Researchers have developed machine learning models to predict user frustration in e-commerce sessions using clickstream data. The study defines frustration based on specific user behaviors like rage bursts and excessive navigation. Both XGBoost and Long Short-Term Memory (LSTM) classifiers were trained on millions of data points, with the LSTM achieving higher accuracy and AUC scores. Notably, the LSTM model demonstrated the ability to reliably predict frustration using only the initial 20 to 30 user interactions. AI
IMPACT This research could enable businesses to proactively identify and address user frustration in real-time, improving customer experience and conversion rates.
RANK_REASON Academic paper detailing a new machine learning approach to predict user frustration. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX
- DagsHub
- Gotit.pub
- Hugging Face
- IArxiv
- Jibin Joseph
- long short-term memory
- ScienceCast
- XGBoost
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
