Multiple Latent Spaces Learning for Cross-Domain Text Classification

Pan, Jianhan; Cui, Teng; Du, Mingjing; Zhang, Qingyang; Song, Bingbing; Qu, Qiaoli

doi:10.1109/acait53529.2021.9730891

Cited by 1 publication

(1 citation statement)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Namely, the commonality refers to the latent information shared across all source domains, which is also held by the target domain and is used as a stable transfer bridge, as well as the specificity refers to the latent information specific to a certain source domain, which matches a part of the target domain and has some power for discriminating document classes. By capturing these attributes, MSTL can achieve better learning performance than traditional single-source transfer learning(SSTL) [3,4]. However, there are some drawbacks to these strategies.…”

Section: Introductionmentioning

confidence: 99%

Multi-source combined transfer learning for text classification

Song,

Pan,

et al. 2023

Third International Conference on Advanced Algorithms and Neural Networks (AANN 2023)

Self Cite

View full text Add to dashboard Cite

The objective of transfer learning is to learn well-established knowledge from label-rich training data to handle learning tasks on the target domain, where the labeled samples can be hardly collected and the data distribution is different from that of the training data. Nowadays, the label information in the target domain may exist in multiple different source domains, but none of which can cover all the information. To make use of the imperfect source domains, many novel transfer learning algorithms explore potential properties among them to improve the completeness of transferred knowledge. However, these works only build on the original source domains, limiting the possibility to uncover additional supplementary information. In this paper, we propose novel multi-source combined transfer learning(MSCTL). MSCTL first replaces original source domains with a new training set space which can provide richer latent information by correlating multiple source domains. Then, through a triplex-complementary strategy, MSCTL extracts and integrates threefold concepts from the space to make the knowledge complementary and integral. Furthermore, we design MSCTL as an optimization problem and provide an iterative algorithm to solve it. Finally, massive experiments are performed to verify the effectiveness of our method.

show abstract

Section: Introductionmentioning

confidence: 99%