Moving Beyond Survey Approaches:

DOI Web Site

Bibliographic Information

Other Title
  • アンケート調査を越えて
  • アンケート調査を越えて : 自然言語処理や機械学習を用いたログデータの活用を模索する
  • アンケート チョウサ オ コエテ : シゼン ゲンゴ ショリ ヤ キカイ ガクシュウ オ モチイタ ログデータ ノ カツヨウ オ モサク スル
  • Exploring Research Applications of Log Data Based on Natural Language Processing and Machine Learning
  • ―自然言語処理や機械学習を用いたログデータの活用を模索する―

Search this article

Abstract

<p>In research activities, it is an important task to appropriately extract abstract concepts that construct the essential part of theory. Although management scholars have used survey techniques for a long time to obtain variables that represent various constructs, these techniques have at the same time made it hard to obtain large samples or time-series data, leaving some theoretical mechanisms difficult to examine. In order to solve these issues, this paper introduces log data, which are massive records of natural human behavior, and also discusses pros and cons of using survey techniques and log data. Furthermore, log data have own problems, one of which is the difficulty to adequately capture psychological and sociological constructs. Thus, as a way to make log data represent the constructs appropriately, this paper examines the applications of machine learning and natural language processing to log data.</p>

Journal

  • Organizational Science

    Organizational Science 55 (1), 16-30, 2021-09-20

    The Academic Association for Organizational Science

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top