下载PDF
实例探究 > Fortune 50 Bank Achieves Significant Performance Gains with Snorkel Flow for News Analytics

Fortune 50 Bank Achieves Significant Performance Gains with Snorkel Flow for News Analytics

技术
  • 分析与建模 - 机器学习
  • 分析与建模 - 预测分析
适用行业
  • 金融与保险
适用功能
  • 商业运营
服务
  • 软件设计与工程服务
挑战
The bank needed an accurate way to tag companies in unstructured news text, link them to identifiers (e.g., stock tickers), and classify mentions by sentiment and other aspects. The existing solutions, including a black box vendor system and internal heuristic approaches, were not meeting the performance requirements. The bank required a more efficient and accurate method to handle the vast amount of unstructured data feeds and derive meaningful insights from them.
关于客户
The customer is a Fortune 50 bank, one of the largest and most influential financial institutions in the world. With a vast array of services ranging from retail banking to investment banking, the bank operates on a global scale, serving millions of customers. The bank is known for its innovation in financial services and its commitment to leveraging cutting-edge technology to enhance its operations. Given its size and scope, the bank deals with an enormous amount of data daily, making efficient data processing and analysis crucial for maintaining its competitive edge.
解决方案
The bank implemented Snorkel Flow to develop an AI-powered news analytics application. This application was designed to monitor press coverage of target companies in unstructured data feeds. Snorkel Flow enabled the bank to accurately tag companies in news text, link them to relevant identifiers such as stock tickers, and classify mentions by sentiment and other aspects. The AI-driven approach provided a significant improvement over the legacy systems, allowing for more precise and faster data processing. The implementation of Snorkel Flow streamlined the bank's ability to analyze news data, providing more accurate and timely insights into market trends and company performance.
运营影响
  • The bank achieved a 25+ point performance gain over the legacy vendor system and internal heuristic approaches.
  • The news analytics application developed with Snorkel Flow reached a +90 F1 score, indicating high accuracy in tagging and classification tasks.
  • The implementation was 45 times faster compared to the traditional hand-labeling process, significantly reducing the time required for data processing.
数量效益
  • +90 F1 score for news analytics application
  • 45x faster compared to hand-labeling
  • +25% performance gain over black box vendor system

相关案例.

联系我们

欢迎与我们交流!

* Required
* Required
* Required
* Invalid email address
提交此表单,即表示您同意 IoT ONE 可以与您联系并分享洞察和营销信息。
不,谢谢,我不想收到来自 IoT ONE 的任何营销电子邮件。
提交

Thank you for your message!
We will contact you soon.