Centroid Terms as Text Representatives

Kubek, Mario; Unger, Herwig

doi:10.1145/2960811.2967150

Cited by 34 publications

(6 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The distance between two terms is determined by the inverse of the DICE-coefficient [21], [22]. The cluster center itself is represented by its centroid term as defined in [23], [24]. A cluster center is formed by the node with the shortest average distance ∆d to every other node in the co-occurrence graph.…”

Section: Novel Graph-based Clustering Algorithms 221 Dynamic Clustering For Segregation Of Co-occurrence Graphs (Dcsg)mentioning

confidence: 99%

Graph-based Clustering Algorithms – A Review on Novel Approaches

Hloch¹,

Kubek²,

Unger³

2021

Adv. sci. technol. eng. syst. j.

Self Cite

View full text Add to dashboard Cite

Classical clustering algorithms often require an a-priori number of expected clusters and the presence of all documents beforehand. From practical point of view, the use of these algorithms especially in more dynamic environments dealing with growing or shrinking corpora therefore is not applicable. Within the last years, graph-based representations of knowledge such as co-occurrence graphs of document corpora have gained attention from the scientific community. Accordingly, novel unsupervised and graph-based algorithms have been recently developed in order to group similar topics, represented by documents or terms, in clusters. The conducted work compares classical and novel graph-based algorithms, showing that classical clustering algorithms in general perform faster than graph-based clustering algorithms. Thus, the authors' focus is to show that the graph-based algorithms provide similar clustering results without requiring an hyperparamter k to be determined a-priori. It can be observed that the identified clusters exhibit an associative relationship reflecting the topical and sub-topical orientation. In addition, it is shown in a more in-depth investigation that the Seqclu (sequential clustering algorithm) can be optimized performance-wise without loss of clustering quality.

show abstract

Section: Novel Graph-based Clustering Algorithms 221 Dynamic Clustering For Segregation Of Co-occurrence Graphs (Dcsg)mentioning

confidence: 99%

Graph-based Clustering Algorithms – A Review on Novel Approaches

Hloch¹,

Kubek²,

Unger³

2021

Adv. sci. technol. eng. syst. j.

Self Cite

View full text Add to dashboard Cite

show abstract

“…If (w a , w b )E (i.e. the words involved are co-occurrents) their distance d(w a , w b ) is easily to be defined as (1) Lecture Notes on Information Theory Vol. 5, No.…”

Section: Fundamentalsmentioning

confidence: 99%

“…Although -differing from semantic approaches-the assigned centroid terms may not represent any semantic meaning of the given text, they are in each case a formally calculable, well-balanced extract of the words used in the text and their content relations. This approach has been used to define the distance of documents [1] and to determine cluster hierarchies [2], too. Additionally, centroids may be used to detect topical shifts, i.e.…”

Section: Hierarchies Of Centroidsmentioning

confidence: 99%

“…Text centroids -inspired from the centre of mass in physics-have been introduced in [1] to represent sentences, paragraphs or whole texts by a single representing term. In addition, it could be shown that a distance measure among centroids may be defined which can be used to determine sematic text similarities and distances as well as to derive a hierarchical clustering algorithm [2] based on them.…”

Section: Introductionmentioning

confidence: 99%

“…Document vectors in this lower-dimensional space can then be compared in the same manner. The new, centroid-based method presented in [1] represents texts by a single centroid term (which is not necessarily contained in the document), which must be usually calculated, only once. Then, every comparison operation is just a single distance measurement on the respective co-occurrence graph, which can be considered a condensed, compact and easily extendible representation of the knowledge of an entity (e.g.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Empiric Experiments with Text Representing Centroids

Kubek¹,

Böhme²,

Unger³

2017

LNIT

Self Cite

View full text Add to dashboard Cite

Centroid terms are comfortable instruments to represent texts, compare them semantically and to even (hierarchically) cluster sets of documents using them. Their determination depends on their topical and conceptual context, i.e. the dynamically changing knowledge of a user represented by the co-occurrence graph. Herein, important properties of centroids as well as their applicability for tasks in natural language processing and text mining shall be discussed and their use justified by a set of experiments. Based on the obtained results, a new approach to detect fine-grained similarities between text documents is derived. 

show abstract

State-of-the-Art Survey on Web Search

Roßrucker

2022

Studies in Big Data

View full text Add to dashboard Cite

Centroid Terms as Text Representatives

Cited by 34 publications

References 6 publications

Graph-based Clustering Algorithms – A Review on Novel Approaches

Graph-based Clustering Algorithms – A Review on Novel Approaches

Empiric Experiments with Text Representing Centroids

State-of-the-Art Survey on Web Search

Contact Info

Product

Resources

About