Data play an increasingly important role in smart data analytics, which facilitate many data-driven applications. The goal of various data markets aims to alleviate the issue of isolated data islands, so as to benefit data circulation. The problem of data pricing is indispensable yet challenging in data trade. In this paper, we conduct a comprehensive survey on the modern data pricing solutions. We divide the data pricing solutions into three major strategies and thirteen models, including query pricing strategy, feature-based data pricing strategy, and pricing strategy in machine learning. It is so far the first attempt to classify so many existing data pricing models. Moreover, we not only elaborate the thirteen specific pricing models within each pricing strategy, but also make in-depth analyses among these models. We also conclude five research directions for the data pricing field, and put forward some novel and interesting data pricing topics. This paper aims at gaining better insights, and directing the future research towards practical and sophisticated pricing mechanisms for better data trade and share.

翻译：数据在智能数据分析中扮演着日益重要的角色，推动了许多数据驱动应用的发展。各类数据市场的目标旨在缓解数据孤岛问题，从而促进数据流通。数据定价问题在数据交易中不可或缺但充满挑战。本文对现代数据定价方案进行了全面综述。我们将数据定价方案划分为三大策略和十三种模型，包括查询定价策略、基于特征的数据定价策略以及机器学习中的定价策略。这是迄今为止首次对如此多的现有数据定价模型进行分类。此外，我们不仅详细阐述了每种定价策略下的具体十三种定价模型，还对这些模型进行了深入分析。我们还总结了数据定价领域的五个研究方向，并提出了一些新颖且有趣的数据定价课题。本文旨在获得更深入的见解，并指导未来研究走向实用且复杂精密的定价机制，以实现更好的数据交易与共享。