Media-based computational influencer network analysis

1. A computer-implemented method for a highly automated media analysis of influencer networks, comprising:

defining, by an application of a computerized selection process of a computer that considers user-entered criteria, one or more scopes of media content to be included in an analysis project;

retrieving, by the computer, relevant media content from a plurality of providers as defined by the one or more scopes of the analysis project, the media content including published text articles and a body of text for each of the text articles;

extracting, by an application of an automatic computerized, linguistic-based and statistically-supported entity extraction process, entities from the text articles, the entities being data including names of people, organizations, locations, and brands recited in the body of text of the text articles in the retrieved media content;

manually associating, selectively and with the aid of the computer, for each of the entities, a functional role of each of the entities for each text article of the retrieved media content;

manually associating, selectively and with the aid of the computer, for each of the entities, a favorability score for each of the entities for each text article of the received media content;

storing the entities in a relational database, where the entities that are co-cited in a body of text of the text articles are linked to each other for an associated text article;

performing, by the computer, a first computation characterizing a network of influence relationships between each of the entities and each text article of the retrieved media content based on the extracted information and the manually associated functional role and favorability of each of the entities; and

performing, by the computer, a second computation characterizing connection properties of individual entities with respect to the other entities;

performing, by the computer, a third computation characterizing connection properties of at least a portion of the overall network of influence, the third computation to include a value for the network's cohesion property and density property; and

outputting a result of the first, second, and third computations to produce a graphical, interactive representation of the network of influence in which a user may select individual entities to examine their associated connection properties, link to other documents, and link to web pages related to the entities, and combinations thereof.

2. The computer-implemented method of claim 1 , wherein said entity extraction process can handle a very high volume media content data.

3. The computer-implemented method of claim 1 , wherein the third computation includes determining an Influence Influence ranking property for each of the entities.

4. The computer-implemented method of claim 1 , wherein the manually associating of the functional role and the favorability score for each of the entities for each text article of the retrieved media content is based on a reading of each text article by a human user.

5. The computer-implemented method of claim 1 , wherein the defining of the one or more scopes of media content to be included in the analysis project includes defining a first duration of time and a second duration of time, and the retrieving of media content, the extracting of information, the manually associating, the first, second, and third computing of a characterization, and the outputting of a result of the computation to provide the indication of influencers associated with the network is performed for each of the first and the second durations of time.

6. The computer-implemented method of claim 1 , wherein the favorability score associated with each of the entities for each text article of the retrieved media content is determined independently for each text article.

7. The computer-implemented method of claim 1 , wherein the text articles are not limited to a specific format or structure.

8. The computer-implemented method of claim 1 , further comprising reformatting the retrieved media content for further processing by the method.

9. The computer-implemented method of claim 1 , wherein a series of network plots for a sequence of subperiods are produced, which can be visualized consecutively, to produce an effect of showing the periodic changes in the network.

10. The computer-implemented method of claim 1 , further comprising outputting a result of the first, second, and third computations to produce a series of graphical representations of the network of influence representing inter-relationships between the entities for several chronological sub-periods, the images of which can be presented consecutively as a movie to visualize a transformation over a period of time.

11. The method of claim 1 , wherein the retrieving is performed by a streaming process.

12. The computer-implemented method of claim 1 , further comprising an automatic computerized process to associate the functional role and the favorability score for each of the entities for each text article of the retrieved media content.

13. The computer-implemented method of claim 1 , further comprising extracting values for structured fields associated with each text article of the retrieved media content, including values associated with at least one of the structured fields of: an author, a title, a date of publication, and a name of a publication associated with each text article.