Dirty Data, Broken Models: Cleaning an Audiobook Dataset for Machine Learning

2026-06-22 · Source: Machine Learning on Medium · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, medium

Summary

This article details the data cleaning process for an Audible audiobook dataset, preparing it for machine learning applications. It identifies critical data quality issues, particularly within the "stars" column, which initially contained 665 unique, unstandardized text values like "4.5 out of 5 stars181 ratings". The cleaning involved extracting "star_rating" and "rating_count" into separate numerical fields, resulting in standardized "star_rating" values such as 1.0, 1.5, up to 5.0. Additionally, the "author" and "narrator" columns were cleaned by removing prefixes like "Writtenby:" and "Narratedby:". The transformed dataset is now suitable for analysis and will be used in a subsequent project phase to build a recommendation system using clustering algorithms like K-Means, Hierarchical Clustering, and DBSCAN.

Key takeaway

For data scientists or ML engineers preparing real-world datasets, meticulously cleaning raw data is non-negotiable for model reliability. If your dataset contains combined text fields like "4.5 out of 5 stars181 ratings" or prefixed strings, you must implement robust parsing and standardization. This ensures your features, like "star_rating" and "rating_count", are numerical and consistent, directly impacting the accuracy and effectiveness of subsequent machine learning models, such as recommendation systems.

Key insights

Real-world datasets demand meticulous cleaning to transform inconsistent, raw data into structured features suitable for machine learning models.

Principles

• High-quality data improves model accuracy.
• Unstructured text often contains multiple features.
• Preprocessing enhances data readability.

Method

Identify combined text fields, extract numerical components using regex, convert to numeric types, and remove unnecessary prefixes from string fields to standardize data for ML.

In practice

• Apply regex to parse complex rating strings.
• Convert extracted rating values to floats.
• Strip "Writtenby:" and "Narratedby:" prefixes.

Topics

• Data Cleaning
• Machine Learning Datasets
• Feature Engineering
• Audiobook Data
• Data Preprocessing
• Recommendation Systems

Best for: Machine Learning Engineer, Data Scientist, AI Student

Related on AIssential

• How to Prepare Data for Machine Learning Models — Effective data preparation is paramount for machine learning model success, often more critical than complex architectures.
• Apache Spark Mastery (Part 3of 6): Wrangling Messy Data — Effective data wrangling in Spark involves systematic cleaning, standardization, and handling of diverse data structures.
• Data Engineering For Machine Learning: How to load, clean and prepare data. — Effective multimodal ML requires robust data pipelines for loading, cleaning, and aligning diverse data types.
• 7 myths about “more data” (and why models get worse) — More data can degrade model performance by introducing noise, distribution shifts, or label poisoning.
• Why Data Cleaning Takes Longer Than the Model In AI — High-quality data preparation is more critical for machine learning success than complex algorithms.
• CoEval: Ranking Language Models for Custom Tasks Without Labeled Data or Trustworthy Benchmarks — The CoEval framework enables reliable, label-free LLM ranking for custom tasks by synthesizing fresh benchmarks and using diverse judge ensembles.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning on Medium.