A Comparative Study of Classification and Clustering Methods from Text of Books

Probierz, Barbara; Kozák, Jan; Hrabia, Anita

doi:10.1007/978-3-031-21967-2_2

Cited by 3 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All book texts were processed according to the steps described in Section 4. We compared our results to the algorithm used in [15]. The results showed that the content of the set had a large impact on the classification results.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…The next step (see Step 4 on Figure 1) was to create a binary matrix, containing information as to whether a given element of the set appeared in the text of the book ("1"-if it occurs in the text, or "0"-if it does not occur). Since we have proven in our previous research that better results are obtained using term weighting (TF) compared to binary [15], we decided to use this fact in our approach as well. We also created matrices in which we weighted the frequency of occurrence of the elements of the set.…”

Section: Methodsmentioning

confidence: 99%

“…The problem of book classification was also raised in [15], where a system was proposed that successfully assigned books to the appropriate category, and in [16], in which BERT (Bidirectional Encoder Representations from Transformers) models supported automatic subject indexing for digital library collections. In contrast, T. Betts et al [17] investigated the usefulness of information extraction techniques in a text categorization task, automatically extracting structured information from the full text of books.…”

Section: Related Workmentioning

confidence: 99%

“…Our text representation was compared with the classic vector approach-binary, TF and T-IDF, as used in [15]. In the case of the classical approach, the matrices we obtained contained 63,280 features.…”

Section: Text Representationmentioning

confidence: 99%

See 3 more Smart Citations

A New Method for Graph-Based Representation of Text in Natural Language Processing

2023

Self Cite

View full text Add to dashboard Cite

Natural language processing is still an emerging field in machine learning. Access to more and more data sets in textual form, new applications for artificial intelligence and the need for simple communication with operating systems all simultaneously affect the importance of natural language processing in evolving artificial intelligence. Traditional methods of textual representation, such as Bag-of-Words, have some limitations that result from the lack of consideration of semantics and dependencies between words. Therefore, we propose a new approach based on graph representations, which takes into account both local context and global relationships between words, allowing for a more expressive textual representation. The aim of the paper is to examine the possibility of using graph representations in natural language processing and to demonstrate their use in text classification. An innovative element of the proposed approach is the use of common cliques in graphs representing documents to create a feature vector. Experiments confirm that the proposed approach can improve classification efficiency. The use of a new text representation method to predict book categories based on the analysis of its content resulted in accuracy, precision, recall and an F1-score of over 90%. Moving from traditional approaches to a graph-based approach could make a big difference in natural language processing and text analysis and could open up new opportunities in the field.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Text Representationmentioning

confidence: 99%

See 2 more Smart Citations

A New Method for Graph-Based Representation of Text in Natural Language Processing

2023

Self Cite

View full text Add to dashboard Cite

show abstract

Enhancing Cross‐Domain Book Classification Through Caching‐Enabled Networks and Transformer Technology

Li,

Zheng,

Bao

et al. 2024

Internet Technology Letters

View full text Add to dashboard Cite

Book classification is a crucial task for libraries and a fundamental aspect of their service offerings. Cross‐domain book classification, in particular, presents significant challenges due to the diversity and complexity of content across different genres and subjects. To tackle these challenges, a user‐oriented strategy employing Transformer network (TN) is developed to fulfill the need for superior content quality and classification. Our proposed method leverages the self‐attention mechanism of TN for precise feature extraction and classification, combining it with principal component analysis to ensure a comprehensive understanding of book content. This integration represents a technical innovation that enhances the model's ability to handle diverse datasets with improved accuracy and robustness. Our approach merges TN with caching‐enabled networks (CEN) to enhance accuracy and robustness. Driven by the necessity for improved cross‐domain classification, our strategy aims to standardize book classifications, thus improving user satisfaction. The primary actions encompass improved classification management, feedback systems, and evaluation frameworks. This work highlights the innovative fusion of TN and CEN, showcasing how these advanced techniques can significantly elevate the performance of library classification systems. Our work demonstrates that high‐quality book classification can significantly improve library services and user experience. Furthermore, it aligns with the broader applications of CEN in emerging networking technologies, showing the potential for cutting‐edge techniques to revolutionize library services.

show abstract