Evaluating the Accuracy of the Labeling System in Web of Science for the Sustainable Development Goals

The dataset is constructed based on 2,434 citation topics from the Web of Science (WoS) database. The data was extracted on March 1, 2024, covering a period from 2015 to 2023, aligning with the United Nations' Sustainable Development Goals (SDGs) timeline, which was adopted in 2015 with a target achievement date of 2030. The dataset encompasses a substantial collection of publications and focuses on 10 out of the 17 SDGs, specifically Poverty (01), Hunger (02), Health (03), Education (04), Energy (07), Cities (11), Consumption (12), Climate (13), Water (14), and Land (15). Through meticulous analysis, it was found that within this timeframe, 1,807 citation topics were successfully linked to one or more of these 10 SDGs.In addition to the direct linking, a Bergen search approach was applied to examine the semantic relevance between the topics and SDG targets. The dataset includes the following columns:<b>Citation Topics Micro</b>: Coherent clusters of documents-a document can belong to only a single micro-topic.<b>Citation Topics Meso</b>: Micro-topics are clustered into larger meso-topics using similar rules.<b>Citation Topics Macro</b>: Meso-topics are brought together into broad macro-topics.<b>WoS</b>: The number of papers associated with the citation topic within the specified timeframe.<b>SDG XX_Action</b>: The number of papers associated with the citation topic that were retrieved using the Bergen action-approach search within the specified timeframe.<b>SDG XX_Topic</b>: The number of papers associated with the citation topic that were retrieved using the Bergen topic-approach search within the specified timeframe.<b>SDG XX_Wos</b>: The number of papers associated with the citation topic that were marked as related to sustainable development by the Web of Science within the specified timeframe.