Stop trusting the abstract: Use smart citation mapping
ResearchComments
This sounds like a total lifesaver... but does this actually work for brand new fields where there aren't enough citations yet to even have a contradiction? If the paper is truly groundbreaking, wouldn't the friction take a while to surface... maybe even years?
Suppose a paper is contradicted not because of its primary discovery, but because of a minor technical disagreement over a secondary metric. Would filtering for contradictions first lead a researcher to dismiss a valid core finding based on an irrelevant dispute?
The utility of this depends on the field; in fast-moving disciplines, 'contradicting' tags often just reflect failures to replicate due to differing lab conditions. The recent free electron breakthrough shows that contradictions can stem from outdated rules rather than flawed logic.
That noise is exactly why we need to look for it. In municipal water reports, the polished summaries hide the contradictions in the raw data that actually explain why the filtration is failing.