2014-02-luton-open-data-101.html

---
layout: reveal
title: Open Data 101
description: University of Bedfordshire, 11 February 2014
author: Dr Tom Heath
twitter: tommyh
---


{% include odi_logo.html %}

{% include standard_title.html %}


<style type="text/css">
@import url(https://fonts.googleapis.com/css?family=Quicksand:400,700,400italic,700italic);
@import url(https://fonts.googleapis.com/css?family=Open+Sans:400italic,700italic,400,700);

reveal {
  font-family: "Open Sans", sans-serif;
  font-size: 36px;
  font-weight: 200;
  letter-spacing: 0.6em;
  color: #333333;
 }
.reveal h1,
.reveal h2,
.reveal h3,
.reveal h4,
.reveal h5,
.reveal h6 {
  margin: 0 0 20px 0;
  color: #333333;
  font-family: "Quicksand", sans-serif;
  line-height: 1.2em;
  letter-spacing: -0.08em;
  text-transform: uppercase;
  text-shadow: none;
 }
</style>

<!-- 
<section>
	<h2>Background</h2>
	<ul>
		<li>Head of Research at ODI</li>
		<li>Senior Research/Data Scientist at Talis Group</li>
		<li>PhD in Social Network-driven Recommender Systems</li>
	</ul>
</section>
-->

<!-- 
<section>
	<h2>Overview</h2>
	<ol>
		<li>The Open Data Institute</li>
		<li>What is Open Data?</li>
		<li>Applications of Open Data</li>
		<li>Considerations for Open Data Consumers</li>
		<li>Pointers and Questions</li>
	</ol>
</section>
-->

<!-- ==== Open Data Institute ==== -->

<section><h2>The Open Data Institute</h2></section>
<section>
	<h2>The Open Data Institute</h2>
	<ul class="centred-list">
		<li><strong>founded in 2012</strong><br/>by Tim Berners-Lee and Nigel Shadbolt</li>
		<li><strong>non-profit, non-partisan company</strong><br/>limited by guarantee</li>
		<li>helping others to be<br/><strong>successful with open data</strong></li>
	</ul>
</section>
<section>
	<h2>ODI Headquarters</h2>
	<h3>Tech City in East London</h3>
	<p><img src="2014-02-luton-open-data-101/lovquist_5587608199_3d50f55ab6_o-cropped.jpg"/></p>
	<p style="font-size:small">photo by <a href="http://www.flickr.com/photos/lovquist/5587608199/">Daniel L&ouml;vquist</a>, used under license <a href="http://creativecommons.org/licenses/by-nc/2.0/deed.en">CC BY-NC 2.0</a></p>
</section>

<!-- ==== Open Data Institute ==== -->

<section>
	<h2>What is Open Data?</h2>
</section>
<section>
	<p>
		<blockquote>A piece of data or content is open if anyone is free to use, reuse, and redistribute it &mdash; subject only, at most, to the requirement to attribute and/or share-alike.</blockquote>
		<span class="quoteattr">&mdash;&nbsp;<a href="http://opendefinition.org/">http://opendefinition.org/</a></span>
	</p>
</section>
<section>
	<h2>The Open Data Deal</h2>
	<ul class="centred-list">
		<li>
			<strong>open data</strong>
			<br/>must have a <em>licence</em> to say it is open
			<br/>(just being on the web is not enough)
		</li>
		<li class="fragment">
			<strong>the license</strong>
			<br/>may impose some constraints:<br/><em>attribution</em> and/or <em>share-alike</em>
		</li>
		<li class="fragment">
			<strong>the upside</strong>
			<br/>you&apos;re free to build cool things<br/>without having to pay for the data or collect it yourself
		</li>
	</ul>
</section>
<section>
	<h2>Big/Open/Personal Data</h2>
	<p><img class="stretch" src="2014-02-luton-open-data-101/venn-diagram-open-big-personal.png" style="border:none; box-shadow:none;"/></p>
	<p style="text-align:center; font-size: smaller;">Image Credit: Ulrich Atz</p>
</section>


