wolfmd / akashicrecords

AkashicRecords

Information Perspective Networks

Overviw

Akashic Records is a unified theory for interactions between the beliefs of individuals.

Problem

The vast network of individuals connected through the Internet has expanded beyond the understanding of any one organization or encyclopedic tool. The combination of a wide variety of opinions and textual content produces patterns and sociological phenomena which have substantial implications for the study of biases and human interaction. This project seeks to move toward an understanding and reconstruction of these relationships as well as their impact on modern social constructs.

Name

The term "Akashic Records" is the theoretical corpus of all human events, thoughts, and emotions that is believed to exist on a non-physical plane of existence. This concept could be seen to be forming in the ethereal data generated through internet communication and interconnection. As more data becomes available on the internet, a clearer image of the whole of humanity with its various sentiments, biases, and intents becomes evident.

Approach

This project utilizes the IBM Watson API which provides sentence-level sentiment analysis, tone analysis, and the deconstruction of statements into more digestible topic units. We look at various frames created by digital entities and compare the sentiment across a selection of events and topical figures.

Fundamental Questions

• Does an entity's background knowledge influence their interpretation of a frame?
• How do entity artifacts impact the global interpretation of frames?

Hypotheses of Emotion Detection

* The more imaginative, open to change, or emotionally aware a tweet seemed, the more likely it was to be liked.

* The more organized, thoughtful, or thorough a tweet seemed, the more likely it was to be retweeted.

* The angrier, stressed, or impulsive a tweet seemed, the less likely it was to be retweeted.

Definitions

Frame

A frame is a system of components related in such a way that in order to understand any one component, one must understand all components in the system. This applies particularly well to textual documents on the internet which interrelated through analyzing the same event or set of topics with a particular mindset or set of values.

• Frames are made up of a coherent set of component subject -> verb -> objects (SVO's).

• Elements are inside of components which are within sentences. A sentence can be made up of one or more components.

• Frame

  • Components (SVO's)
    • Elements
      • Entities
        • Extracts people, companies, organizations, cities, geographic features, and other entities from your content, and optionally detects the sentiment of each entity.
      • Keywords
        • Determines important keywords in the text, ranks them by relevance, and optionally detects the sentiment of each keyword.
      • Concepts
        • Identifies general concepts that aren’t necessarily directly referenced in the text. For example, analysis of an article about BMW, Audi, and Porsche is likely to return the "Automotive industry” concept.
    • Mood
      • Tone
        • Emotion
          • Joy, Happy, Satisfied
          • Fear, Scared, Paranoid
          • Sad, Quiet, Withdrawn
          • Disgust, Offensive, Unpleasent
          • Anger, Hostile, Conflicting
        • Social Tendencies
          • Openness
          • Organized
          • Extraverted
          • Agreeable
          • Neurotic
        • Language
          • Analytical: Understandable, Believable
      • Sentiment
        • Positive
        • Negative