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

In recent years, with large language models (LLMs) achieving state-of-the-artperformance in context understanding, increasing efforts have been dedicated todeveloping LLM-enhanced sequential recommendation (SR) methods. Consideringthat most existing LLMs are not specifically optimized for recommendationtasks, adapting them for SR becomes a critical step in LLM-enhanced SR methods.Though numerous adaptation methods have been developed, it still remains asignificant challenge to adapt LLMs for SR both efficiently and effectively. Toaddress this challenge, in this paper, we introduce a novel side sequentialnetwork adaptation method, denoted as SSNA, for LLM enhanced SR. SSNA featuresthree key designs to allow both efficient and effective LLM adaptation. First,SSNA learns adapters separate from LLMs, while fixing all the pre-trainedparameters within LLMs to allow efficient adaptation. In addition, SSNA adaptsthe top-a layers of LLMs jointly, and integrates adapters sequentially forenhanced effectiveness (i.e., recommendation performance). We compare SSNAagainst five state-of-the-art baseline methods on five benchmark datasets usingthree LLMs. The experimental results demonstrate that SSNA significantlyoutperforms all the baseline methods in terms of recommendation performance,and achieves substantial improvement over the best-performing baseline methodsat both run-time and memory efficiency during training. Our analysis shows theeffectiveness of integrating adapters in a sequential manner. Our parameterstudy demonstrates the effectiveness of jointly adapting the top-a layers ofLLMs.

Towards Efficient and Effective Adaptation of Large Language Models for Sequential Recommendation