Title:
PRE-TRAINED DEEP NEURAL NETWORK MODEL-BASED USER CHURN PREDICTION SYSTEM
Document Type and Number:
WIPO Patent Application WO/2022/012605
Kind Code:
A1
Abstract:
Disclosed herein is a pre-trained deep neural network model-based user churn prediction system, comprising: a course teaching platform module provides a teacher object and a student object with a work terminal platform related to a pre-determined educational course; a corpus collection module collects a voice dialogue, text dialogue and teaching video of the teacher object and the student object in a classroom during the operation of the work terminal platform; a churn status module marks classroom corpuses of lost users as churn, and marks the remaining class corpuses as non-churn; a prediction algorithm module, according to classroom corpuses collected by the corpus collection module and the classroom corpuses marked by the churn status module, calculates a user churn prediction result by means of a trained deep neural network model; and a real-time monitoring module monitors the course teaching platform module in real time, and displays the user churn prediction result calculated by the prediction algorithm module to the teacher object. The described user churn prediction system achieves the effect of adjusting an educational counseling policy.
Inventors:
WANG XIN (CN)
XU ZHAOHUI (CN)
XU ZHAOHUI (CN)
Application Number:
PCT/CN2021/106382
Publication Date:
January 20, 2022
Filing Date:
July 15, 2021
Export Citation:
Assignee:
SHANGHAI SQUIRREL CLASSROOM ARTIFICIAL INTELLIGENCE TECH CO LTD (CN)
International Classes:
G06Q10/04
Domestic Patent References:
| WO2010130733A1 | 2010-11-18 |
Foreign References:
| CN111898810A | 2020-11-06 | |||
| CN110263326A | 2019-09-20 | |||
| CN110378812A | 2019-10-25 | |||
| CN110991381A | 2020-04-10 |
Other References:
YIN WEN: "Dropout Study Based on Learners’ Behavior in Massive Open Online Courses", CHINA MASTER’S THESES FULL-TEXT DATABASE, 15 January 2019 (2019-01-15), XP055888009
Attorney, Agent or Firm:
BEYOND ATTORNEYS AT LAW (CN)
Download PDF: