Advances in Business Statistics, Methods and Data Collection

24Six Years of Machine Learning in the Bureau of Labor Statistics

Alexander Measure

U.S. Bureau of Labor Statistics, Washington, D.C., USA

24.1 Introduction

The past 10 years have seen dramatic advances in the computer's ability to process and understand human language. In a thoroughly publicized 2011 event, a computer beat the world's best human players at the trivia game of Jeopardy by extracting answers from massive collections of written text (Markoff 2011). The following years have brought many related advances that now impact our lives in countless ways. Computers accurately translate between dozens of languages, consumers regularly issue verbal commands to devices that respond intelligently, and anyone can get answers to millions of questions by simply typing them into a search engine.

These and many other advances have been made possible by rapid progress in the fields of machine learning (teaching computers to learn from data) and natural language processing (using computers to understand language). Increasingly, these topics are also of interest in official statistics. In 2019, for example, the UN Group for the Modernization of Official Statistics launched a Machine Learning project which attracted more than 120 participants from 33 national and 4 international organizations (Julien 2021). New venues such as the BigSurv conference have also emerged to address this growing interest. Nonetheless, there is relatively little information about how statistical organizations ...