Qin Luo

Mr. Qin Luo

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 January 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
Edmund Venant NdimboEdmund Venant Ndimbo
Edmund Venant Ndimbo
QLQin Luo
Qin Luo
Gimo Carlos FernandoGimo Carlos Fernando
Gimo Carlos Fernando
XYXu Yang
Xu Yang
Bang WangBang Wang
Bang Wang
A conversational system is an artificial intelligence application designed to interact with users in natural language, providing accurate and contextually relevant responses. Building such systems for low-resource languages like Swahili presents significant challenges due to the limited availability of large-scale training datasets. This paper proposes a Retrieval-Augmented Generation-based system to address these challenges and improve the quality of Swahili conversational AI. The system leverages fine-tuning, where models are trained on available Swahili data, combined with external knowledge retrieval to enhance response accuracy and fluency. Four models—mT5, GPT-2, mBART, and GPT-Neo—were evaluated using metrics such as BLEU, METEOR, Query Performance, and inference time. Results show that Retrieval-Augmented Generation consistently outperforms fine-tuning alone, particularly in generating detailed and contextually appropriate responses. Among the tested models, mT5 with Retrieval-Augmented Generation demonstrated the best performance, achieving a BLEU score of 56.88%, a METEOR score of 72.72%, and a Query Performance score of 84.34%, while maintaining relevance and fluency. Although Retrieval-Augmented Generation introduces slightly longer response times, its ability to significantly improve response quality makes it an effective approach for Swahili conversational systems. This study highlights the potential of Retrieval-Augmented Generation to advance conversational AI for Swahili and other low-resource languages, with future work focusing on optimizing efficiency and exploring multilingual applications.

14 October 2024 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
BLBin Liu
Bin Liu
QLQin Luo
Qin Luo
Bang WangBang Wang
Bang Wang
Learning representations from pairwise comparisons has achieved significant success in various fields, including computer vision and information retrieval. In recommendation systems, collaborative filtering algorithms based on pairwise learning are also rooted in this approach. However, a major challenge in collaborative filtering is the lack of labels for negative instances in implicit feedback data, leading to the inclusion of false negatives among randomly selected instances. This issue causes biased optimization objectives and results in biased parameter estimation. In this paper, we propose a novel method to address learning biases arising from implicit feedback data and introduce a modified loss function for pairwise learning, called debiased pairwise loss (DPL). The core idea of DPL is to correct the biased probability estimates caused by false negatives, thereby adjusting the gradients to more closely approximate those of fully supervised data. Implementing DPL requires only a small modification to the existing codebase. Experimental studies on public datasets demonstrate the effectiveness of the proposed method.

10 March 2026 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
BLBin Liu
Bin Liu
XLXiaohong Liu
Xiaohong Liu
QLQin Luo
Qin Luo
ZSZiqiao Shang
Ziqiao Shang
JCJielei Chu
Jielei Chu
TLTianrui Li
Tianrui Li
Pairwise learning underpins implicit collaborative filtering, yet its effectiveness is often hindered by sparse supervision, noisy interactions, and popularity-driven exposure bias. In this paper, we propose Variational Bayesian Personalized Ranking (VarBPR), a tractable variational framework for implicit-feedback pairwise learning that offers principled exposure controllability and theoretical interpretability. VarBPR reformulates pairwise learning as variational inference over discrete latent indexing variables, explicitly modeling noise and indexing uncertainty, and divides training into two stages: variational inference, which solve variational posteriors, and variational learning, which updates model parameters based on these posteriors. In the variational inference stage, we develop a variational formulation that integrates preference alignment, denoising, and popularity debiasing under a unified ELBO/regularization objective, deriving closed-form posteriors with clear control semantics: the prior encodes a target exposure pattern, while temperature/regularization strength controls posterior-prior adherence. As a result, exposure controllability becomes an endogenous and interpretable outcome of variational inference. In the variational learning stage, we propose a posterior-compression objective that reduces the ideal ELBO's computational complexity from polynomial to linear, with the approximation justified by an explicit Jensen-gap upper bound. Theoretically, we provide interpretable generalization guarantees by identifying a structural error component and revealing the opportunity cost of prioritizing certain exposure patterns (e.g., long-tail), offering a concrete analytical lens for designing controllable recommender systems. Empirically, We validate VarBPR across popular backbones; it demonstrates consistent gains in ranking accuracy, enables controlled long-tail exposure, and preserves the linear-time complexity of BPR.


29 July 2023 in arXiv
Verified
Preprint
Open access
PDF file
BLBin Liu
Bin Liu
QLQin Luo
Qin Luo
Bang WangBang Wang
Bang Wang
Learning contrastive representations from pairwise comparisons has achieved remarkable success in various fields, such as natural language processing, computer vision, and information retrieval. Collaborative filtering algorithms based on pairwise learning also rooted in this paradigm. A significant concern is the absence of labels for negative instances in implicit feedback data, which often results in the random selected negative instances contains false negatives and inevitably, biased embeddings. To address this issue, we introduce a novel correction method for sampling bias that yields a modified loss for pairwise learning called debiased pairwise loss (DPL). The key idea underlying DPL is to correct the biased probability estimates that result from false negatives, thereby correcting the gradients to approximate those of fully supervised data. The implementation of DPL only requires a small modification of the codes. Experimental studies on five public datasets validate the effectiveness of proposed learning method.