More Like this

1. Top-k Community Similarity Search Over Large Road-Network Graphs

   https://doi.org/10.1109/ICDE51399.2021.00206  

   Rai, Niranjan; Lian, Xiang (April 2021, IEEE 37th International Conference on Data Engineering)

   null (Ed.)
2. On the K-theory of division algebras over local fields

   https://doi.org/10.1007/s00222-019-00909-x  

   Hesselholt, Lars; Larsen, Michael; Lindenstrauss, Ayelet (July 2019, Inventiones mathematicae)

   Let K be a complete discrete valuation field with finite residue field of characteristic p, and let D be a central division algebra over K of finite index d. Thirty years ago, Suslin and Yufryakov showed that for all prime numbers ℓ different from p and integers j≥1 , there exists a "reduced norm" isomorphism of ℓ-adic K-groups Nrd_{D/K}:K_j(D,Z_ℓ)→K_j(K,Z_ℓ) such that d⋅Nrd_{D/K} is equal to the norm homomorphism N_{D/K}. The purpose of this paper is to prove the analogous result for the p-adic K-groups. To do so, we employ the cyclotomic trace map to topological cyclic homology and show that there exists a "reduced trace" equivalence Trd_{A/S}:THH(A|D,Z_p)→THH(S|K,Z_p) between two p-complete cyclotomic spectra associated with D and K, respectively. Interestingly, we show that if p divides d, then it is not possible to choose said equivalence such that, as maps of cyclotomic spectra, d⋅Trd_{A/S} agrees with the trace Tr_{A/S}, although this is possible as maps of spectra with T-action 

   more » « less
3. Effective and efficient top-k query processing over incomplete data streams

   https://doi.org/10.1016/j.ins.2020.08.011  

   Ren, Weilong; Lian, Xiang; Ghazinour, Kambiz (January 2021, Information Sciences)

   null (Ed.)
4. K− over K+ multiplicity ratio for kaons produced in DIS with a large fraction of the virtual-photon energy

   https://doi.org/10.1016/j.physletb.2018.09.052  

   Akhunzyanov, R.; Alexeev, M.G.; Alexeev, G.D.; Amoroso, A.; Andrieux, V.; Anfimov, N.V.; Anosov, V.; Antoshkin, A.; Augsten, K.; Augustyniak, W.; et al (November 2018, Physics Letters B)
5. K-Paths: Reasoning over Graph Paths for Drug Repurposing and Drug Interaction Prediction

   https://doi.org/10.1145/3711896.3737011  

   Abdullahi, Tassallah; Gemou, Ioanna; Nayak, Nihal V; Murtaza, Ghulam; Bach, Stephen H; Eickhoff, Carsten; Singh, Ritambhara (August 2025, KDD '25: Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining)

   Biomedical knowledge graphs (KGs) encode rich, structured information critical for drug discovery tasks, but extracting meaningful insights from large-scale KGs remains challenging due to their complex structure. Existing biomedical subgraph retrieval methods are tailored for graph neural networks (GNNs), limiting compatibility with other paradigms, including large language models (LLMs). We introduce K-Paths, a model-agnostic retrieval framework that extracts structured, diverse, and biologically meaningful multi-hop paths from dense biomedical KGs. These paths enable prediction of unobserved drug-drug and drug-disease interactions, including those involving entities not seen during training, thus supporting inductive reasoning. K-Paths is training-free and employs a diversity-aware adaptation of Yen's algorithm to extract the K shortest loopless paths between entities in a query, prioritizing biologically relevant and relationally diverse connections. These paths serve as concise, interpretable reasoning chains that can be directly integrated with LLMs or GNNs to improve generalization, accuracy, and enable explainable inference. Experiments on benchmark datasets show that K-Paths improves zero-shot reasoning across state-of-the-art LLMs. For instance, Tx-Gemma 27B improves by 19.8 and 4.0 F1 points on interaction severity prediction and drug repurposing tasks, respectively. Llama 70B achieves gains of 8.5 and 6.2 points on the same tasks. K-Paths also boosts the training efficiency of EmerGNN, a state-of-the-art GNN, by reducing the KG size by 90% while maintaining predictive performance. Beyond efficiency, K-Paths bridges the gap between KGs and LLMs, enabling scalable and explainable LLM-augmented scientific discovery. We release our code and the retrieved paths as a benchmark for inductive reasoning. 

   more » « less