<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Data Mining &amp; Analysis on Wasi's Log</title><link>https://wasi0013.com/categories/data-mining--analysis/</link><description>Recent content in Data Mining &amp; Analysis on Wasi's Log</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 21 Sep 2025 19:22:41 +0600</lastBuildDate><atom:link href="https://wasi0013.com/categories/data-mining--analysis/index.xml" rel="self" type="application/rss+xml"/><item><title>Detecting and Fixing Data Quality Problems in Data Pipelines and ML Systems</title><link>https://wasi0013.com/2025/09/21/detecting-and-fixing-data-quality-problems-in-data-pipelines-and-ml-systems/</link><pubDate>Sun, 21 Sep 2025 19:22:41 +0600</pubDate><guid>https://wasi0013.com/2025/09/21/detecting-and-fixing-data-quality-problems-in-data-pipelines-and-ml-systems/</guid><description>&lt;p&gt;Early in my career, I built many web scrapers, some of these early scrapy scrapers used to aggregate product pricing data from multiple e-commerce websites. The initial dataset looked complete, and basic analysis suggested reliable trends. Once I integrated the data into a dashboard for decision-making, inconsistencies appeared due to slight differences in HTML structure across sites. These hidden issues caused subtle quiet failure, but the impact stacked up quickly after weeks. This was a nice example of the hidden risk of poor data quality problems in data pipelines.&lt;/p&gt;</description></item></channel></rss>