Computer implemented method and system for integrative causal modeling and transfer
A hypergraph user interface for interaction with a system for integrative causal modeling from natural language text and including a natural language text input element configured to receive a designation of a natural language text file or such a file. The input element causes the designation or natural language text file to be communicated to at least one natural language processing module. The hypergraph user interface further includes a causal hypergraph visualization element representative of directed hypergraph data generated from the at least one natural language processing module. Selection of a hypergraph element of the causal hypergraph representation causes the hypergraph user interface to communicate an indication of the hypergraph element to a querying module and generate an informative element indicative of at least one of a causal link, a causal statement, or a sentence of the at least one natural text file on which the hypergraph element is based.
1. A hypergraph user interface for interaction with a system for integrative causal modeling from natural language text, the hypergraph user interface comprising:
a natural language text input element configured to receive at least one of a designation of a natural language text file or a natural language text file and, in response to the received at least one designation or natural language text file, cause the at least one designation or natural language text file to be communicated to at least one natural language processing module of the system; and
a causal hypergraph visualization element representative of directed hypergraph data generated from the at least one natural language processing module of the system,
wherein, in response to selection of a hypergraph element of the causal hypergraph representation, the hypergraph user interface is configured to communicate an indication of the hypergraph element to a querying module of the system and generate, in association with the causal hypergraph visualization, an informative element indicative of at least one of a causal link, a causal statement, or a sentence of the at least one natural text file on which the hypergraph element is based.
2. The hypergraph user interface of claim 1 , wherein, when the directed hypergraph data includes multivariable hypergraph data, at least one hypergraph element of the casual hypergraph visualization indicates multivariable, directed causality.
3. The hypergraph user interface of claim 1 , wherein the at least one natural language processing module of the system includes at least one module suitable for transfer learning.
4. The hypergraph user interface of claim 1 , wherein, in response to the selection of the hypergraph element, the hypergraph user interface is configured to reorient the casual hypergraph visualization element to emphasize the selected hypergraph element.
5. The hypergraph user interface of claim 1 , wherein the informative element includes at least one of a table or chart.
6. The hypergraph user interface of claim 1 , wherein the causal hypergraph visualization element includes a plurality of nodes associated with a plurality of taxonomy categories.
7. The hypergraph user interface of claim 6 , wherein the causal hypergraph visualization element includes a plurality of directed edges connecting two or more nodes of the plurality of nodes, wherein each directed edge includes an indicator indicative of at least one of a number of causal statements associated with the directed edge, the strength of a causal association of at least one causal statement associated with the directed edge, or a strength of the association between at least one of the connected nodes and at least one casual statement associated with the directed edge.
8. The hypergraph user interface of claim 6 , wherein the causal hypergraph visualization element includes a first-order nodal diagram, wherein the first-order nodal diagram includes a performance node representing the nodes of the directed hypergraph data associated with the taxonomy categories and a non-performance node representing one or more portions of causal statements not associated with at least one of the taxonomy categories.
9. The hypergraph user interface of claim 8 , wherein the hypergraph user interface is configured to generate, in response to a selection of one of the performance node or the non-performance node, a second-order causal hypergraph visualization element including nodes associated with the selected node of the first-order nodal diagram.
10. The hypergraph user interface of claim 9 , wherein the second-order causal hypergraph visualization element includes a plurality of edges, each edge connecting at least one node associated with a taxonomy category to at least one other node, the plurality of edges indicating multivariable, directed hypergraph data.
11. The hypergraph user interface of claim 1 , wherein the hypergraph user interface is configured such that at least one of a layout of the hypergraph user interface including at least one of the causal hypergraph visualization element or the informative element or a user-initiated change in the at least one of the causal hypergraph visualization element or the informative element is provided as visualization training data to a visualization module configured to generate the causal hypergraph visualization element or the informative element.
12. The hypergraph user interface of claim 11 , wherein the visualization module comprises a deep learning network, and wherein the visualization training data is utilized to train the deep learning network.
13. A system for integrative causal modeling from natural language text, the system comprising:
a causal identification module comprising instructions stored in at least one memory and executable by one or more processors to cause the causal identification module to identify, from at least one natural language text file, a plurality of causal links utilizing natural language processing, each causal link including a causal portion and an output portion;
a normalization module comprising instructions stored in at least one memory and executable by one or more processors to cause the normalization module to identify whether each causal portion and each output portion of each causal link is associated with a taxonomy category of a plurality of predefined taxonomy categories; and
a visualization and querying module comprising instructions stored in at least one memory and executable by one or more processors to cause the visualization and querying module to generate a causal hypergraph visualization utilizing interface hardware associated with the system, the causal hypergraph visualization generated from directed hypergraph data representing the causal statements and associations with the predefined taxonomy categories of the at least one natural language text file.
14. The system of claim 13 , wherein the visualization and querying module further comprising instructions stored in the at least one memory and executable by the one or more processors to cause the visualization and querying module to:
receive a selection of an element of the causal hypergraph visualization; and
generate, in response to the selected element of the causal hypergraph visualization, at least one of a causal link indicated by the selected element, at least one causal statement on which the at least one causal link is based, at least one sentence of the at least one natural language text file on which the at least one causal link is based, or a combination of the preceding.
15. The system of claim 13 , wherein the causal identification module further comprising instructions stored in the at least one memory and executable by the one or more processors to cause the causal identification module to:
receive the at least one natural language text file; and
identify a plurality of causal statements included in the at least one natural language text file utilizing natural language processing,
wherein the casual links are identified from the identified plurality of causal statements.
16. The system of claim 13 , wherein the visualization and querying module further comprising instructions stored in the at least one memory and executable by the one or more processors to cause the visualization and querying module to:
generate the causal hypergraph visualization utilizing at least one of a haptic interface device, a virtual reality interface device, or a multimodal interface device; or
receive a selection of an element of the causal hypergraph visualization communicated from at least one of the haptic interface device, the virtual reality interface device, or the multimodal interface device.
17. The system of claim 13 , wherein the directed hypergraph data comprises multivariable, directed hypergraph data.
18. The system of claim 13 , further comprising a hypergraph data generation module comprising instructions stored in at least one memory and executable by one or more processors to cause the hypergraph data generation module to:
generate the directed hypergraph data based on the identified plurality of causal links and the associations of the causal links with the plurality of predefined taxonomy categories.
19. The system of claim 13 , further comprising a translation module communicatively coupled to the visualization and querying module, the at least one natural language text file, and at least one data file including the directed hypergraph data and data indicative of the casual links normalized with respect to the predefined taxonomy categories.
20. The system of claim 13 , wherein each of the causal identification module and the normalization module includes at least one module suitable for transfer learning.