What the F-measure doesn't measure: Features, Flaws, Fallacies and Fixes

03/22/2015

∙

The F-measure or F-score is one of the most commonly used single number measures in Information Retrieval, Natural Language Processing and Machine Learning, but it is based on a mistake, and the flawed assumptions render it unsuitable for use in most contexts! Fortunately, there are better alternatives.

READ FULL TEXT

What the F-measure doesn't measure: Features, Flaws, Fallacies and Fixes

Lecture Notes on Neural Information Retrieval

Report on the 3rd Joint Workshop on Bibliometric-enhanced Information Retrieval and Natural Language Processing for Digital Libraries (BIRNDL 2018)

Information retrieval in folktales using natural language processing

Using Local Optimality Criteria for Efficient Information Retrieval with Redundant Information Filters

Inter-Annotator Agreement in the Wild: Uncovering Its Emerging Roles and Considerations in Real-World Scenarios

Explainability of Text Processing and Retrieval Methods: A Critical Survey

ATOL: Measure Vectorisation for Automatic Topologically-Oriented Learning

What the F-measure doesn't measure: Features, Flaws, Fallacies and Fixes

Related Research

Lecture Notes on Neural Information Retrieval

Report on the 3rd Joint Workshop on Bibliometric-enhanced Information Retrieval and Natural Language Processing for Digital Libraries (BIRNDL 2018)

Information retrieval in folktales using natural language processing

Using Local Optimality Criteria for Efficient Information Retrieval with Redundant Information Filters

Inter-Annotator Agreement in the Wild: Uncovering Its Emerging Roles and Considerations in Real-World Scenarios

Explainability of Text Processing and Retrieval Methods: A Critical Survey

ATOL: Measure Vectorisation for Automatic Topologically-Oriented Learning