A novel textual data augmentation method for identifying comparative text from user-generated content

Wei, Na; Zhao, Shengchuan; Liu, Jing; Wang, Shenghui

doi:10.1016/j.elerap.2022.101143

Cited by 9 publications

(2 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, although #1 in Table 4 lacks a keyword, it implies comparative information. On the other hand, CSR and SVM generally predict non-comparative text including selected keywords or rules as comparative text [53], such as the example #2 in Table 4. While deep learning-based methods have the ability to correctly identify them.…”

Section: Resultsmentioning

confidence: 99%

A review for comparative text mining: From data acquisition to practical application

Wei

Zhao

Liu

et al. 2023

Journal of Information Science

Self Cite

View full text Add to dashboard Cite

Social media provides customers with great opportunities to share their opinions regarding certain products and services. Comparative text, as an important expression form, deserves further exploration, since it contains considerable comparative information between different products and services. In this study, we review existing research on Comparative Text Mining (CTM) in the past 16 years. Basic concepts related to CTM are first described, and a general research framework is subsequently proposed. We then dive into each component of the research framework, ranging from data acquisition, comparative text identification (CTI), comparative relation extraction (CRE), to potential applications. In addition, we conduct extensive experimental analysis on existing methods for CTI and CRE, and clarify their limitations. Accordingly, we provide corresponding insights, and point out future research directions.

show abstract

Section: Resultsmentioning

confidence: 99%

A review for comparative text mining: From data acquisition to practical application

Wei

Zhao

Liu

et al. 2023

Journal of Information Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…Shorten and Khoshgoftaar et al [25] used a deletion method with word lists and randomized deletion probabilities, to randomly delete some words in a sentence by deactivating the word list and preset probability. However, these methods are based on simple rules or dictionaries, which lack a deep understanding of the text and make it difficult to provide additional effective information to assist image classification [26]. Other approaches try different strategies.…”

Section: Related Workmentioning

confidence: 99%

Cross‐modal knowledge learning with scene text for fine‐grained image classification

Xiong,

Mao,

Wang

et al. 2024

IET Image Processing

View full text Add to dashboard Cite

Scene text in natural images carries additional semantic information to aid in image classification. Existing methods lack full consideration of the deep understanding of the text and the visual text relationship, which results in the difficult to judge the semantic accuracy and the relevance of the visual text. This paper proposes image classification based on Cross modal Knowledge Learning of Scene Text (CKLST) method. CKLST consists of three stages: cross‐modal scene text recognition, text semantic enhancement, and visual‐text feature alignment. In the first stage, multi‐attention is used to extract features layer by layer, and a self‐mask‐based iterative correction strategy is utilized to improve the scene text recognition accuracy. In the second stage, knowledge features are extracted using external knowledge and are fused with text features to enhance text semantic information. In the third stage, CKLST realizes visual‐text feature alignment across attention mechanisms with a similarity matrix, thus the correlation between images and text can be captured to improve the accuracy of the image classification tasks. On Con‐Text dataset, Crowd Activity dataset, Drink Bottle dataset, and Synth Text dataset, CKLST can perform significantly better than other baselines on fine‐grained image classification, with improvements of 3.54%, 5.37%, 3.28%, and 2.81% over the best baseline in mAP, respectively.

show abstract

MVFSIGM: multi-variant feature-based synthesis image generation model for improved stability using generative adversarial network

Vathana

Babu

2023

Soft Comput

View full text Add to dashboard Cite

A novel textual data augmentation method for identifying comparative text from user-generated content

Cited by 9 publications

References 44 publications

A review for comparative text mining: From data acquisition to practical application

A review for comparative text mining: From data acquisition to practical application

Cross‐modal knowledge learning with scene text for fine‐grained image classification

MVFSIGM: multi-variant feature-based synthesis image generation model for improved stability using generative adversarial network

Contact Info

Product

Resources

About