Crowdsourcing is a practice where individuals, organizations, or companies seek to obtain ideas, feedback, solutions, or contributions from a large group of people. This concept leverages the collective intelligence and diversity of the crowd, enabling it to solve complex problems, generate new ideas, create content, or gather significant amounts of data. However, traditional crowdsourcing platforms have privacy concerns as sensitive data might be at risk of exposure. Moreover, a lack of transparency may further lead to mistrust and hesitance among potential contributors.

This dissertation presents algorithmic design for building blockchain-based systems and privacy-preserving crowdsourcing applications. First, we design a crowdsourcing system atop a public blockchain where a crowdsourcing task can involve an unlimited number of data providers but with a constant transaction cost. Second, we propose an off-chain payment system which aims to achieve high-throughput and collateral-efficient instant payment for routine purchase. Third, we introduce a privacy-preserving framework designed for reliable aggregation of crowdsourced text data. Lastly, we build a cryptocurrency blocklisting service which supports private and highly efficient blocklist query scheme, as well as a framework for shareholders to evaluate the quality of blocklists while suppressing individual biasing and coercive manipulation. The presented research would improve the adoption of crowdsourcing and shed light on the integration with blockchain technologies.