Dynamic semantic textual document clustering using frequent terms and named entity
Wael M. S. Yafooz, Siti Z. Z. Abidin, Nasiroh Omar, Rosenah A. Halim
Abstract
Wael M. S. Yafooz, Siti Z. Z. Abidin, Nasiroh Omar, Rosenah A. Halim
Abstract
Data is mostly stored in digital format rather than hard copy because the former is safer, more secure, smaller in size, and faster to retrieve than the latter. With the increasing number of electronic documents to be organized for users to obtain knowledge and integrate information, document clustering has been applied by grouping textual documents based on their similarities. Many attempts have been made to perform textual document clustering with highly accurate results (i.e., close to nature classes) and high processing performance. However, such proposed techniques work in batch (or static) mode in which performance tend to be sacrificed with the use of all the terms in the document, at times resulting in overlapping or scalability issues. Few studies that focus on dynamic clustering also reported on performance issues. This paper contributes in the investigation of textual document clustering approaches and highlights the importance of using dynamic clustering in mining frequent terms with included named entity. This method is used to achieve high efficiency and high-quality data clustering. The method is also beneficial to be used in textual document clustering algorithms for many text domain applications.
OpenAlex reports 9 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Data is mostly stored in digital format rather than hard copy because the former is safer, more secure, smaller in size, and faster to retrieve than the latter. With the increasing number of electronic documents to be organized for users to obtain knowledge and integrate information, document clustering has been applied by grouping textual documents based on their similarities. Many attempts have been made to perform textual document clustering with highly accurate results (i.e., close to nature classes) and high processing performance. However, such proposed techniques work in batch (or static) mode in which performance tend to be sacrificed with the use of all the terms in the document, at times resulting in overlapping or scalability issues. Few studies that focus on dynamic clustering also reported on performance issues. This paper contributes in the investigation of textual document clustering approaches and highlights the importance of using dynamic clustering in mining frequent terms with included named entity. This method is used to achieve high efficiency and high-quality data clustering. The method is also beneficial to be used in textual document clustering algorithms for many text domain applications.
Key concepts: Document clustering, Computer science, Cluster analysis, Information retrieval, Clustering high-dimensional data, Brown clustering, Scalability, Data mining