Zhong Guo Jing Ying Bao·2025-04-04 22:18Group 1: Anti-Fraud Models - The anti-fraud model developed by China Telecom has shown significant effectiveness in identifying and preventing potential fraud cases, utilizing data from call records, roaming behavior, and devices [1] - General anti-fraud big data models are capable of real-time monitoring of fraudulent activities and patterns, enhancing fraud prevention capabilities through multi-dimensional data analysis, including user behavior and credit ratings [3] Group 2: Data Asset Utilization - Data circulation plays an increasingly vital role in economic development, with the OECD estimating an average profit growth rate of about 10% across industries due to data flow, and up to 32% in sectors like digital platforms and finance [4] - The National Data Bureau is collaborating with the State-owned Assets Supervision and Administration Commission to formulate a three-year action plan aimed at enhancing the data efficiency of state-owned enterprises, transitioning data resources from storage to utilization [5] - A state-owned energy enterprise has achieved over 200 million yuan in annual data service revenue through market-oriented data operations, improving data collaboration efficiency in the industry by 40% [5] Group 3: Data Infrastructure and Innovations - China Telecom has amassed over 600PB of data in its data lake, creating a multi-source data foundation that has led to significant advancements in business, operational, and management digitization [6] - The company has developed the industry's first voice model capable of recognizing 50 dialects, based on a dataset of 500,000 hours of anonymized audio [6] - China Mobile and China Unicom have also reported successful implementations of their anti-fraud models, with Unicom specifically optimizing algorithms to combat GoIP fraud gangs [6]