Skip to content

Instantly share code, notes, and snippets.

Show Gist options
  • Select an option

  • Save thundergolfer/cc5698ee212afb1ae96af927b80665fe to your computer and use it in GitHub Desktop.

Select an option

Save thundergolfer/cc5698ee212afb1ae96af927b80665fe to your computer and use it in GitHub Desktop.

Me On Reddit.com / [Your Username]


Some Reddit users are fantastic and dedicated contributors to subreddits, but don't have a way to easily communicate it. Once you connect your profile on meonreddit.com, the site will display your profile with only professionally oriented content.

Here's an example for me. I'm no guru.

example meonreddit profile

How It Works

coming-soon

About The author

I'm a software engineering student from Melbourne, and I'm looking for a great internship anywhere on the globe. Here's my LinkedIn Profile.


Am I In A Bubble?

Most people don't realise they are in an information bubble, or 'echo-chamber'. Are you?

http://www.sciencedirect.com/science/article/pii/S0959378015000369 - Network analysis reveals open forums and echo chambers in social media discussions of climate change

https://www.facebook.com/notes/facebook-data-team/rethinking-information-diversity-in-networks/10150503499618859 - Rethinking Information Diversity in Networks

http://gradworks.umi.com/36/85/3685237.html - Social Structure and Information Diversity in Online Social Networks


If I Lived Forever

How long would it take to do everything you'd ever want to do?

The Scraping Code

If you want to learn from some basic web-scraping code in python check out the scripts in scraping_scripts.

The Post

coming soon


neuroscience-rundown-for-AI-tech

A short(ish) primer on neuroscience and the brain for those trying to build a digital genius.

How strong is the analogy of the human brain as a computer?

todo

Brain processing speed vs. computer processing speed

todo

Brain memory capacity vs. computer memory capacity

todo

Research Resources For This Page

  1. Computational Neuroscience for Advancing Artificial Intelligence: Models, Methods and Applications
  2. What Can AI Get From Neuroscience?, http://www.neurolab.gatech.edu/wp/wp-content/uploads/potter/publications/Potter-NeuroscienceForAIchapter.pdf

THE WAKEUP SHOCK WATCH

Currently In Development!

Intention

When I don't really have to get out of bed, I likely won't get out of bed. I really should get out of bed though, because 8 hours is enough sleep and there's so much to do in life that 16 hour days barely cut it, let alone 14.

I've always had in the back of my mind something that would definitely get me out of bed. The wakeup shock watch is that something.

When finished it will include the following features:

  1. Electric shock delivered to wrist if wearer remains in bed past alarm
  2. Electronic Locking so you can't just take it off and keep snoozing
  3. Proximity Detection to know if your still in bed/bedroom
  4. Speakers with custom alarm sounds
  5. LCD display with GUI
  6. Other.. regular old alarm watch stuff

Utilised Technology

  • Raspberry Pi 3 Model B
  • Arduino
  • 3D Printing
  • Computer-Aided 3D Modelling
  • C++
  • A cheap chinese electronic lock

Do It Yourself Sitcom Entrance Machine

Inspired by this video I have decided to build and neatly package a door sensor that can play sitcom entrance riffs when people pass through the door. This little machine will hopefully move with me to my first apartment where is will soon annoy all my guests.

Current Project Status - May 2016

All parts have been purchased except for a microSD card, which is bought and on its way here.


answers-to-common-AI-Qs

A.I from the perspective of people who write papers, not magazines


noam-comments

A comment system based on the values of the Noam Editor


Ontology onto Model

Evaluate an embedding space's compatibility with an ontological model.

It just crossed my mind that at times word embedding spaces, trained to approximate some number of ontologies, are employed in Machine Learning architectures that rely on on the fidelity of those approximated ontologies to the actual ontology/ontologies of the problem space.

It seems to me (a newbie) that the datasets used to train these ML models maintain an implicit or explicit set of ontologies, and that during training and use it may be important that these models do not pervert these ontologies.

Speculative Benefits of this hypothetical tool

  • Increased model intepretability
  • Improvements to model training process
  • Can reveal problems in existing ontologies, or reveal latent (right word?) ontologies

Vague roadmap

  • Understand how to analyse vector embedding spaces
  • Understand how to assess whether an embedding space supports a particular ontological relationship
  • Figure out how to easily and quickly specify an ontology
  • Figure out how to ingest multiple existing ontology specification types
  • Figure out how to visualise it all

under-the-surface

Blog Post on Google News's 'fact-checking tag' feature: https://blog.google/topics/journalism-news/labeling-fact-check-articles-google-news/

Paper on "Information Structure", which really deals with the structure of our language (linguistics) rather than the structure of media/content. Not super relevant but still exploring things in a similar way. LINK

Another paper on Information Structure (IS). Information Structure: Notional Distinctions, Ways of Expression

This looks much more like what we're talking about. See this -> "This papers aims at studying the existence of links between the structure of online communications and the contents they are composed of." Emotional Links between Structure and Content in Online Communications

"Content Structure" Landing Page from TECH WHIRL. The following seems like its of interest..

The discipline of structured authoring, or structured writing, evolved out of the particular need to structure and organize complex content for presentation to consumers of that content across various media. This TechWhirl research area provides a cross-section of information that introduces the concepts of content structure and structured writing, as well as tools and best practices for managing and producing content efficiently and effectively.

TechWhirl.com

Haven't read into this too much, but it seems to be revealing on the notion of there being 'layers of information' or 'layers of meaning'. Two-Dimensional Semantics

Extraction of references from scientific papers. Extraction of References Using Layout and Formatting Information from Scientific Articles. An existing extractor is called ParsCit

Relates to this stuff very closely I think - https://www.newscientist.com/article/mg22530102-600-google-wants-to-rank-websites-based-on-facts-not-links/


auto-tai-lopez

'read' a book a day. get knawwwledgeee!


shNERlock

Easy to follow Named Entity Recognition (NER) for Reddit content. Written in Python


Take Me Somewhere Android App

Take Me Somewhere is my first app that I am developing. The basic idea is to provide relaxing layered audio that evokes the feeling of being in a certain place on earth.

Current Development Status

Initial stages. Development is postponed until the completion of a higher-priority project. Estimated restart time: Early October 2016

Installation

App will eventually be available on the Android Play Store.

User Guide

Coming after app launch.


Narrative ~ Meaning From Data

A quantified self, and from it the story of your life


Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment