Borealis
Ruest, Nick;
Sala, Christine;
Thurman, Alex
—
2020-02-16
<p>Web archive derivatives of the <a href="https://archive-it.org/collections/1757">Avery Library Historic Preservation and Urban Planning</a> collection from <a href="https://archive-it.org/home/Columbia">Columbia University Libraries</a>. The derivatives were created with the <a href="https://github.com/archivesunleashed/aut/">Archives Unleashed Toolkit</a> and <a href="https://cloud.archivesunleashed.org/">Archives Unleashed Cloud</a>.</p>
<p>The <strong>cul-1757-parquet.tar.gz</strong> derivatives are in the <a href="https://parquet.apache.org/">Apache Parquet format</a>, which is a <a href="http://en.wikipedia.org/wiki/Column-oriented_DBMS">columnar storage</a> format. These derivatives are generally small enough to work with on your local machine, and can be easily converted to Pandas DataFrames. See <a href="https://github.com/archivesunleashed/notebooks/blob/master/datathon-nyc/parquet_pandas_stonewall.ipynb">this</a> notebook for examples.</p>
<p><strong>Domains</strong></p>
<pre>
<code class="language-java">.webpages().groupBy(ExtractDomainDF($"url").alias("url")).count().sort($"count".desc)</code></pre>
<p>Produces a DataFrame with the following columns:</p>
<ul>
<li>domain</li>
<li>count</li>
</ul>
<p><strong>Web Pages</strong></p>
<pre>
<code class="language-java">.webpages().select($"crawl_date", $"url", $"mime_type_web_server", $"mime_type_tika", RemoveHTMLDF(RemoveHTTPHeaderDF(($"content"))).alias("content"))</code></pre>
<p>Produces a DataFrame with the following columns:</p>
<ul>
<li>crawl_date</li>
<li>url</li>
<li>mime_type_web_server</li>
<li>mime_type_tika</li>
<li>content</li>
</ul>
<p><strong>Web Graph</strong></p>
<pre>
<code class="language-java">.webgraph()</code></pre>
<p>Produces a DataFrame with the following columns:</p>
<ul>
<li>crawl_date</li>
<li>src</li>
<li>dest</li>
<li>anchor</li>
</ul>
<p><strong>Image Links</strong></p>
<pre>
<code class="language-java">.imageLinks()</code></pre>
<p>Produces a DataFrame with the following columns:</p>
<ul>
<li>src</li>
<li>image_url<br />
</li>
</ul>
<p>The <strong>cul-1757-auk.tar.gz </strong>derivatives<strong> </strong>are the <a href="https://cloud.archivesunleashed.org/derivatives">standard set of web archive derivatives</a> produced by the Archives Unleashed Cloud.</p>
<ul>
<li><strong>Gephi </strong>file, which can be loaded into <a href="https://gephi.org/">Gephi</a>. It will have basic characteristics already computed and a basic layout.</li>
<li><strong>Raw Network</strong> file, which can also be loaded into <a href="https://gephi.org/">Gephi</a>. You will have to use that network program to lay it out yourself.</li>
<li><strong>Full text</strong> file. In it, each website within the web archive collection will have its full text presented on one line, along with information around when it was crawled, the name of the domain, and the full URL of the content.</li>
<li><strong>Domains count</strong> file. A text file containing the frequency count of domains captured within your web archive.</li>
</ul>
<p>Due to file size restrictions in Scholars Portal Dataverse, each of the derivative files needed to be split into 1G parts. These parts can be joined back together with <code>cat</code>. For example:
<pre><code class="bash">cat cul-1757-parquet.tar.gz.part* > cul-1757-parquet.tar.gz</code></pre>
</p>