Stack Data Attends the First Insurance Technology Entrepreneurship Week • Xingchenhui Insurance Entrepreneurship Project Roadshow

2019-12-11

In recent years, hundreds of insurance technology startups have emerged in China, driving service upgrades through technological advancements and giving rise to numerous entrepreneurial projects that connect insurance services with various insurance scenarios such as healthcare, wellness, elderly care, travel, and consumption. Currently, the Chinese insurance industry is undergoing transformation and upgrading, accelerating its development cycle and urgently requiring fresh blood and deep industry communication to continuously integrate technological achievements with the real economy, drive the development of the insurance industry with new momentum, and better serve the people. Against this backdrop, the first Insurance Technology Entrepreneurship Week · Xingchenhui Insurance Entrepreneurship Project Roadshow has emerged. Insurance technology companies, pharmaceutical enterprises, health management companies, and investment institutions hope to use this roadshow as an opportunity to promote cross-industry communication and collaboration between the insurance industry and the internet, technology, healthcare, wellness, elderly care, and security industries, forming sustained industry exchanges and cooperation. Tian Jiao, Vice President of Zhanlue Data Business, took the stage to deliver a speech, focusing on data collection, model construction, knowledge base development, and intelligent production management. She then introduced how Zhanlue Data carries out innovation in health insurance risk management: First, establish a "knowledge base + image recognition" system. For example, at the data entry end, OCR technology is used to assist human input, reducing data collection time and improving the quality of collection. Under the premise of continuously updating the knowledge base, fuzzy matching is fully realized through the underlying knowledge base, filling in project information, improving feedback collection for manual work, and developing data quality inspection tools that are in line with actual production. Second, build a medical knowledge graph and use it as a core product to gain insights into the entire physiological cycle of disease evolution from perspectives such as disease population, drug information, medical expenses, and treatment distribution, solidifying the foundation of risk control. Third, establish a natural language model and use NLP algorithms for classification to generalize biases. In the natural language processing stage, combined with deep learning technology, machines can better grasp human language, possessing multiple functions such as machine translation, sentiment analysis, information extraction, text classification, intelligent question answering, and language generation. When applied to specific business scenarios, the natural language model can quickly classify diseases, thereby solving problems such as inaccurate data entry, non-medical professional audit work, and human error, significantly improving the automation level of the system. Fourth, establish a model for abnormal diagnosis and treatment trajectories, using time series features, anomaly detection, and isolation forest algorithms for abnormal detection, screening out cases with significant differences in medical logic and treatment trajectories compared to patients at the same level, and providing risk alerts. Fifth, closely focus on product design, pricing data support, and risk control claims to infuse risk control genes into the blood of product design.

RECENT AI NEWS

RECENT AI TOOLS