A comparison of extrinsic clustering evaluation metrics based on formal constraints

By Enrique Amigó, Julio Gonzalo, J. Artiles, M. Felisa Verdejo. Information retrieval (Boston), 2008

884 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Information Systems, Signal Processing

View Paper: https://www.semanticscholar.org/paper/af2636dd2ec5a2777f69913b7a8bc3229440c42b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Information Systems

In Signal Processing

Citations in the corpus

Time-series clustering - A decade review (2015, background): “Based on these approaches, many indices are presented in different articles [217,221].”
Learning similarity functions for topic detection in online reputation monitoring (2014, methodology): “Note that, in clustering tasks, R&S are equivalent to the well-known BCubed Precision and Recall measures [3].”
Asking without Telling: Exploring Latent Ontologies in Contextual Representations (2020, background): “B 3 is argued to have favorable properties (Amig´o et al., 2009) and allows for label-wise scoring by restricting to points with speciﬁc gold labels.”
Incomplete Multiview Spectral Clustering With Adaptive Graph Learning (2020, background): “There are many criterions to evaluate the clustering performances of different methods [52].”
MagFace: A Universal Representation for Face Recognition and Quality Assessment (2021, methodology): “F-measure [3] are employed as the evaluation metrics.”
A survey on data stream clustering and classification (2015, background): “The BCubed measure, which computes the average correctness over the dataset, is stated to meet the above constraints at the same time [13].”
IARPA Janus Benchmark-B Face Dataset (2017, positive): “4 report the BCubed Precision and Recall [2] as well as Fmeasure for each clustering sub-protocol.”
An empirical evaluation of zero resource acoustic unit discovery (2017, positive): “Purity and B-Cubed F1 score [26, 27] are used as evaluation metrics.”
People reidentification in surveillance and forensics (2013, background): “However, the same metrics adopted for clustering evaluation could potentially be introduced [Amigó et al. 2009].”
Deep Comprehensive Correlation Mining for Image Clustering (2019, background): “BCubed [1] is a metric to evaluate the quality of partitions in clustering.”
Towards Meaningful Statements in IR Evaluation: Mapping Evaluation Measures to Interval Scales (2021, background): “[6, 7, 8, 9] and Sebastiani [81] face this issue from a formal and theoretical point of view, applying it to various tasks such as ranking, filtering, diversity and quantification, while Moffat [64] adopts a more numerical approach.”
Global-Local GCN: Large-Scale Label Noise Cleansing for Face Recognition (2020, methodology): “The clustering metrics BCubed precision (P), recall (R) and F1-measure (F) [4] are adapted for the cleansing tasks.”
Learning similarity metrics for event identification in social media (2010, background): “Figure 3: NMI and B-Cubed scores on the Upcoming and Last.fm test datasets for All-Text (A), Tags (B), ENS-PART (C), ENS-SIM (D), CLASS-SVM (E), and CLASS-LR (F).”
IARPA Janus Benchmark - C: Face Dataset and Protocol (2018, methodology): “BCubed precision, recall, and F-measure values are reported for the clustering protocol [1].”