- This release introduces a new mechanism by which users can provide feedback on the accuracy of the metadata ChemMeta extracts from each peer-reviewed article contributing to the general acceptance score for a hypothesis of bodily injury
ChemMeta summarizes the scientific evidence investigating more than 33,000 chemicals and drugs across some 13,000 hypotheses of bodily injury. The peer-reviewed articles underlying these hypotheses number in the hundreds of thousands and the metadata (e.g., relevance, outcome, study subject) extracted from the majority of these articles has been accomplished via machine-learning algorithm rather than human analysis.
This release introduces a mechanism by which users can provide feedback on the accuracy of the metadata extracted from each article. Article feedback can be provided in the literature explorer modal, which is accessed by clicking on the general acceptance score for a hypothesis of interest in the chemical profile page (1). General feedback about ChemMeta can still be provided via the Feedback link located on the bottom right of every page (2).
Within the literature explorer modal, users can choose either to provide feedback about an article that they believe is missing from the hypothesis (3) or about an article they believe has been coded incorrectly (4). An indication is provided if the user or a colleague from their company has already provided feedback about an article in the literature table (5).
Users providing feedback about a specific article can indicate the nature of the feedback (e.g., the article is not relevant, the outcome or study subject is incorrect) via buttons (6), in free text (7), or both.
Praedicat's response along with the user's original feedback is posted in the right-hand feedback pane (8). Praedicat endeavors to respond to user feedback within two business days and correct miscoded articles within two content releases.