Description-Enhanced Label Embedding Contrastive Learning for Text Classification

Zhang, Kun; Wu, Le; Lv, Guangyi; Chen, Enhong; Ruan, Shulan; Liu, Jing; Zhang, Zhiqiang; Zhou, Jun; Wang, Meng

doi:10.1109/tnnls.2023.3282020

Cited by 5 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dalam metode ini, setiap kategori atau label diubah menjadi vektor biner dengan panjang yang setara dengan jumlah total kategori yang berbeda dalam variabel tersebut. Semua elemen vektor memiliki nilai nol, kecuali indeks yang sesuai dengan kategori, yang diwakili oleh angka 1 [14].…”

Section: One-hot Encodingunclassified

Studi Kasus Feature Engineering Untuk Data Teks: Perbandingan Label Encoding dan One-Hot Encoding Pada Metode Linear Regresi

Herdian,

Kamila,

Agung Musa Budidarma

2024

Technologia

View full text Add to dashboard Cite

Di dalam pemodelan pembelajaran mesin (Machine Learning), data terbagi menjadi jenis data numerik dan jenis data teks. Tetapi Machine Learning lebih cenderung efektif dalam mengenali pola pada jenis data numerik karena algoritma Machine Learning, terutama yang berbasis statistik dan matematika, dirancang untuk memproses dan menganalisis data numerik. Sehingga bentuk data teks harus dirubah ke dalam bentuk data numerik yang merupakan bagian dari Feature Engineering. Pada penelitian ini, peneliti membanding sebuah hasil akurasi dari prediksi Machine Learning yaitu linear regresi pada teks label data yang telah dilakukan perubahan menjadi numerik dengan metode Feature engineering Label Encoding dan juga Feature Engineering One-Hot Encoding. Pada penelitian ini didapatkan hasil R-Square untuk Label Encoding 0.54 dan R-Squared untuk One-Hot Encoding 0.85 (hasil One-Hot Encoding lebih baik). Sehingga tentu saja yang harus dipilih untuk model yang dibuat adalah Feature Engineering One-Hot Encoding. Untuk kedepannya bisa dilakukan pengujian dengan metode lain untuk merubah data teks menjadi numerik seperti Bags of Words (BoW), Term Frequency-Inverse Document Frequency (TF-IDF), dan yang lainnya.

show abstract

Section: One-hot Encodingunclassified

Studi Kasus Feature Engineering Untuk Data Teks: Perbandingan Label Encoding dan One-Hot Encoding Pada Metode Linear Regresi

Herdian,

Kamila,

Agung Musa Budidarma

2024

Technologia

View full text Add to dashboard Cite

show abstract

“…introduce an attention mechanism that measures the compatibility of embeddings of input and labels. Additional information can be incorporated in learning label embeddings, such as label hierarchy (Chatterjee et al, 2021;Zhang et al, 2022a;Miyazaki et al, 2019) and textual description of labels (Zhang et al, 2023).…”

Section: Related Workmentioning

confidence: 99%

Towards Unification of Discourse Annotation Frameworks

Fu¹

2022

Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop

View full text Add to dashboard Cite

Discourse information is difficult to represent and annotate. Among the major frameworks for annotating discourse information, RST, PDTB and SDRT are widely discussed and used, each having its own theoretical foundation and focus. Corpora annotated under different frameworks vary considerably. To make better use of the existing discourse corpora and achieve the possible synergy of different frameworks, it is worthwhile to investigate the systematic relations between different frameworks and devise methods of unifying the frameworks. Although the issue of framework unification has been a topic of discussion for a long time, there is currently no comprehensive approach which considers unifying both discourse structure and discourse relations and evaluates the unified framework intrinsically and extrinsically. We plan to use automatic means for the unification task and evaluate the result with structural complexity and downstream tasks. We will also explore the application of the unified framework in multi-task learning and graphical models.

show abstract

“…Nevertheless, task-specific label descriptions are often insufficient in real-world scenarios. Subsequently, Zhang et al [16] proposed a description-enhanced label embedding comparative learning method, which integrates external knowledge to obtain label description information. However, they acknowledged that the introduction of external knowledge may lead to unexpected fine-grained noise issues and therefore designed an interaction module to filter out the noise.…”

Section: Introductionmentioning

confidence: 99%

Innovative Label Embedding for Food Safety Comment Classification: Fusion of Self-Semantic and Self-Knowledge Features

Zhang,

Liu,

Feng

et al. 2024

HighTech. Innov. J.

View full text Add to dashboard Cite

Food safety comment classification represents a specialized task within the realm of text classification. The objective is to efficiently identify a large volume of food safety comments, aiding relevant authorities in timely food analysis and safety alerts. Traditional methods typically employ one-hot encoding for label processing. However, in real-world situations, classified labels often convey valuable semantic information and guidance. This paper introduces an innovative approach to enhance the classification performance of food safety comments by embedding label information. Initially, we extracted generic sentiment pivot words from various classification labels as label description information. Subsequently, we employ a joint embedding approach to integrate this label description information into the text. This process will pool the expressions of the pivot word into the corresponding sentiment labels in the known domains after averaging to get the embedded expression. This aims to acquire highly detailed self-semantic feature vectors and self-knowledge feature vectors that are integrated with labeled descriptive information. Then, feed the semantic representation of comments and the word-embedded representation of labeled description information into a time-step-based multilayer Bi-LSTM and a step-based multilayer CNN, respectively. Ultimately, we concatenate these two feature vectors to facilitate matching, thereby fusing the self-semantic and self-knowledge features of labeled description information to train a classification model for food safety comments. Experimental results on the food safety comment dataset showcase a noteworthy improvement of 1.74% and 1.27% in Macro_Precision and Macro_F1 metrics, respectively, compared to BERT, BERT-RNN, and BERT-CNN. Through extensive ablation experiments and additional studies, our method effectively embeds labeling information, demonstrating a clear advantage over traditional methods in the task of classifying food safety comments. Doi: 10.28991/HIJ-2024-05-01-013 Full Text: PDF

show abstract

Description-Enhanced Label Embedding Contrastive Learning for Text Classification

Cited by 5 publications

References 0 publications

Studi Kasus Feature Engineering Untuk Data Teks: Perbandingan Label Encoding dan One-Hot Encoding Pada Metode Linear Regresi

Studi Kasus Feature Engineering Untuk Data Teks: Perbandingan Label Encoding dan One-Hot Encoding Pada Metode Linear Regresi

Towards Unification of Discourse Annotation Frameworks

Innovative Label Embedding for Food Safety Comment Classification: Fusion of Self-Semantic and Self-Knowledge Features

Contact Info

Product

Resources

About