ABSTRACT

The increasing volume of images published online in a wide variety of contexts requires communication researchers to address this reality by analyzing visual content at a large scale. Ongoing advances in computer vision to automatically detect objects, concepts, and features in images provide a promising opportunity for communication research. We propose a research protocol for Automated Visual Content Analysis (AVCA) to enable large-scale content analysis of images. It offers inductive and deductive ways to use commercial pre-trained models for theory building in communication science. Using the example of corporations’ website images on sustainability, we show in a step-by-step fashion how to classify a large sample (N = 21,876) of images with unsupervised and supervised machine learning, as well as custom models. The possibilities and pitfalls of these approaches are discussed, ethical issues are addressed, and application examples for future communication research are detailed.

摘要

在各种情况下在线发布的图像数量不断增加，要求传播研究人员通过大规模分析视觉内容来解决这一现实。自动检测图像中的对象，概念和特征的计算机视觉技术的不断发展为交流研究提供了广阔的机遇。我们提出了一种用于自动视觉内容分析（AVCA）的研究协议，以实现图像的大规模内容分析。它提供归纳和演绎方式，以使用商业预训练模型来构建通信科学中的理论。以公司关于可持续性的网站图像为例，我们逐步展示了如何使用无监督和监督的机器学习以及自定义模型对大样本（N = 21,876）图像进行分类。