We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Computer Science > Computation and Language

Title: TransCouplet:Transformer based Chinese Couplet Generation

Abstract: Chinese couplet is a special form of poetry composed of complex syntax with ancient Chinese language. Due to the complexity of semantic and grammatical rules, creation of a suitable couplet is a formidable challenge. This paper presents a transformer-based sequence-to-sequence couplet generation model. With the utilization of AnchiBERT, the model is able to capture ancient Chinese language understanding. Moreover, we evaluate the Glyph, PinYin and Part-of-Speech tagging on the couplet grammatical rules to further improve the model.
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as: arXiv:2112.01707 [cs.CL]
  (or arXiv:2112.01707v1 [cs.CL] for this version)

Submission history

From: Kuan-Yu Chiang [view email]
[v1] Fri, 3 Dec 2021 04:34:48 GMT (359kb,D)

Link back to: arXiv, form interface, contact.