<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
        <title>Search on Data Henrik - Life in IT</title>
        <link>https://data-henrik.de/tags/search/</link>
        <description>Recent content in Search on Data Henrik - Life in IT</description>
        <generator>Hugo -- gohugo.io</generator>
        <language>en-us</language>
        <lastBuildDate>Mon, 22 Jun 2026 11:44:00 +0200</lastBuildDate><atom:link href="https://data-henrik.de/tags/search/index.xml" rel="self" type="application/rss+xml" /><item>
        <title>Db2 experiments with vectors and similarity search</title>
        <link>https://data-henrik.de/2026/06/db2-vector-geo-similarity/</link>
        <pubDate>Mon, 22 Jun 2026 11:44:00 +0200</pubDate>
        
        <guid>https://data-henrik.de/2026/06/db2-vector-geo-similarity/</guid>
        <description>&lt;img src="https://data-henrik.de/images/202606_Db2_vector_geo_map_markham.avif" alt="Featured image of post Db2 experiments with vectors and similarity search" /&gt;&lt;p&gt;Over the past days, I got back to using the &lt;a class=&#34;link&#34; href=&#34;https://www.ibm.com/docs/en/db2/12.1.0?topic=list-vector-values&#34;  target=&#34;_blank&#34; rel=&#34;noopener&#34;
    &gt;vector feature in Db2&lt;/a&gt; for &lt;a class=&#34;link&#34; href=&#34;https://data-henrik.de/2025/11/db2-vector-geo-search/&#34;  target=&#34;_blank&#34; rel=&#34;noopener&#34;
    &gt;geographic search&lt;/a&gt;. What would happen, so I asked myself, when instead of just latitute and longitude I would use population as a third dimension? How would the similarity search turn out? The experiments proofed to be quite interesting. And it turned the focus on the aspect that the way embeddings are generated, turning objects into vector data, has a significant impact on result quality and usefulness.&lt;/p&gt;</description>
        </item>
        
    </channel>
</rss>
