This project was a part of Big data hackathon conducted by Govt of Austraila at University of Sydney. Following is the submission link: http://hackerspace.govhack.org/content/watercooler

GovHack 2014 Descriptive Project Page

Team: Water Coolers - Big Data, Bite-Size Team members: Akhil Karun M 0469 618 962 Jennifer Quealy M 0404 847 302 Assisted by Yuriy Tyshetskiy Supported by GovHack SydUni 2014 Details of Project: Our hack is called "Big-data Bite-size" - our aim: to regularly 'find and feed' data from GovDatasets to interested 'parties' (i.e. specialists, researchers, journalists, individuals), based on their selection of a 'favourite' research topic, as can be found in any Gov Dataset We have built a simple but effective app platform, that provides a list and links to popular research topic 'favourites', which can be selected, for push-feeds Once registered, the party will receive push notifications when favourite data sets update, & find new & 'quite interesting' material as it is found & added to data sets, ready to access for informed use & practice Our challenge was to popularise the use of data sets which may currently be daunting in their number, complexity, accessibility and usefulness We have built complex algorithms to visualise data from across data sets, to make it more accessible and attractive for parties, using a flash card approach The big aim is to have Big data Bite-size be the feed of choice for parties wanting to access and use a real and timely evidence-based data, for credibility and integrity, in their specialist, research and journalism practice; these people are generally time-poor and (evidence) data-hungry We aim to help parties move from 'old school' to 'awesome' research pathways, where:

"Old school" data = Days, weeks and months at the computer, trawling obscure and potentially thousands of data sets for what you need to know and without the technical ability to access and 'massage' the enormous stores of GovData data sets. It's like trying to find a needle in a haystack, where the haystack keeps getting bigger and bigger, and tossed and re-formed, and the needle almost impossible to find, simply for the amount and complexity of data that is potentially of interest to a party. "Old School" data also relies on word of mouth and slower forms of notification and alerts.

"Awesome" data = choosing "Big-Data Bite-sized", selecting what interests you, then getting real-time and specified data feeds, by push notification, - daily or as you choose - and to your mobile device. Awesome! It's brain food to feed and soothe the hungry beast of accessible research. It's every Dataset manager's dream - to get their data in hue he hands of the specialists who can create policy and drive practice change.

Data sets used:

ABS statabs.gov.au 2011 Census of population and housing, demography and health - Time series profiles, total household income weekly by rent Time series profiles LGA T24 State LGA Suburbs

Our Video link: Outcomes for Big Data - Bite Size: brings big data out of the servers and into the hands of regular people and for specialists, journalists, researchers and agencies - and makes for 'quite interesting' news tidbits, but backed by and linked to evidence-based research, for discussion and sharing, as parties gather around the coffee stand or water cooler takes in real government data from across the range of GovData data types, sources and values - allowing more awareness and engagement with the broad range of GovDatasets analyses data through complex algorithms, to generate simple, useful and shareable knowledge - that is visualised using an API (e.g. We are turning data sets into Motion charts and into personalised objects) curates really 'Quite Interesting' facts for people to have at their fingertips, and to uncover the 'geek' in everyone - another step to making science accessible and engaging for anyone allows anyone including health specialists, journalists and other parties to receive push notifications (alerts) for 'favourite' topics, like alcohol-related liver disease, or geo-located rental costs in income brackets Allows clients to put themselves in the picture of the national average on a mass range of indicators, and to build and see their own time trajectory related to national averages allows researchers to find the data to use with their questions and analyses. Code: Graphics: Mashups: Apps: Website URL

Workshopping notes on Water Cooler Team Hack

Other helpers: Yuriy Tyshetskiy 0432036011 James Polley

Two hacks: Water cooler (all data sets potentially) Health bites (only health - as an example)

Features

Add me Financial risk boundaries Rental vs income Social place indicators Difference niche: getting data that google doesn't access - feed Troll all data sets SMM sharing Need personal history

Big data is the big haystack - how to find the needle? Troll all data sets Allow people to pick their data set e.g. Health, nutrition, weather, finance, organic farming, farmers markets, gardening, motoring, travel, cooking Feed the latest news and releases (in small bites e.g. Twitter sized) to all who select - giving a tidbit of the research (e.g. A great statistic) and then links to the research, papers, other links and expertise Get the latest real news (evidence based) pure data, not gossip. To drive people to health institutions Help Doctors to troll the latest research in their particular field or relevant to their type of patient (eg liver cancer diabetes etc)

Hack needs: UI design Architecture (GAE) Video Demo: Video scribe Recording Structure strong base Script Data analytics ('R') - simple information: Graph Description Maps R scripts Details