<!-- ==== Applications ==== -->

<section><h2>Applications of<br/>Open Data</h2></section>
<section>
	<h2>Prescribing Analytics</h2>
	<h3>The Question</h3>
	<ul class="centred-list">
		<li class="fragment">annual prescription drugs bill for NHS is <strong>~&pound;10 billion</strong></li>
		<li class="fragment">potential cost savings from prescribing <strong>generic drugs</strong></li>
		<li class="fragment"><strong>how big are the savings?</strong></li>
	</ul>
</section>
<section>
	<h2>Prescribing Analytics</h2>
	<h3>The Method</h3>
	<ul class="centred-list">
		<li class="fragment">prescription data for england, published monthly<br/>anonymised, under an open license</li>
		<li class="fragment">work with domain specialists<br/>select one drug class: statins</li>
		<li class="fragment">find all cases where generic could have been prescribed<br/>with no difference in patient outcomes</li>
		<li class="fragment">what is the cost difference?</li>
	</ul>
</section>
<section>
		<img src="2014-02-luton-open-data-101/prescribinganalytics.png"><br/>
		<span class="quoteattr"><a href="http://prescribinganalytics.com/">http://prescribinganalytics.com/</a></span>
</section>
<section>
	<h2>Prescribing Analytics</h2>
	<h3>The Results</h3>
	<ul class="centred-list">
		<li class="fragment">potential costs savings to national health service of<br/><strong>&pound;200 million per year</strong></li>
		<li class="fragment"><strong>for just one drug class</strong></li>
	</ul>
</section>
<section>
	<h2>Fire Stations in London</h2>
	<ul class="centred-list">
		<li class="fragment">proposal to close a number of <strong>fire stations</strong>
				<br/>in central london</li>
		<li class="fragment">candidates for closure identified by
				<br/><strong>non-transparent analysis/process</strong></li>
		<li class="fragment"><strong>questions</strong>
				<br/>is the analysis fair/accurate?
				<br/>how can citizens engage in the process?
		</li>
	</ul>
</section>
<section>
	<img src="2014-02-luton-open-data-101/london-fire-cropped.png"><br/>
	<span class="quoteattr"><a href="http://london-fire.labs.theodi.org/">http://london-fire.labs.theodi.org/</a></span>
</section>
<section>
	<h2>Fire Stations in London</h2>
	<h3>Conclusions</h3>
	<ul class="centred-list">
		<li class="fragment">findings broadly compatible with initial analysis</li>
		<li class="fragment">significantly cheaper than official analysis?</li>
		<li class="fragment"><strong>tool is publicly available</strong>
				<br/>citizens can explore the options
		</li>
	</ul>
</section>


<!-- ==== Considerations for Data Consumers ==== -->

<section><h2>Considerations for <br/>Open Data Consumers</h2></section>

<section id="handout">
	<h2>Overview</h2>
    <ol class="fragment">
      <li style="color:#D60303">Accessibility</li>
      <li style="color:#2254F4">Ownership and licensing</li>
			<li style="color:#0DBC37">Form</li>
			<li style="color:#FF6700">Quality</li>
			<li style="color:#B13198">Support</li>
    </ol>
</section>


<section id="handout1">
	<h2 style="color:#D60303">Accessibility (1/5)</h2>
    <ul class="centred-list">
      <li class="fragment"><strong>is the data already available?</strong><br/>
      if so, where?</li>
      <li class="fragment"><strong>how can you access it?</strong><br/>
      dumps? an <em>api</em>?</li>
			<li class="fragment"><strong>in what format is the data published?</strong><br/>
			csv? xml? json? pdf?!</li>
    </ul>
</section>

<section id="handout2">
	<h2 style="color:#2254F4">Ownership and licensing (2/5)</h2>
    <ul class="centred-list">
      <li class="fragment"><strong>who publishes the data?</strong></li>
      <li class="fragment"><strong>are they the originator of the data?</strong></li>
			<li class="fragment"><strong>under what licence is the data published?</strong></li>
			<li class="fragment"><strong>does it contain personal data?</strong></li>
    </ul>
</section>

