Exemplar-based Visualization of Large Document Corpus (InfoVis2009-1115) Export

@article{citeulike:5998168, abstract = {With the rapid growth of the World Wide Web and electronic information services,text corpus is becoming available on-line at an incredible rate.By displaying text data in a logical layout (e.g., color graphs),text visualization presents a direct way to observe the documentsas well as understand the relationship between them.In this paper, we propose a novel technique, Exemplar-based Visualization (EV), to visualizean extremely large text corpus. Capitalizing on recent advances in matrixapproximation and decomposition, EV presents a probabilistic multidimensional projection modelin the low-rank text subspace with a sound objective function. The probability of each document proportion to the topics is obtained through iterative optimization andembedded to a low dimensional space using parameter embedding.By selecting the representative exemplars, we obtain a compactapproximation of the data. This makes the visualization highly efficient and flexible. In addition, the selected exemplars neatly summarize the entire data set and greatly reduce the cognitiveoverload in the visualization, leading to an easier interpretation oflarge text corpus. Empirically, we demonstrate the superior performance of EVthrough extensive experiments performed on the publicly available text data sets.}, author = {Chen, Yanhua and Wang, Lijun and Dong, Ming and Hua, Jing}, citeulike-article-id = {5998168}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/TVCG.2009.140}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=5290725}, day = {23}, doi = {10.1109/TVCG.2009.140}, journal = {Visualization and Computer Graphics, IEEE Transactions on}, keywords = {infovis, text, visualization}, month = {October}, number = {6}, pages = {1161--1168}, posted-at = {2009-10-23 16:52:42}, priority = {2}, title = {Exemplar-based Visualization of Large Document Corpus (InfoVis2009-1115)}, url = {http://dx.doi.org/10.1109/TVCG.2009.140}, volume = {15}, year = {2009} }

TY - JOUR ID - citeulike:5998168 L3 - citeulike-article-id:5998168 N2 - With the rapid growth of the World Wide Web and electronic information services,text corpus is becoming available on-line at an incredible rate.By displaying text data in a logical layout (e.g., color graphs),text visualization presents a direct way to observe the documentsas well as understand the relationship between them.In this paper, we propose a novel technique, Exemplar-based Visualization (EV), to visualizean extremely large text corpus. Capitalizing on recent advances in matrixapproximation and decomposition, EV presents a probabilistic multidimensional projection modelin the low-rank text subspace with a sound objective function. The probability of each document proportion to the topics is obtained through iterative optimization andembedded to a low dimensional space using parameter embedding.By selecting the representative exemplars, we obtain a compactapproximation of the data. This makes the visualization highly efficient and flexible. In addition, the selected exemplars neatly summarize the entire data set and greatly reduce the cognitiveoverload in the visualization, leading to an easier interpretation oflarge text corpus. Empirically, we demonstrate the superior performance of EVthrough extensive experiments performed on the publicly available text data sets. IS - 6 JF - Visualization and Computer Graphics, IEEE Transactions on EP - 1168 TI - Exemplar-based Visualization of Large Document Corpus (InfoVis2009-1115) VL - 15 SP - 1161 KW - infovis KW - text KW - visualization AU - Chen, Yanhua AU - Wang, Lijun AU - Dong, Ming AU - Hua, Jing PY - 2009/10/23/ UR - http://dx.doi.org/10.1109/TVCG.2009.140 ER -