Systems and methods for the combination and display of social and textual content
View Patent ↗Techniques are presented for determining a corpus of content portions, each content portion associated with at least one element. A first set of feature values is determined for each content portion. Clusters of content portions are then determined based on the first set of feature values. The features values are optionally associated with topics. Structural links between the elements are determined based on a second set of feature values. A layout of the element is then determined based on the clusters and the structural links. Optionally the N-most dominant topics are determined and also used to inform the layout of the elements in a display.
1. A computer-implemented method for the combination and display of social and textual content comprising the steps of:
determining a corpus of content portions;
determining elements associated with the content portions;
determining a first set of feature values based on the determined elements wherein the first set of feature values comprise at least one of:
textual elements drawn from the content potions, and topics derived from the content portions;
determining content indicators associated with the content portions;
determining clusters of content portions based on the first set of feature values;
determining structural links between the content portion indicators based on a second set of feature values: wherein the second set of feature values indicate a type and a strength of inter-relationships between the content portions associated with the content portion indicators;
determining a mapping of the clusters of content portions onto the dimensions of a display device that preserve the pair-wise distances between clusters;
determining a layout of the content portion indicators based on the structural links derived from the second set of feature values and the mapping of clusters derived from the first set of feature values; and
displaying a graph of the content portion indicators based on the layout.
2. The method of claim 1 , wherein the determining a mapping is based on multi-dimensional scaling.
3. The method of claim 2 , wherein the multi-dimensional scaling is based on at least one of: a Kohonen map and a self-organizing map, such that the clustered content portions are transformed from N-dimensions to a 2-dimensional display space.
4. The method of claim 1 , in which the clusters are determined based on a similarity metric.
5. The method of claim 4 , in which the similarity metric is at least one of: Euclidian distance, and cosine distance.
6. The method of claim 1 , in which the clustering is based on at least one of: Agglomerative clustering and k-means.
7. The method of claim 1 , in which the texts are digital texts.
8. The method of claim 7 , in which the digital texts are at least one of: messages, computer code, author identified papers and web pages.
9. The method of claim 1 wherein the determined clusters contain at least two distinct content portions.
10. The computer-implemented method of claim 1 , wherein the structural links indicate the strength of inter-relationships between the content portions by varying the width of the edge representing structural link on the displayed graph.
11. A system for combining and displaying social and textual content comprising:
an input/output circuit for determining a corpus of content portions;
a microprocessor to execute the following functions:
determine a corpus of content portions;
determine elements associated with the content portions;
determine a first set of feature values based on the determined elements, wherein the first set of feature values comprise at least one of:
textual elements drawn from the content portions and topics derived from the content portions;
determine content indicators associated with the content portions;
determine clusters of content portions based on the first set of feature values;
determine structural links between the content portion indicators based on a second set of feature values: wherein the second set of feature values indicate a type and a strength of inter-relationships between the content portions associated with the content portion indicators;
determine a mapping of the clusters of content portions onto the dimensions of a display device that preserve the pair-wise distances between clusters;
determine a layout of the content portion indicators based on the structural links derived from the second set of feature values and the mapping of clusters derived from the first set of feature values; and
display a graph of the content portion indicators based on the layout.
12. The system of claim 1 , wherein the determining a mapping is based on multi-dimensional scaling.
13. The system of claim 12 , in which the multi-dimensional scaling is based on at least one of: a Kohonen map and a self-organizing map, such that the clustered content portions are transformed from N-dimensions to a 2-dimensional display device.
14. The system of claim , in which the clusters are determined based on a similarity metric.
15. The system of claim 14 , in which the similarity metric is at least one of: Euclidean distance, and cosine distance.
16. The system of claim 1 , in which the clustering is based on at least one of: Agglomerative clustering and k-means.
17. The system of claim , in which the texts are digital texts.
18. The system of claim 17 , in which the digital texts are at least one of: messages, computer code, author identified papers and web pages.
19. The system of claim 1 , wherein the clusters formed by the cluster determination circuit contain at least two distinct content portions.
20. The system of claim 1 ,
wherein the first set of feature values are based on at least one of:
word stemming,
lemmatization, and
text transformations; and
wherein the second set of feature values is based on at least one of:
inter-relationships between the content portion indicators.
21. A non-transient computer-readable storage medium having computer executable instructions for causing a computer to perform the steps of:
determining a corpus of content portions;
determining elements associated with the content portions;
determining a first set of feature values based on the determined elements wherein the first set of feature values comprise at least one of:
textual elements drawn from the content potions, and topics derived from the content portions;
determining content indicators associated with the content portions;
determining clusters of content portions based on the first set of feature values;
determining structural links between the content portion indicators based on a second set of feature values: wherein the second set of feature values indicate a type and a strength of inter-relationships between the content portions associated with the content portion indicators;
determining a mapping of the clusters of content portions onto the dimensions of a display device that preserve the pair-wise distances between clusters;
determining a layout of the content portion indicators based on the structural links derived from the second set of feature values and the mapping of clusters derived from the first set of feature values; and
displaying a graph of the content portion indicators based on the layout.