English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Phrase mining is a fundamental text mining task that aims to identify qualityphrases from context. Nevertheless, the scarcity of extensive gold labelsdatasets, demanding substantial annotation efforts from experts, renders thistask exceptionally challenging. Furthermore, the emerging, infrequent, anddomain-specific nature of quality phrases presents further challenges indealing with this task. In this paper, we propose LMPhrase, a novelunsupervised context-aware quality phrase mining framework built upon largepre-trained language models (LMs). Specifically, we first mine quality phrasesas silver labels by employing a parameter-free probing technique calledPerturbed Masking on the pre-trained language model BERT (coined as Annotator).In contrast to typical statistic-based or distantly-supervised methods, oursilver labels, derived from large pre-trained language models, take intoaccount rich contextual information contained in the LMs. As a result, theybring distinct advantages in preserving informativeness, concordance, andcompleteness of quality phrases. Secondly, training a discriminative spanprediction model heavily relies on massive annotated data and is likely to facethe risk of overfitting silver labels. Alternatively, we formalize phrasetagging task as the sequence generation problem by directly fine-tuning on theSequence-to-Sequence pre-trained language model BART with silver labels (coinedas Generator). Finally, we merge the quality phrases from both the Annotatorand Generator as the final predictions, considering their complementary natureand distinct characteristics. Extensive experiments show that our LMPhraseconsistently outperforms all the existing competitors across two differentgranularity phrase mining tasks, where each task is tested on two differentdomain datasets.

Language Model as an Annotator: Unsupervised Context-aware Quality Phrase Generation