Abstract: A system and method for thematically grouping documents into clusters is provided. Concepts are extracted from a plurality of documents. The concepts include nouns or noun phrases. A number of occurrences for each concept are determined within each document. A bounded range is applied to the concepts and a subset of the concepts is selected by removing the concepts that fall outside the bounded range. The bounded range includes upper edge conditions and lower edge conditions. Themes are generated from the subset of concepts by identifying two or more concepts with common semantic meaning. Clusters of the documents are generated based on the themes.

Abstract: A system and method for clustering unstructured documents is provided. Documents having terms with frequencies of occurrence that satisfy upper and lower edge conditions are selected. Concepts are generated for the selected documents. The selected documents are grouped into clusters of the documents. A weight for each of the clusters is evaluated. A similarity value is determined from the frequencies of occurrence for at least one of the terms from the concepts and the cluster weights for each selected document. Each selected document is assigned into one such cluster based on the similarity value of the selected document.

Abstract: A system and method for reorienting clusters within a display is provided. Clusters are maintained within a display. Each cluster includes a center located at a distance relative to a common origin for the display and a radius measured from the center. A pair of the clusters is selected and a bounding region is determined for each cluster in the pair by forming a pair of tangent vectors about the cluster and originating at the common origin. The bounding regions of the clusters in the pair are compared. The distance from the common origin of one of the clusters in the pair is increased upon overlap of the bounding regions as a perspective-corrected distance, which is determined as a function of the distances, the radii, and an angle between tangent vectors. The one cluster is moved to reorient the cluster's center at the perspective-corrected distance in the display.

Abstract: A system and method for clustering unstructured documents is provided. Documents having terms with frequencies of occurrence that satisfy upper and lower edge conditions are selected. Concepts are generated for the selected documents. The selected documents are grouped into clusters of the documents. A weight for each of the clusters is evaluated. A similarity value is determined from the frequencies of occurrence for at least one of the terms from the concepts and the cluster weights for each selected document. Each selected document is assigned into one such cluster based on the similarity value of the selected document.

Abstract: A system and method for providing a perspective-corrected representation of a multi-dimensional cluster is presented. Clusters are displayed based on independent spatial orientations within multiple dimensions. For each such spatial orientation, the clusters are located at an independent distance from a common origin. A relationship between a pair of the clusters is picked and analyzed. A criteria for perspective correction is sought. The criteria is applied to the pair of clusters to determine whether the clusters need be reoriented. The clusters are reoriented in the display by a perspective-corrected distance based on the relationship between the pair of clusters.

Abstract: A system and method for dynamically evaluating latent concepts in unstructured documents is disclosed. A multiplicity of concepts are extracted from a set of unstructured documents into a lexicon. The lexicon uniquely identifies each concept and a frequency of occurrence. A frequency of occurrence representation is created for the documents set. The frequency representation provides an ordered corpus of the frequencies of occurrence of each concept. A subset of concepts is selected from the frequency of occurrence representation filtered against a pre-defined threshold. A group of weighted clusters of concepts selected from the concepts subset is generated. A matrix of best fit approximations is determined for each document weighted against each group of weighted clusters of concepts.

Abstract: A system and method for providing a perspective-corrected representation of a multi-dimensional cluster is presented. Clusters are displayed based on independent spatial orientations within multiple dimensions. For each such spatial orientation, the clusters are located at an independent distance from a common origin. A relationship between a pair of the clusters is picked and analyzed. A criteria for perspective correction is sought. The criteria is applied to the pair of clusters to determine whether the clusters need be reoriented. The clusters are reoriented in the display by a perspective-corrected distance based on the relationship between the pair of clusters.

Abstract: A pair of convex clusters are selected. Each convex cluster has a center of mass located at an original fixed distance from a common origin, and is oriented along a vector formed at a fixed angle from a common polar axis. A span is measured between the centers of mass of each convex cluster. A segment is measured from the center of mass of each convex shape to a point closest to the other convex shape along the span. A new fixed distance from the common origin for the center of mass for one of the convex clusters located along the vector for that convex cluster is evaluated if the span is less than the sum of the segments of the convex clusters. The pair of convex clusters are displayed rendered using at least the new fixed distance for the center of mass of the one convex cluster.

