PWA-PEM for Latent Tree Model and Hierarchical Topic Detection

Liu, Zhuchen; Chen, Hao; Li, Jie; Yu, Yi

doi:10.1007/978-3-030-00828-4_19

Cited by 1 publication

(2 citation statements)

References 9 publications

(2 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Liu et al [21] proposed a hierarchical latent tree analysis (HLTA) to determine a topic of documents. Balakrishnan and Chopra [22] adopted the method of gradient acceleration optimization to improve the execution efficiency of progressive EM HLTA (PEM-HLTA) for topic detection. References [12], [13], [21], and [22] only considered how to build LTM with a large amount of data.…”

Section: Updating Latent Tree Modelmentioning

confidence: 99%

“…Balakrishnan and Chopra [22] adopted the method of gradient acceleration optimization to improve the execution efficiency of progressive EM HLTA (PEM-HLTA) for topic detection. References [12], [13], [21], and [22] only considered how to build LTM with a large amount of data. However, in current applications, new data frequently enter the system [23].…”

Section: Updating Latent Tree Modelmentioning

confidence: 99%

See 1 more Smart Citation

Modeling of Multilayer Multicontent Latent Tree and Its Applications

Lin

Chiu

Wang

et al. 2021

IEEE Trans. Comput. Soc. Syst.

View full text Add to dashboard Cite

Latent tree model (LTM) is a probabilistic tree-structured graphical model, which can reveal the hidden hierarchical causal relations among data contents and play a key role in explainable artificial intelligence. However, because current LTM modeling techniques are only suitable for singlecontent variable, the applications of LTMs are somewhat limited. Toward this end, a multilayer LTM (ML-LTM) is first presented to deal with the hierarchical clustering issues of multicontent variables. Second, we further develop an ML-LTMbased multicontent recommendation system. Our experiment results show that the proposed ML-LTM can achieve 90% recommendation accuracy, but the current LTM can only has 20%. Third, we propose an incremental update approach for ML-LTM that can save five-sixth updating time comparing with the whole-model retraining approach for achieving the same recommendation accuracy.

show abstract