Browse Prior Art Database

Topic Driven Information Spread Analysis of Social Media

IP.com Disclosure Number: IPCOM000245384D
Publication Date: 2016-Mar-04
Document File: 4 page(s) / 145K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a system to process social media data in order to produce an understanding of the way specific types of information spread through the platform. The system performs an analysis of social media posts and other content as related to a cluster of opinions/beliefs, followed by an analysis of topographical webs to determine how different kinds of content spread.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 4

Topic Driven Information Spread Analysis of Social Media

To successfully introduce products, technologies, and ideas into the marketplace, companies utilize social media. For example, the company can examine the presence of products in social media discussions; however, the company is still not able to recognize the best marketing methods without making assumptions about the audience. The insight into the effects of a marketing campaign typically comes after the campaign is started. Current technologies target stereotypes and make broad generalizations of target groups. Personalized information given to the individuals most capable of eliciting change can only obtained after the effects of a campaign; a large part of social media possesses untapped potential.

The novel contribution is a system to process social media data in order to produce an understanding of the way specific types of information spread through the platform. With this understanding, analysts can identify the individuals within the platform, the individuals' respective roles relative to the content, and the individuals' effectiveness in these roles. The solution comprises two novel analytical components.

The first component is an analysis of posts and other content on social media as related to a cluster of opinions/beliefs. This generalizes similar content and analyzes how like content spreads through the same network as a topographical web over time. This allows users to identify how individual nodes participate in a dialogue as related to a topic. Points on each map are categorized as generators, propagators, and consumers.

The second component is an analysis each of these topographical webs to determine how different kinds of content spread. The process looks at which nodes are active for particular clusters, and examine what roles those nodes play with specific cluster information. The insight provided enables comparison between opinions/beliefs and the spread of those opinions/beliefs through social media. More significantly, the analysis provides the ability to target marketing material to the individuals in a network

who are the generators and propagators. This insight encourages the spread of information through the system by accelerating the pace at which a topographical web matures.

Figure 1: Components and process flow

The system captures data from a social media platform in terms of time and content. It then applies Natural Language Processing (NLP) to the posts and a set of clusters relates like content. Producing these clusters offers an understanding of the relationships between ideas, opinions, and beliefs as ideas spread through social media over time. An understanding of how certain clusters can be topographically

represented provides insight into individual behavior and ways to optimize the spread of

1


Page 02 of 4

information. Ultimately, making topographical data available by cluster alongside node properties...