Home

Hashmapd WIKI

This is the hub of our main intranet.

Other Resources:

There is also the dropbox, the mailing list, the issues list, code itself, the lean canvas, and the kanban project page.

Welcome to New People - info at high level

Getting started with Hashmapd - getting correct libraries installed

Algorithmic

• Measuring the probability that a tweet is in English
• Tokenize Data Algorithm
• Choose Tokens Algorithm
• Auto-Encoder Algorithm Stuff
• Layout Algorithm
• Labeling Algorithm
• Visualisation Algorithm

Technical - Offline (Training & Data Retrieval )

Offline / Command Line Steps

• Choosing People for Analysis
• Get Raw Data
• Prepare
• Train
• Get Codes
• Get Coordinates
• Get Map

The main Twitter Data / Project

• The rate limit problem
• Twitter/User Data Set

SysAdmin

• CouchDB Server/s
• GPU Enabled Server

Technical - Online (Web Service)

Website / Visualization Layer

• Notes on Canvas visualisation

Web API Layer

• TO DO

Service Layer

• Tokenizer
• Histogram Generator: ( Tokens -> Histogram, per user)
• HashCode Generator: (Histogram -> Hashcode (given Weights))
• TSNE Online: (Hashcode->Coordinate)
• Token Density Data: (-> Map Density, per token)
• Label Service: (-> Best 'labels' for given area of map)
• Map Tile Generator ** TO DO **

Middle Layer

• Task Manager / Thread Pool

Database and support

• User Documents and Download
• Raw Tweet Docs

SysAdmin

• CouchDB Server/s
• Download Service

Technical - Development Environment Stuff

• Setting up a python virtual environment
• Compiling cython code Cython instructions

Incidental Data Sets / Projects

• #EQNZ Tweets DataSet/Project
• MusicHype DataSet/Project
• MNIST DataSet/Project
• Reuters DataSet/Project

Business Stuff

• TO DO