test


<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>SCRATCH</title>
<meta name="description" content="This file was exported from WriteMonkey editor.">
<style type="text/css">
@import url(http://fonts.googleapis.com/css?family=Vollkorn:400,400italic,700,700italic&subset=latin);

html, body {
        padding:1em;
        margin:auto;
        max-width:42em;
        background:#fefefe;
    }
body {
    font: 1.3em "Vollkorn", Palatino, Times;
    color: #333;
    line-height: 1.4;
    text-align: justify;
    }
header, nav, article, footer {
    width: 700px;
    margin:0 auto;
    }
article {
    margin-top: 4em;
    margin-bottom: 4em;
    min-height: 400px;
    }
footer {
    margin-bottom:50px;
    }
video {
    margin: 2em 0;
    border:1px solid #ddd;
    }

nav {
    font-size: .9em;
    font-style: italic;
    border-bottom: 1px solid #ddd;
    padding: 1em 0;
    }
nav p {
    margin: 0;
    }

/* Typography
-------------------------------------------------------- */

h1 {
    margin-top: 0;
    font-weight: normal;
    }
h2 {
    font-weight: normal;
    }
h3 {
    font-weight: normal;
    font-style: italic;
    margin-top:3em;
    }
p {
    margin-top:0;
    -webkit-hypens:auto;
    -moz-hypens:auto;
    hyphens:auto;
    }
ul {
    list-style: square;
    padding-left:1.2em;
    }
ol {
    padding-left:1.2em;
    }
blockquote {
    margin-left: 1em;
    padding-left: 1em;
    border-left: 1px solid #ddd;
    }
code {
    font-family: "Consolas", "Menlo", "Monaco", monospace, serif;
    font-size: .9em;
    background: white;
    }
a {
    color: #2484c1;
    text-decoration: none;
    }
a:hover {
    text-decoration: underline;
    }
a img {
    border:none;
    }
h1 a, h1 a:hover {
    color: #333;
    text-decoration: none;
    }
hr {
    color : #ddd;
    height : 1px;
    margin: 2em 0;
    border-top : solid 1px #ddd;
    border-bottom : none;
    border-left: 0;
    border-right: 0;
    }
p#heart{
    font-size: 2em;
    line-height: 1;
    text-align: center;
    color: #ccc;
    }
.red {
    color:#B50000;
    }

/* Home Page
--------------------------- */

body#index li {
    margin-bottom: 1em;
    }


/* iPad
-------------------------------------------------------- */
@media only screen and (max-device-width: 1024px) {
body {
    font-size: 120%;
    line-height: 1.4;
    }
} /* @iPad */

/* iPhone
-------------------------------------------------------- */
@media only screen and (max-device-width: 480px) {
body {
    text-align: left;
    }
article, footer {
    width: auto;
    }
article {
    padding: 0 10px;
    }
} /* @iPhone */
</style>

</head>
<body>
<h1>ADS - Headstart</h1>
<p>Date: 30.07.2015 - 17:00</p>
<p>Participants: Roman Chyla/Peter Kraker/Asura Enkhbayar</p>
<h2>Questions</h2>
<ol>
<li>Why are the raw reader id's not public?</li>
<li>Is there a possibility to access more than the last 90-days?</li>
<li>How does the trending() operator work in detail and is there any possibility to quantify the co-reads between to articles?</li>
</ol>
<h2>Headstart</h2>
<ul>
<li>ADS data (<em>readers</em>, <em>abstract</em>, <em>title</em>)</li>
</ul>
<blockquote>
<p>Readers -&gt; Co-readership -&gt; Clustering/Placement
Abstract, title -&gt; Naming of clusters</p>
</blockquote>
<ul>
<li>Clustering/Placement</li>
</ul>
<blockquote>
<p>Co-read matrix is used to find clusters in the data (mainly hclust)
Placement with n-MDS, t-SNE
t-SNE -&gt; Clustering on the 2d data
Using cosine similarity for the co-read data</p>
</blockquote>
<ul>
<li>Graph-based approach</li>
</ul>
<blockquote>
<p>Export the co-read matrix to a network file
Import in gephi and run modularity based community detection
Force-directed placement of documents
Export the positions of the nodes</p>
</blockquote>

</body>
</html>