A comparison study of clustering validity indices

Chouikhi, Hasna; Charrad, Malika; Ghazzali, Nadia

doi:10.1109/gscit.2015.7353330

Cited by 35 publications

(28 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…NbClust provides 30 different indices (e.g., Gap statistic or Silhouette) for determining the optimal number of clusters based on a ''majority rule'' approach ( Fig. 4; Chouikhi, Charrad & Ghazzali, 2015). Membership in these clusters was evaluated via Jaccard similarity index comparing, pairwise, all three clusters from each of the three surveys to determine Table 2 Jaccard similarity clusters resulting from K-Means applied to questionnaire results.…”

Section: Clusteringmentioning

confidence: 99%

Data-driven classification of the certainty of scholarly assertions

Prieto

Deus

Waard

et al. 2020

PeerJ

View full text Add to dashboard Cite

The grammatical structures scholars use to express their assertions are intended to convey various degrees of certainty or speculation. Prior studies have suggested a variety of categorization systems for scholarly certainty; however, these have not been objectively tested for their validity, particularly with respect to representing the interpretation by the reader, rather than the intention of the author. In this study, we use a series of questionnaires to determine how researchers classify various scholarly assertions, using three distinct certainty classification systems. We find that there are three distinct categories of certainty along a spectrum from high to low. We show that these categories can be detected in an automated manner, using a machine learning model, with a cross-validation accuracy of 89.2% relative to an author-annotated corpus, and 82.2% accuracy against a publicly-annotated corpus. This finding provides an opportunity for contextual metadata related to certainty to be captured as a part of text-mining pipelines, which currently miss these subtle linguistic cues. We provide an exemplar machine-accessible representation—a Nanopublication—where certainty category is embedded as metadata in a formal, ontology-based manner within text-mined scholarly assertions.

show abstract

Section: Clusteringmentioning

confidence: 99%

Data-driven classification of the certainty of scholarly assertions

Prieto

Deus

Waard

et al. 2020

PeerJ

View full text Add to dashboard Cite

show abstract

“…However, like most clustering algorithms, PAM requires a predefined cluster number of k and cluster centers (medoids) those are difficult to choose. Researchers have presented various clustering strategies [32][33][34][35], but no consensus has been reached even on the definition of a cluster. One simple and effective method was proposed to recognize the clusters (cluster centers and numbers) [36][37][38].…”

Section: Defining the Cluster Centers And Numbersmentioning

confidence: 99%

Clustering Coastal Land Use Sequence Patterns along the Sea–Land Direction: A Case Study in the Coastal Zone of Bohai Bay and the Yellow River Delta, China

Ding

Zhang

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

Specific knowledge about the characteristics of coastal land use along the sea–land direction helps to better understand the spatial heterogeneity of coastal land use, which could offer scientific support for rational land management and the sustainable development of the coastal zone. However, the traditional methods including buffering or the regional method are hard to extract detailed spatial structure and location correlations of coastal land use along the sea–land direction. Therefore, we developed a model, called sequence–based clustering of coastal land use pattern (SCCLUP), to mine the coastal land use sequence patterns (CLUSPs) along the sea–land direction. As a case study in the major coastal zone of Bohai Bay and the Yellow River Delta from 1990 to 2010, we found that: (1) The land use showed a sequential distribution along the sea–land direction. And the land use closed to shoreline and inland boundary had relative stable sequential location along the sea–land direction. However, the middle land uses had dynamic sequential locations that led to multiple CLUSPs; (2) due to the increasing percent of construction land, the artificial level of CLUSPs was continuously increasing and new CLUSPs tended to distribute around port areas. Different CLUSPs with similar land use sequential relationships tended to have similar land use structure along the sea–land direction; (3) the land uses sequential location along the sea–land direction revealed the actual distance of land use to the shoreline and had a tight correlation with environmental factors (salinity, water, and landform). The land use with large increasing and wide adaptivity (like construction land) had a large impact on the changes of CLUSPs in the study area. Therefore, strong control should be provided for the excessive expansion of land use like construction land to limit the over changes in land use pattern along the sea–land direction. Additionally, the spatial heterogeneity of land use along the sea–land direction should be considered to a better understanding of anthropic impacts on the coastal zone.

show abstract

“…NbClust provides 30 different indices (e.g., Gap statistic or Silhouette) for determining the optimal number of clusters based on a "majority rule" approach ( Fig. 4; Chouikhi, Charrad & Ghazzali, 2015). Membership in these clusters was evaluated via Jaccard similarity index comparing, pairwise, all three clusters from each of the three surveys to determine which clusters were most alike ( Table 2).…”

Section: Statistical Analysis Of Agreementmentioning

confidence: 99%

Data-driven classification of the certainty of scholarly assertions

Prieto

Deus

Waard

et al. 2019

Preprint

View full text Add to dashboard Cite

The grammatical structures scholars use to express their assertions are intended to convey various degrees of certainty or speculation. Prior studies have suggested a variety of categorization systems for scholarly certainty; however, these have not been objectively tested for their validity, particularly with respect to representing the interpretation by the reader, rather than the intention of the author. In this study, we use a series of questionnaires to determine how researchers classify various scholarly assertions, using three distinct certainty classification systems. We find that there are three distinct categories of certainty along a spectrum from high to low. We show that these categories can be detected in an automated manner, using a machine learning model, with a cross-validation accuracy of 89.2% relative to an author-annotated corpus, and 82.2% accuracy against a publicly-annotated corpus. This finding provides an opportunity for contextual metadata related to certainty to be captured as a part of text-mining pipelines, which currently miss these subtle linguistic cues. We provide an exemplar machine-accessible representation - a Nanopublication - where certainty category is embedded as metadata in a formal, ontology-based manner within text-mined scholarly assertions.

show abstract

A comparison study of clustering validity indices

Cited by 35 publications

References 24 publications

Data-driven classification of the certainty of scholarly assertions

Data-driven classification of the certainty of scholarly assertions

Clustering Coastal Land Use Sequence Patterns along the Sea–Land Direction: A Case Study in the Coastal Zone of Bohai Bay and the Yellow River Delta, China

Data-driven classification of the certainty of scholarly assertions

Contact Info

Product

Resources

About