The disclosed embodiments relate to a computer system that facilitates identifying additional tax information. During operation, the computer system receives tax information. Next, the computer system determines relevance of the additional tax information to the tax information using a metric indicative of a statistical relationship between the tax information and the additional tax information in a tax-information data structure. This metric is based on co-existence of the tax information and the additional tax information at different hierarchical levels in income-tax documents. Moreover, the tax-information data structure includes tax information and associated context information from income-tax documents. Next, the computer system identifies the additional tax information based on the relevance and a threshold, and the computer system provides the additional tax information.