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

Download:

Current browse context:

cs

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 ScienceWISE logo

Computer Science > Computation and Language

Title: Applying a Generic Sequence-to-Sequence Model for Simple and Effective Keyphrase Generation

Abstract: In recent years, a number of keyphrase generation (KPG) approaches were proposed consisting of complex model architectures, dedicated training paradigms and decoding strategies. In this work, we opt for simplicity and show how a commonly used seq2seq language model, BART, can be easily adapted to generate keyphrases from the text in a single batch computation using a simple training procedure. Empirical results on five benchmarks show that our approach is as good as the existing state-of-the-art KPG systems, but using a much simpler and easy to deploy framework.
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as: arXiv:2201.05302 [cs.CL]
  (or arXiv:2201.05302v1 [cs.CL] for this version)

Submission history

From: Md Faisal Mahbub Chowdhury [view email]
[v1] Fri, 14 Jan 2022 04:50:28 GMT (6149kb,D)

Link back to: arXiv, form interface, contact.