2008 IEEE International Conference on Computational Cybernetics 2008
DOI: 10.1109/icccyb.2008.4721382
|View full text |Cite
|
Sign up to set email alerts
|

Comparison of Two Document Clustering Techniques which use Neural Networks

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
2
0
1

Year Published

2017
2017
2024
2024

Publication Types

Select...
3
2
2

Relationship

0
7

Authors

Journals

citations
Cited by 8 publications
(3 citation statements)
references
References 11 publications
0
2
0
1
Order By: Relevance
“…Akar dari pohon atau dendogram menggambarkan jumlah cluster, sementara daunnya akan menggambarkan letak satu elemen atau objek dalam satu cluster tertentu [14]. Dalam penelitian ini, algoritma yang digunakan adalah pendekatan bottom-up manner, yaitu dimulai dari mengasumsikan bahwa jumlah cluster adalah sama dengan jumlah elemen dan pada setiap penambahan height, cluster akan digabungkan hingga jumlah yang diinginkan atau sesuai dengan penentuan k optimal [15].…”
Section: Hierarchical Clusteringunclassified
“…Akar dari pohon atau dendogram menggambarkan jumlah cluster, sementara daunnya akan menggambarkan letak satu elemen atau objek dalam satu cluster tertentu [14]. Dalam penelitian ini, algoritma yang digunakan adalah pendekatan bottom-up manner, yaitu dimulai dari mengasumsikan bahwa jumlah cluster adalah sama dengan jumlah elemen dan pada setiap penambahan height, cluster akan digabungkan hingga jumlah yang diinginkan atau sesuai dengan penentuan k optimal [15].…”
Section: Hierarchical Clusteringunclassified
“…Each section/subsection of the documents is transformed into a vector space, where the entries of vector are TF-IDF values. Finally, cosine similarity [20], is used to investigate hierarchical relevance between the reformulated query (expanded query) and sections/subsections in a Vector Space Model (VSM) model. N -grams (N <= 3) is used to create new tokens from the preprocessed sections, new tokens are included with uni-grams in a Bag of words [24].…”
Section: Document Reconstructionmentioning
confidence: 99%
“…This final metric is a scaled variant of cosine similarity. Such a measure of cosine similarity has wide usage across several fields, of which we cite a few examples . Although this method is the most abstract—and perhaps least intuitive—of the three, we have found it useful in identifying what we have termed “dynamic domains.” These domains are (not necessarily local) sets of residues with similar dynamics, as indicated by their correlated motion vectors.…”
Section: Introductionmentioning
confidence: 99%