A simple and efficient text matching model based on deep interaction

Yu, Chuanming; Xue, Haodong; Jiang, Yifan; Luo, An; Li, Gang

doi:10.1016/j.ipm.2021.102738

Cited by 23 publications

(12 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The MAP(%) and MRR(%) of the model on the WikiQA test set.Results marked with l are reported by Tay et al[35], m are reported by Yu et al[36] , e are reported by Wang et al[24], f are reported by Chen et al[25], g are reported by Yang et al[13] …”

mentioning

confidence: 65%

DEIM: An effective deep encoding and interaction model for sentence matching

Jiang¹,

Zhao²,

Cui³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

mentioning

confidence: 65%

DEIM: An effective deep encoding and interaction model for sentence matching

Jiang¹,

Zhao²,

Cui³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…To evaluate the effectiveness of the proposed model, we compare MGMSN with six current mainstream text matching algorithms, they are Deep Structured Semantic Model (DSSM) ( Huang et al (2013) ), Siamese LSTM network (Siamese LSTM) ( Neculoiu et al (2016) ), Attention-Based Convolutional Neural Network for Modeling Sentence Pairs (ABCNN) ( Yin et al (2016) ), Enhance Sequential Inference Model (ESIM) ( Greff et al (2016) ), Deep Interactive Text Matching (DITM) model ( Yu C. et al (2021) ), and Frame-based Multi-level Semantics Representation (FMSR) model ( Guo et al (2021) ).…”

Section: Experiments Results and Analysismentioning

confidence: 99%

MGMSN: Multi-Granularity Matching Model Based on Siamese Neural Network

Wang

Yang

2022

Front. Bioeng. Biotechnol.

View full text Add to dashboard Cite

show abstract

“…DITM (Yu et al, 2021): The model utilizes a multiple iterative interaction module coupled with a multiperspective pooling layer to capture the interaction information between short texts comprehensively.…”

Section: Methodsmentioning

confidence: 99%

“…Due to the length limitation of the texts, existing models for short‐text matching cannot accurately understand the semantic meaning of critical words with limited contextual information, leading to decreased performance. Compared with the interaction‐based approaches, for example, ESIM (Chen et al, 2017), RE2 (Yang et al, 2019), ESAN (Hu et al, 2020), and DITM (Yu et al, 2021), the proposed model has significantly improved the performance of short‐text matching due to external knowledge.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A lightweight semantic‐enhanced interactive network for efficient short‐text matching

Xue

Luo

et al. 2022

Asso for Info Science & Tech

Self Cite

View full text Add to dashboard Cite

Knowledge‐enhanced short‐text matching has been a significant task attracting much attention in recent years. However, the existing approaches cannot effectively balance effect and efficiency. Effective models usually consist of complex network structures leading to slow inference speed and the difficulties of applications in actual practice. In addition, most knowledge‐enhanced models try to link the mentions in the text to the entities of the knowledge graphs—the difficulties of entity linking decrease the generalizability among different datasets. To address these problems, we propose a lightweight Semantic‐Enhanced Interactive Network (SEIN) model for efficient short‐text matching. Unlike most current research, SEIN employs an unsupervised method to select WordNet's most appropriate paraphrase description as the external semantic knowledge. It focuses on integrating semantic information and interactive information of text while simplifying the structure of other modules. We conduct intensive experiments on four real‐world datasets, that is, Quora, Twitter‐URL, SciTail, and SICK‐E. Compared with state‐of‐the‐art methods, SEIN achieves the best performance on most datasets. The experimental results proved that introducing external knowledge could effectively improve the performance of the short‐text matching models. The research sheds light on the role of lightweight models in leveraging external knowledge to improve the effect of short‐text matching.

show abstract

A simple and efficient text matching model based on deep interaction

Cited by 23 publications

References 25 publications

DEIM: An effective deep encoding and interaction model for sentence matching

DEIM: An effective deep encoding and interaction model for sentence matching

MGMSN: Multi-Granularity Matching Model Based on Siamese Neural Network

A lightweight semantic‐enhanced interactive network for efficient short‐text matching

Contact Info

Product

Resources

About