Method and system for information retrieval and aggregation from inferred user reasoning
The visual analytic system enables information retrieval within large text collections. Typically, users have to directly and explicitly query information to retrieve it. With this system and process, the reasoning of the user is inferred from the user interaction they perform in a visual analytic tool, and the appropriate information to query, process, and visualize is systematically determined.
1. A data visualization and interaction system comprising:
a. a processor;
b. a set of eight display devices configured for displaying a spatial layout for implementing multiple levels of aggregation of information within a single spatialization;
c. an interaction component, wherein the system utilizes user interactions from which the system infers analytical reasoning and steers an underlying analytic model, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
d. a generation component, using the processor, configured for generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information, wherein the system adjusts distribution among the multiple levels of aggregation based on inferred topics of interest calculated from the user interactions performed.
2. The system of claim 1 wherein the multiple levels of aggregation include terms, phrases, snippets, and full documents.
3. The system of claim 1 wherein the distance between terms in the spatial layout represents relative similarity between the terms.
4. The system of claim 1 wherein the user interactions comprise semantic interactions that implement direct manipulation of information within the spatialization.
5. The system of claim 1 wherein the multiple levels of aggregation are shown to a user within the same spatial layout through calculating saliency of the information via a similarity matrix and weighting schema.
6. The system of claim 5 wherein each level of aggregation is assembled into a hierarchy based on a sum of the weights of its entities.
7. The system of claim 5 wherein the calculated salient information is loaded into the client and rendered in the visualization.
8. The system of claim 1 wherein user interactions include moving terms, resizing terms, pinning terms, deleting terms, and text searching for terms or phrases.
9. A method of interactive visual exploration of information comprising:
a. implementing, using a processor, multiple levels of aggregation of information within a single spatialization on a set of eight display devices;
b. implementing user interactions from which analytical reasoning is inferred, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset;
c. generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information; and
d. adjusting distribution among the multiple levels of aggregation based on inferred topics of interest calculated from the user interactions performed.
10. The method of claim 9 further comprising steering an underlying analytic model.
11. The method of claim 9 wherein the multiple levels of aggregation include terms, phrases, snippets, and full documents.
12. The method of claim 9 the distance between terms in the spatialization represents relative similarity between the terms.
13. The method of claim 9 wherein the user interactions comprise semantic interactions that implement direct manipulation of information within the spatialization.
14. The method of claim 9 wherein the multiple levels of aggregation are shown to a user within the same spatialization through calculating saliency of the information via a similarity matrix and weighting schema.
15. The method of claim 14 wherein each level of aggregation is assembled into a hierarchy based on a sum of the weights of its entities.
16. The method of claim 14 wherein the calculated salient information is loaded into the client and rendered in the visualization.
17. The method of claim 9 wherein user interactions include moving terms, resizing terms, pinning terms, deleting terms, and text searching for terms or phrases.
18. An apparatus comprising:
a. a non-transitory memory for storing an application, the application for:
i. implementing multiple levels of aggregation of information within a single spatialization;
ii. implementing user interactions from which analytical reasoning is inferred, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
iii. generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information; and
iv. adjusting distribution among the multiple levels of aggregation based on inferred topics of interest calculated from the user interactions performed; and
b. a processor coupled to the non-transitory memory, the processor configured for processing the application.
19. The apparatus of claim 18 further comprising steering an underlying analytic model.
20. The apparatus of claim 18 wherein the multiple levels of aggregation include terms, phrases, snippets, and full documents.
21. The apparatus of claim 18 the distance between terms in the spatialization represents relative similarity between the terms.
22. The apparatus of claim 18 wherein the user interactions comprise semantic interactions that implement direct manipulation of information within the spatialization.
23. The apparatus of claim 18 wherein the multiple levels of aggregation are shown to a user within the same spatialization through calculating saliency of the information via a similarity matrix and weighting schema.
24. The apparatus of claim 23 wherein each level of aggregation is assembled into a hierarchy based on a sum of the weights of its entities.
25. The apparatus of claim 23 wherein the calculated salient information is loaded into the client and rendered in the visualization.
26. The apparatus of claim 18 wherein user interactions include moving terms, resizing terms, pinning terms, deleting terms, and text searching for terms or phrases.
27. The apparatus of claim 18 wherein the information includes an entity-by-entity similarity matrix, wherein each node is an entity and a weight of the edges between nodes is determined by a similarity score from the entity-by-entity similarity matrix, further wherein each entity is assigned a snippet, computed using a Lucene index to search for a short subset of a document that has a density of terms above a density threshold that have a similarity score above a similarity threshold.
28. A data visualization and interaction system comprising:
a. a processor;
b. a set of eight display devices configured for displaying a spatial layout for implementing multiple levels of aggregation of information within a single spatialization;
c. an interaction component, wherein the system utilizes user interactions from which the system infers analytical reasoning and steers an underlying analytic model, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
d. a generation component, using the processor, configured for generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information, wherein the multiple levels of aggregation are shown to the user within the same spatial layout through calculating saliency of the information via a similarity matrix and weighting schema.
29. A method of interactive visual exploration of information comprising:
a. implementing, using a processor, multiple levels of aggregation of information within a single spatialization on a set of eight display devices;
b. implementing user interactions from which analytical reasoning is inferred, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
c. generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information, wherein the multiple levels of aggregation are shown to the user within the same spatial layout through calculating saliency of the information via a similarity matrix and weighting schema.
30. An apparatus comprising:
a. a non-transitory memory for storing an application, the application for:
i. implementing multiple levels of aggregation of information within a single spatialization;
ii. implementing user interactions from which analytical reasoning is inferred, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
iii. generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information, wherein the multiple levels of aggregation are shown to the user within the same spatial layout through calculating saliency of the information via a similarity matrix and weighting schema; and
b. a processor coupled to the non-transitory memory, the processor configured for processing the application.
31. A data visualization and interaction system comprising:
a. a processor;
b. a set of eight display devices configured for displaying a spatial layout for implementing multiple levels of aggregation of information within a single spatialization;
c. an interaction component, wherein the system utilizes user interactions from which the system infers analytical reasoning and steers an underlying analytic model, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
d. a generation component, using the processor, configured for generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information, wherein the user interactions include moving terms, resizing terms, pinning terms, deleting terms, and text searching for terms or phrases.
32. A method of interactive visual exploration of information comprising:
a. implementing, using a processor, multiple levels of aggregation of information within a single spatialization on a set of eight display devices;
b. implementing user interactions from which analytical reasoning is inferred, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
c. generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information, wherein the user interactions include moving terms, resizing terms, pinning terms, deleting terms, and text searching for terms or phrases.
33. An apparatus comprising:
a. a non-transitory memory for storing an application, the application for:
i. implementing multiple levels of aggregation of information within a single spatialization;
ii. implementing user interactions from which analytical reasoning is inferred, wherein user reasoning on a small subset of the information is extrapolated into a larger dataset; and
iii. generating a graphical representation based on the information, wherein the graphical representation represents each data object as a node with edges representing similarities between each data object, wherein the edges are invisible, further wherein a user reveals a hierarchical topic relationship by hovering over the information in the multiple levels of aggregation of information, wherein the user interactions include moving terms, resizing terms, pinning terms, deleting terms, and text searching for terms or phrases; and
b. a processor coupled to the non-transitory memory, the processor configured for processing the application.