The dataset consists of two main sections related to the language analysis of the latest 10-K or 10-Q financial reports. The first section contains language metrics from the most recent report, including financial sentiment, the percentage of words related to the financial domain by language type (constraining, interesting, litigious, uncertainty), readability scores, lexical metrics, and text statistics. The second part captures the changes between the two most recent reports for each company. The analysis is done for the whole report and for specific sections, like Risk Factors and MD&A.