<section id="handout3">
	<h2 style="color:#0DBC37">Form (3/5)</h2>
    <ul class="centred-list">
      <li class="fragment"><strong>prior processing</strong><br/>
      is the data in raw or summary form?<br/>
      how has it been processed?</li>
			<li class="fragment"><strong>form</strong><br/>
			shape, granularity, etc.<br/>
			how will these affect your analysis/product/application?<br/>
			is the form compatible with other data you are using?</li>
			<li class="fragment"><strong>transformations</strong><br/>
			what syntactic and semantic transformations<br/>will you need to make?</li>
    </ul>
</section>

<section id="handout4">
	<h2 style="color:#FF6700">Quality (4/5)</h2>
    <ul class="centred-list">
      <li class="fragment"><strong>currency and regularity</strong><br/>
      how current is the data?<br/>
      how regularly is it updated?<br/>
			for how long will it be published?<br/>
			what is the commitment by the publisher?</li>
			<li class="fragment"><strong>comprehensibility</strong><br/>
			do you understand all the fields and their context?</li>
			<li class="fragment"><strong>accuracy</strong><br/>
			what do you know about the accuracy of the data?<br/>
			how is missing data handled?</li>
    </ul>
</section>

<section id="handout5">
	<h2 style="color:#B13198">Support (5/5)</h2>
    <ul class="centred-list">
			<li class="fragment"><strong>(how) is the data set documented?</strong></li>
			<li class="fragment"><strong>does the metadata make sense?</strong></li>
      <li class="fragment"><strong>is there a place you can report errors in the data?</strong></li>
			<li class="fragment"><strong>does the publisher offer support in any way?</strong></li>
    </ul>
</section>



<!-- ==== Pointers and Questions ==== -->

<section>
	<h2>Pointers</h2>
	<ul class="centred-list" style="font-size:0.6em;">
		<li>Example Data Sources: <a href="http://is.gd/example_data_sources">http://is.gd/example_data_sources</a></li>
		<li>OpenRefine (data cleansing): <a href="http://openrefine.org/">http://openrefine.org/</a></li>
		<li>R (statistical programming): <a href="http://r-project.org/">http://r-project.org/</a></li>
		<li>Hadoop (big data platform): <a href="http://hadoop.apache.org/">http://hadoop.apache.org/</a></li>
		<li>D3 (visualisation library): <a href="http://d3js.org/">http://d3js.org/</a></li>
		<li>These Slides: <a href="http://theodi.github.io/presentations/2014-02-luton-open-data-101.html">http://is.gd/open_data_101</a></li>
	</ul>
</section>

<section>
  <h2>Questions?</h2><br>
  <img height="320;" src="brand/odi_dynamic_logo.png" alt="ODI" style="border:none; box-shadow:none;"> 
  <hr>
  <p style="text-align:center">Tom Heath · Open Data Institute</p>
  <p style="text-align:center; font-size: smaller">
		<a href="mailto:tom.heath@theodi.org">tom.heath@theodi.org</a> ·
		<a href="https://twitter.com/tommyh">@tommyh</a>
	</p>
</section>

<!-- <section>
  <h2>Acknowledgements</h2>
  <p style="text-align:center"><strong>DaPaaS</strong> and <strong>OpenDataMonitor</strong> projects</p>
  <p style="text-align:center">funded by EC FP7 grants 610988 &amp; 611988</p>
  <img height="320;" src="brand/odi_dynamic_logo.png" alt="ODI" style="border:none; box-shadow:none;"> 
  <hr>
  <p style="text-align:center">Tom Heath · Open Data Institute</p>
  <p style="text-align:center; font-size: smaller">
		<a href="mailto:tom.heath@theodi.org">tom.heath@theodi.org</a> ·
		<a href="https://twitter.com/tommyh">@tommyh</a>
	</p>
</section> -->


{% include odi_tech_team.html %}