In real-world scenarios, users tend to engage with a small set of popular items, while a large number of long-tail items receive little to no interaction. This long-tail phenomenon substantially impairs recommendation quality. Although prior approaches have attempted to address this issue, the absence of sufficient collaborative signals remains a major obstacle. With the advent of Large Language Models (LLMs), recent studies have explored leveraging LLM-derived semantics to enrich recommendation models. These approaches aim to incorporate textual or contextual knowledge to compensate for limited user-item interactions. A key challenge, however, lies in effectively integrating semantic signals with collaborative representations, which originate from different modalities and learning dynamics. To tackle this, We propose a novel framework, called FCRLLM (the Flipped Classroom with LLM), for long-tail sequential recommendation that aligns collaborative and LLM-based semantic representations. The flipped classroom mechanism dynamically updates the teacher representation to align with the student's attention, enabling more effective integration of semantic and collaborative information. This alignment is implemented via an energy-based formulation inspired by Hopfield networks. To validate its effectiveness, we conduct extensive experiments on three real-world datasets and demonstrate that FCRLLM consistently improves recommendation performance regardless of item popularity or user activity.

We propose Spectral Edge Encoding (SEE), a parameter-free framework that quantifies each edge's contribution to the global structure by measuring spectral shifts in the Laplacian eigenvalues. SEE captures the low-frequency sensitivity of edges and integrates these scores into graph Transformer attention logits as a structure-aware bias. When applied to the Moiré Graph Transformer (MoiréGT) and evaluated on seven MoleculeNet classification benchmarks, SEE consistently improves ROC-AUC performance. In particular, MoiréGT+SEE achieves an average ROC-AUC of 85.3%, approximately 7.1 percentage points higher than the previous state-of-the-art model UniCorn (78.2%). Moreover, SEE preserves molecular topology and enables edge-level interpretability, offering a practical alternative to sequence-based chemical language models. These results demonstrate that spectrum-informed attention can simultaneously enhance performance and transparency in graph-based molecular modeling.

Large Language Models (LLMs) excel in various NLP tasks but remain underexplored in recommendation systems. This study proposes the Sequential Emotion-Aware LLM-Based Personalized Recommendation System (SEALR ) to leverage sentiment analysis in user-generated reviews, tracking emotional changes and extracting sentiment labels. It integrates candidate items produced by sequential models with user behavior data into an LLM, enhancing personalization. Experiments on Amazon and Yelp datasets explore the effect of varied candidate pool sizes and instruction-based fine-tuning ratios, demonstrating significant performance gains. The combination of sentiment insights and user behavior data effectively accommodates diverse user preferences and contexts.

Deep learning models exhibit a dependency on peripheral attributes of input data, such as shapes and colors, leading the models to become biased towards these certain attributes that result in subsequent degradation of performance. In this paper, we alleviate this problem by presenting~\sysname, a novel framework that tackles dataset bias by leveraging generative models to amplify bias and facilitate the learning of debiased representations of the classifier. Our method involves three major steps. We initially train a biased classifier, denoted as f_b, on a biased dataset and extract the top-K biased-conflict samples. Next, we train a generator solely on a bias-conflict dataset comprised of these top-K samples, aiming to learn the distribution of bias-conflict samples. Finally, we re-train the classifier on the newly constructed debiased dataset, which combines the original and amplified data. This allows the biased classifier to competently learn debiased representation. Extensive experiments validate that our proposed method effectively debiases the biased classifier.

In real-world recommender model deployments, the models are typically retrained and deployed repeatedly. It is the rule-of-thumb to periodically retrain recommender models to capture up-to-date user behavior and item trends. However, the harm caused by delayed model updates has not been investigated extensively yet. in this perspective paper, we formulate the delayed model update problem and quantitatively demonstrate the delayed model update actually harms the model performance by increasing the number of cold users and cold items increase and decreasing overall model performances. These effects vary across different domains having different characteristics. Upon these findings, we further argue that although the delayed model update has negative effects on online recommender model deployment, yet it has not gathered enough attention from research communities. We argue our verification of the relationship between the model update cycle and model performance calls for further research such as faster model training, and more efficient data pipelines to keep the model more up-to-date with the latest user behaviors and item trends.