US 2017/0371970 A1Published application

Automated storyline content selection and qualitative linking based on context

AssigneeInternational Business Machines Corporation

InventorsHarish Bharti, Sanjib Choudhury, Kshitij K. Raval, Rajesh K. Saxena

Life of this patent

Estimating the term…

Loading patent…

Abstract

A huge volume of unstructured content is available on the internet. Social media websites, news outlets, subject matter expert sites, forums, government organization sites, non-government organization sites, etc., collectively provide a rich source of raw material for any kind of story writing, for example, for movies, novels, television, etc. In some embodiments of the present invention, content is intelligently searched from diverse sources. Embodiments of the present invention make use of such unstructured content, to provide raw material upon which to base a cohesive and appealing story, in part by applying graphing theory to: (i) represent content gathered in the search as a graph, with each element of content assigned to a node of the graph; (ii) qualitatively link the nodes; and/or (iii) identify important nodes which potentially become central to a storyline.

Claims16

Dependent claims

A computer-implemented method of creating an automated story, the method comprising:

receiving a plurality of content data sets with each content data set including network addressable content useful in creating narrative content for the automated story;
creating a story data graph data structure stored in a graph database, with the creation of the graph data structure including:
defining a plurality of nodes, with each node of the plurality of nodes respectively corresponding to a content data set of the plurality of content data sets, and with each node including the content data of the respectively corresponding content data set and contextual metadata of the respectively corresponding to attributes of a context of the corresponding context data set, and
defining a plurality of connections among and between the nodes based upon the contextual metadata of the nodes;
for each given node of the plurality of nodes, determining an aggregate connectedness value based upon the connections in which the given node is involved;
identifying, by a processor set, a subset of recommended candidate nodes of the plurality of nodes based, at least in part, upon the aggregate connectedness values of the nodes of the plurality of nodes;
and
recommending the recommended candidate nodes of the subset of recommended candidate nodes for use in the automated story.

5 more elements

A computer program product for creating an automated story, the computer program product comprising a computer readable storage medium having stored thereon:

first program instructions programmed to receive a plurality of content data sets with each content data set including network addressable content useful in creating narrative content for the automated story;
second program instructions programmed to create a story data graph data structure stored in a graph database, with the creation of the graph data structure including:
third program instructions programmed to define a plurality of nodes, with each node of the plurality of nodes respectively corresponding to a content data set of the plurality of content data sets, and with each node including the content data of the respectively corresponding content data set and contextual metadata of the respectively corresponding to attributes of a context of the corresponding context data set, and
fourth program instructions programmed to define a plurality of connections among and between the nodes based upon the contextual metadata of the nodes;
for each given node of the plurality of nodes, third program instructions programmed to determine an aggregate connectedness value based upon the connections in which the given node is involved;
fifth program instructions programmed to identify, by a processor set, a subset of recommended candidate nodes of the plurality of nodes based, at least in part, upon the aggregate connectedness values of the nodes of the plurality of nodes;
and
sixth program instructions programmed to recommend the recommended candidate nodes of the subset of recommended candidate nodes for use in the automated story.

5 more elements

Dependencies are inferred from the claim text.