The article compares different ways to group data together, called clustering, using three specific techniques: k-means clustering, hierarchical clustering, and Self-Organizing Map. The researchers tested these methods on online product reviews and found that k-means clustering and Self-Organizing Map were the most accurate for analyzing clusters. They also discovered that the accuracy of clustering online reviews depends on the structure and amount of sentences. This study shows that clustering techniques can help improve text summarization for customer review systems.