# A study of text classification algorithms for live-streaming e-commerce comments based on improved BERT model

## TL;DR

This paper introduces an improved BERT model with a hierarchical structure to better classify short comments in live-streaming e-commerce.

## Contribution

The novel contribution is a hierarchical BERT model that improves classification accuracy and efficiency for e-commerce bullet comments.

## Key findings

- The hierarchical BERT model significantly improves classification accuracy for e-commerce bullet comments.
- The model efficiently extracts valuable information from large volumes of short, diverse comments.
- Combining BERT's semantic understanding with hierarchical categorization enhances marketing effectiveness.

## Abstract

As e-commerce live streaming becomes increasingly popular, the textual analysis of bullet comments is becoming more and more important. Bullet comments is characterized by its brevity, diverse content, and vast quantity. Faced with these challenges, this study proposes an improved BERT model based on a hierarchical structure for classifying e-commerce bullet comments. First, a parent class BERT model is trained to categorize bullet comments into six designated categories (parent categories). Subsequently, subclass BERT models are trained to classify bullet comments into subcategories. The model combines BERT’s profound semantic comprehension with the closely categorized capabilities of the hierarchical structure. Empirical evidence shows that the proposed model significantly improves classification accuracy and efficiency, aiding in further analysis of bullet comments, extracting valuable information, and achieving effective marketing.