下载PDF
语音识别
概述
语音识别应用程序识别口语中的单词和短语，并将它们转换为机器可读的格式。随着机器学习的快速改进和智能助手（如亚马逊的 Alexa、苹果的 Siri 和微软的 Cortana）的发展，语音识别已经获得了突出和更广泛的应用。语音识别系统使消费者能够通过简单地与技术进行交互，从而实现免提请求、提醒和其他简单任务。一些语音识别系统需要“训练”，其中单个说话者将文本或孤立的词汇读入系统。该系统分析人的特定声音模式并使用它来微调对该人语音的识别，从而提高准确性。在商业环境中，应用程序也可以使用公司独有的词汇和短语进行训练。这在通常使用首字母缩略词或专业语言的情况下很有用。
案例研究.
Case Study
Driving Two Million Vehicles with Real-Time Data
For ARI, customer service is key. Seven global customer contact centers handle about 4 million calls annually, and this sea of daily information needs to be analyzed, which took up to two days manually, so negatively affecting service levels. With the amount of data ARI collects doubling every 14 months, their in-house ETL data warehouse limited their ability to fully help customers reduce fleet costs and gain operational efficiencies.
