[Pycon] [new paper] "Michael Penkov" - Rain + dragon = Electricity: Unravelling the mysteries of ideograms with Python.

[info a pycon.it]

Title: Rain + dragon = Electricity: Unravelling the mysteries of ideograms with Python.
Duration: 45 (includes Q&A)
Q&A Session: 0
Language: en
Type: Talk

Abstract: When does the Sun meet the Moon? What did the Axe say to the Tree? Why does the Child go to School? Welcome to the exciting, puzzling world of ideograms, an ancient writing system from a faraway land!

[Ideograms][1] are a writing system where each symbol represents an idea or concept. Examples of natural languages using ideograms include Chinese, Japanese, and Korean. Command of these languages requires memorizing, reading, and writing THOUSANDS of ideograms, causing much frustration to language learners. This talk brings Python and graph theory to the rescue, organizing the ideograms and exposing meaningful (and sometimes, fun) relationships between them.

This talk is for a general audience. There will be some code examples, but they will be in idiomatic Python, so people without advanced Python skills will still be able to appreciate them. This talk does not assume any previous knowledge of ideograms or natural languages that use them. I will introduce the basic concepts and keep things simple. Knowledge of basic graph theory would be helpful but not essential, because I will give a brief intro. The audience will learn how to apply tools like Python and graph theory to problems that at first do not seem to be solvable with those tools. The audience will see how Python makes it easy to leverage existing datasets, work with graphs, and achieve insight into a problem. Using fun and entertaining examples of real ideograms, I hope to encourage people to look for new and better ways to use Python to solve existing problems.

There's a short write-up about the ideogram work [here][2]. The source code for the write-up is [here][3]. I haven't presented this material anywhere before.

  [1]: https://en.wikipedia.org/wiki/Ideogram
  [2]: http://michael.penkov.id.au/blog/2017/11/29/heisig17.html
  [3]: https://github.com/mpenkov/heisig

Tags: [u'japanese', u'networkx', u'graph', u'nlp']