基于机器学习技术的骚扰电话识别技术研究

Research on harassing phone recognition technology based on machine learning

随着移动通信服务的不断发展,广泛存在的骚扰电话违法行为愈演愈烈,成为当前通信运营商急需解决及优化治理的主要通信违法行为,造成了严重的负面社会影响。本文提出基于机器学习及大数据分析的骚扰电话拨打行为分析的治理方案。方案基于移动通信网络在客户使用过程中产生的各类语音信令数据,通过机器学习技术,对保存在大数据集群上的信令数据进行分析与训练,实现更加精准的骚扰电话分析研判策略制定,形成基于“大数据采集、策略制定、分析研判,封堵处置”的闭环管理流程,达到业务应用优化的目的。

With the continuous development of mobile communication services,the widespread illegal behavior of harassing phone calls has become increasingly severe,becoming the main communication illegal behavior that communication operators urgently need to solve and optimize governance,causing serious negative social impacts.This project proposes a governance plan for analyzing harassing phone calls based on machine learning and big data analysis technology.The plan is based on various voice signaling data generated by mobile communication networks during customer usage.Based on machine learning technology,the signaling data stored in big data clusters is analyzed and trained to achieve more accurate analysis and judgment strategies for harassing phone calls.Establish a closed-loop management process based on"big data collection,strategy formulation,analysis and judgment,and blockage and disposal"to achieve the research goal of optimizing on-site business applications.