Leader, Sentiment Analysis of Audi A6L User Reviews
Text
Emotion RecognitionI led a project on sentiment analysis of Audi A6L user reviews, focusing on data cleaning, semantic processing, and annotation. The tasks included text de-duplication, outlier removal, word segmentation, and multi-level stopword filtering to prepare high-quality labeled text data. Fine-grained sentiment analysis was performed, with manual review and annotation confirmed for emotion extraction and label accuracy across key dimensions. • Processed and annotated 4000+ textual user comments to create a 3997-sample clean dataset • Applied pre-trained models for emotion recognition with verification and cross-validation • Constructed and interpreted LDA topic models to identify themes and sentiment clusters • Generated labeled data for downstream product optimization and consumer behavior studies.