Scientific communities usually have a strong publication bias towards works with positive results, and the Semantic Web is not an exception. However, negative or inconclusive results are fundamental for the progress of the research areas and are as valuable as positive results.

This workshop provides a forum to discuss non-confirmatory results, their role in the progress of the Web Semantic research field, and methodologies and best practices to publish them. Thus, NoISE aims at diminishing the taboo of reporting negative results, and encouraging researchers to share experimental protocols, applied methodologies, and documented approaches that refute their research goals.


Nowadays, there is a strong positive publication bias: approaches reporting positive, and significant, results are more likely to be published at conferences and journals. Statistically significant outcome, e.g., from benchmarking, have higher chances of being fully reported, discarding knowledge about what is proven false or not working. However, such testimonies are fundamental in scientific research and occur more frequently than positive results.

A considerable effort is required to evaluate alternative approaches that not always end up overcoming the state-of-the-art in all or certain aspects. Unfortunately, these non-confirmatory observations are not always publicly documented, causing other researchers to repeat them over and over again. This creates a huge overhead in our community’s progress.

Well documented negative or inconclusive results should

  • reveal fundamental flaws and obstacles in commonly used methods,
  • act as catalyst for formalizing empirical knowledge, and
  • gain insights that contribute to the progress of Semantic Web technologies.

However, publishing such negative results currently lacks incentives, as authors may never get attributed or cited.

This workshop addresses the way Semantic Web research deals with the insufficient evidence of negative results, following the trend, as other disciplines do, to this fight against negative publication bias. The uppermost goal is to turn intuitive knowledge on Semantic technologies into justifiable arguments and gain formal proofs of sub-optimal solutions.