Abstract: A system and method for dynamically evaluating latent concepts in unstructured documents is disclosed. A multiplicity of concepts are extracted from a set of unstructured documents into a lexicon. The lexicon uniquely identifies each concept and a frequency of occurrence. A frequency of occurrence representation is created for the documents set. The frequency representation provides an ordered corpus of the frequencies of occurrence of each concept. A subset of concepts is selected from the frequency of occurrence representation filtered against a pre-defined threshold. A group of weighted clusters of concepts selected from the concepts subset is generated. A matrix of best fit approximations is determined for each document weighted against each group of weighted clusters of concepts.

Abstract: A system and method for dynamically evaluating latent concepts in unstructured documents is disclosed. A multiplicity of concepts are extracted from a set of unstructured documents into a lexicon. The lexicon uniquely identifies each concept and a frequency of occurrence. A frequency of occurrence representation is created for the documents set. The frequency representation provides an ordered corpus of the frequencies of occurrence of each concept. A subset of concepts is selected from the frequency of occurrence representation filtered against a pre-defined threshold. A group of weighted clusters of concepts selected from the concepts subset is generated. A matrix of best fit approximations is determined for each document weighted against each group of weighted clusters of concepts.

Abstract: A pair of convex clusters are selected. Each convex cluster has a center of mass located at an original fixed distance from a common origin, and is oriented along a vector formed at a fixed angle from a common polar axis. A span is measured between the centers of mass of each convex cluster. A segment is measured from the center of mass of each convex shape to a point closest to the other convex shape along the span. A new fixed distance from the common origin for the center of mass for one of the convex clusters located along the vector for that convex cluster is evaluated if the span is less than the sum of the segments of the convex clusters. The pair of convex clusters are displayed rendered using at least the new fixed distance for the center of mass of the one convex cluster.

Abstract: A system and method for generating a visualized data representation preserving independent variable geometric relationships is described. A pair of convex clusters are selected. Each convex cluster is rendered on a display. Each convex cluster has a center of mass located at an original fixed distance from a common origin, and is oriented along a vector formed at a fixed angle from a common polar axis. A span is measured between the centers of mass of each convex cluster. For each convex cluster, a segment is measured from the center of mass of each convex shape to a point closest to the other convex shape along the span. A new fixed distance from the common origin for the center of mass for one of the convex clusters located along the vector for that convex cluster is evaluated if the span is less than the sum of the segments of the convex clusters. The pair of convex clusters are displayed rendered using at least the new fixed distance for the center of mass of the one convex cluster.

Abstract: A system and method for efficiently generating cluster groupings in a multi-dimensional concept space is described. A plurality of terms is extracted from each document in a collection of stored unstructured documents. A concept space is built over the document collection. Terms substantially correlated between a plurality of documents within the document collection are identified. Each correlated term is expressed as a vector mapped along an angle ? originating from a common axis in the concept space. A difference between the angle ? for each document and an angle ? for each cluster within the concept space is determined. Each such cluster is populated with those documents having such difference between the angle ? for each such document and the angle ? for each such cluster falling within a predetermined variance.

Abstract: A system and method for efficiently generating cluster groupings in a multi-dimensional concept space is described. A plurality of terms are extracted from each document in a collection of stored unstructured documents. A concept space is built over the document collection. Terms substantially correlated between a plurality of documents within the document collection are identified. Each correlated term is expressed as a vector mapped along an angle &thgr; originating from a common axis in the concept space. A difference between the angle &thgr; for each document and an angle &sgr; for each cluster within the concept space is determined. Each such cluster is populated with those documents having such difference between the angle &thgr; for each such document and the angle &sgr; for each such cluster falling within a predetermined variance.