Sources of data for the article "Between "pravda" and "istina": Conceptualizing Truth in Putin and Medvedev's Political Communication

The Dekoder corpus was originally provided as a structured JSON dataset. Prior to analysis, the data were preprocessed using a Python script developed for the purposes of this study. The script parsed the JSON records and recursively extracted textual material from the relevant fields, including text, utterance, content, and line, while excluding non-textual metadata. Recursive extraction was necessary because relevant textual material could occur at different levels of the nested JSON structure. The extracted texts were subsequently consolidated into a single UTF-8 encoded plain-text file, which constituted the working corpus for the subsequent analysis. The resulting corpus was checked against the original dataset to verify that relevant textual material had been retained and that no unintended duplication had occurred. Invalid or malformed JSON records were excluded during the parsing process. The assigned values to extract were "правда" and "истина", along with their immediate context within a +/-5 words. This is the link the original documents from which Python extracted the terms, and concrete extracts can be further found by accessing each page individually and scanning it.