1 more Smart Citation

Demonstrating conceptual dynamics in an evolving text collection

Daranyi

Wittek

2013

J Am Soc Inf Sci Tec

View full text Add to dashboard Cite

Based on real world user demands, we demonstrate how animated visualisation of evolving text corpora displays the underlying dynamics of semantic content. To interpret the results, one needs a dynamic theory of word meaning. We suggest that conceptual dynamics as the interaction between kinds of intellectual, emotional etc. content, and language, is key for such a theory. We demonstrate our methodology by two-way seriation which is a popular technique to analyse groups of similar instances and their features, as well as the connections between the groups themselves. The two-way seriated data may be visualised as a two-dimensional heat map or as a three-dimensional landscape where colour codes or height correspond to the values in the matrix. In this paper we focus on two-way seriation of sparse data in the Reuters-21568 test collection. To achieve a meaningful visualisation thereof we introduce a compactly supported convolution kernel similar to filter kernels used in image reconstruction and geostatistics. This filter populates the high-dimensional sparse space with values that interpolate nearby elements, and provides insight into the clustering structure. We also extend two-way seriation to deal with online updates of both the row and column spaces, and, combined with the convolution kernel, demonstrate a three-dimensional visualisation of dynamics.

show abstract

Section: Mapping Evolving Semantic Structuresmentioning

confidence: 99%

Section: Mapping Evolving Semantic Structuresmentioning

confidence: 99%

Demonstrating conceptual dynamics in an evolving text collection

Daranyi

Wittek

2013

J Am Soc Inf Sci Tec

View full text Add to dashboard Cite

show abstract

“…Two-way seriation assumes a matrix representation, or a two-dimensional layout of the data, where rows correspond to instances and columns correspond to features that describe the individual instances (this assignment of rows and columns is arbitrary and can happen the other way). This approach has a history of over a hundred years (see [23] for an overview), and it is commonly used in a wide variety of information visualization methods, including microarray data [5,11], binary matrices [24], and others [6]. Seriation is similar to clustering, and two-way seriation is similar to biclustering: instances with similar feature subsets are grouped together in regular patterns.…”

Section: Seriation and Two-way Seriationmentioning

confidence: 99%

“…We regard seriation as "sequencing objects along a continuum that rely upon a symmetric proximity measure defined between the objects to be seriated" [19]. Seriation is widely used in the visualization of binary matrices [6,24], and it is also used in genetic research to understand which genes are activated simultaneously [5,11].…”

Section: Introductionmentioning

confidence: 99%

Two-way incremental seriation in the temporal domain with three-dimensional visualization: Making sense of evolving high-dimensional datasets

Wittek

2013

Computational Statistics & Data Analysis

View full text Add to dashboard Cite

Two-way seriation is a popular technique to analyse groups of similar instances and their features, as well as the connections between the groups themselves. The two-way seriated data may be visualized as a two-dimensional heat map or as a three-dimensional landscape where colour codes or height correspond to the values in the matrix. To achieve a meaningful visualization of high-dimensional data, a compactly supported convolution kernel is introduced, which is similar to filter kernels used in image reconstruction and geostatistics. This filter populates the high-dimensional space with values that interpolate nearby elements, and provides insight into the clustering structure. Ordinary two-way seriation is also extended to deal with updates of both the row and column spaces. Combined with the convolution kernel, a three-dimensional visualization of dynamics is demonstrated on two data sets, a news collection and a set of microarray measurements.

show abstract

“…to discover these hidden structures, we perform hierarchical clustering (single-linkage and complete-linkage) and reorder the correlation matrix using the ordering suggested by the dendrogram. The reordering process is called seriation [125][126][127], and we see that the seriated correlation matrix in Fig. 5.2.2(c)∼(d) is more structured and we can identify a few clusters just through visual inspection.…”

Section: Hierarchical Clusteringmentioning

confidence: 99%

Macroscopic & mesoscopic dynamics of financial markets

Teh¹

View full text Add to dashboard Cite

With hundreds trillion dollars of capital floating in the stock market, it is extremely important to understand market structures and dynamics of stock markets. In this thesis, we studied the macroscopic and mesoscopic dynamics of financial markets, from the econophysics (a marriage between physics and economics) point of view. When econophysicists study stock markets, they frequently borrow methods developed in other areas of physics. However, because of the nature of their problems, econophysicists sometimes also invent new methods. In this thesis, we have also contributed methodological innovations (MI), to contrast the phenomenological discoveries (PD) that we have also made. The first of these methodological innovations is (MI1) the method of partial hierarchical clustering (PHC), a supervised clustering method with the advantage of using multiple thresholds to determine clusters. Through the PHC results, we demonstrated (PD1) the existence of hierarchical structures in the Singapore Exchange and Hong Kong Stock Exchange: from market sectors, to country markets, and to global markets. Furthermore, we also investigated the dynamics of these hierarchical structures across market crashes. To do this, we (MI2) extend the complete-linkage hierarchical clustering algorithm, to obtain robust clusters of stocks with high intra-cluster homogeneity and high inter-cluster heterogeneity. By visualizing these robust clusters using (MI3) the fusion-fission diagram, we observed that (PD2) when approaching market crashes, the movements of stock prices become synchronized, causing most of stocks to merge into a giant cluster. Right after the crash, this giant cluster fragmented and thereafter mixed strongly. This discovery points us to the fusion-fission processes in the market, which we can exploit to forecast market crashes. We assume that the traders strategies form strategy clusters in the constantly changing strategy space, and the stock price movements are governed by the dynamics of these strategy clusters. Moreover, this dynamic can be described by a statistical physics fusion-fission model: the soup-of-groups model (SoG). We (MI4) derived a mean-field SoG forecasting equation and showed that some market crashes can be predicted. Specifically, by fitting the continuous returns of the component stocks of the Straits Times Index to the SoG forecasting equation, we (PD3) found episodes of heightened crash likelihoods close to the Chinese Correction (27 Feb 2007), beginning of the Subprime Crisis (17 Aug 2007), and the Asian Correction (9 Mar 2009), with early warning four to six months prior to the crashes.

show abstract

Visual matrix explorer for collaborative seriation

Cited by 8 publications

References 19 publications

Demonstrating conceptual dynamics in an evolving text collection

Demonstrating conceptual dynamics in an evolving text collection

Two-way incremental seriation in the temporal domain with three-dimensional visualization: Making sense of evolving high-dimensional datasets

Macroscopic & mesoscopic dynamics of financial markets

Contact Info

Product

Resources

About