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

Incremental Named Entity Recognition (INER) involves the sequential learningof new entity types without accessing the training data of previously learnedtypes. However, INER faces the challenge of catastrophic forgetting specificfor incremental learning, further aggravated by background shift (i.e., old andfuture entity types are labeled as the non-entity type in the current task). Toaddress these challenges, we propose a method called task Relation Distillationand Prototypical pseudo label (RDP) for INER. Specifically, to tacklecatastrophic forgetting, we introduce a task relation distillation scheme thatserves two purposes: 1) ensuring inter-task semantic consistency acrossdifferent incremental learning tasks by minimizing inter-task relationdistillation loss, and 2) enhancing the model's prediction confidence byminimizing intra-task self-entropy loss. Simultaneously, to mitigate backgroundshift, we develop a prototypical pseudo label strategy that distinguishes oldentity types from the current non-entity type using the old model. Thisstrategy generates high-quality pseudo labels by measuring the distancesbetween token embeddings and type-wise prototypes. We conducted extensiveexperiments on ten INER settings of three benchmark datasets (i.e., CoNLL2003,I2B2, and OntoNotes5). The results demonstrate that our method achievessignificant improvements over the previous state-of-the-art methods, with anaverage increase of 6.08% in Micro F1 score and 7.71% in Macro F1 score.

Task Relation Distillation and Prototypical Pseudo Label for Incremental Named Entity Recognition