Knowledge graph completion (KGC) is a widely used method to tackle incompleteness in knowledge graphs (KGs) by making predictions for missing links. Description-based KGC leverages pre-trained language models to learn entity and relation representations with their names or descriptions, which shows promising results. However, the performance of description-based KGC is still limited by the quality of text and the incomplete structure, as it lacks sufficient entity descriptions and relies solely on relation names, leading to sub-optimal results. To address this issue, we propose MPIKGC, a general framework to compensate for the deficiency of contextualized knowledge and improve KGC by querying large language models (LLMs) from various perspectives, which involves leveraging the reasoning, explanation, and summarization capabilities of LLMs to expand entity descriptions, understand relations, and extract structures, respectively. We conducted extensive evaluation of the effectiveness and improvement of our framework based on four description-based KGC models and four datasets, for both link prediction and triplet classification tasks.
Model editing aims to precisely modify the behaviours of large language models (LLMs) on specific knowledge while keeping irrelevant knowledge unchanged. It has been proven effective in resolving hallucination and out-of-date issues in LLMs. As a result, it can boost the application of LLMs in many critical domains (e.g., medical domain), where the hallucination is not tolerable. In this paper, we propose two model editing studies and validate them in the medical domain: (1) directly editing the factual medical knowledge and (2) editing the explanations to facts. Meanwhile, we observed that current model editing methods struggle with the specialization and complexity of medical knowledge. Therefore, we propose MedLaSA, a novel Layer-wise Scalable Adapter strategy for medical model editing. It employs causal tracing to identify the precise location of knowledge in neurons and then introduces scalable adapters into the dense layers of LLMs. These adapters are assigned scaling values based on the corresponding specific knowledge. To evaluate the editing impact, we build two benchmark datasets and introduce a series of challenging and comprehensive metrics. Extensive experiments on medical LLMs demonstrate the editing efficiency of MedLaSA, without affecting irrelevant knowledge that is not edited.
Although biomedical entity linking (BioEL) has made significant progress with pre-trained language models, challenges still exist for fine-grained and long-tailed entities. To address these challenges, we present BioELQA, a novel model that treats Biomedical Entity Linking as Multiple Choice Question Answering. BioELQA first obtains candidate entities with a fast retriever, jointly presents the mention and candidate entities to a generator, and then outputs the predicted symbol associated with its chosen entity. This formulation enables explicit comparison of different candidate entities, thus capturing fine-grained interactions between mentions and entities, as well as among entities themselves. To improve generalization for long-tailed entities, we retrieve similar labeled training instances as clues and concatenate the input with retrieved instances for the generator. Extensive experimental results show that BioELQA outperforms state-of-the-art baselines on several datasets.
Emotion-Cause Pair Extraction (ECPE) involves extracting clause pairs representing emotions and their causes in a document. Existing methods tend to overfit spurious correlations, such as positional bias in existing benchmark datasets, rather than capturing semantic features. Inspired by recent work, we explore leveraging large language model (LLM) to address ECPE task without additional training. Despite strong capabilities, LLMs suffer from uncontrollable outputs, resulting in mediocre performance. To address this, we introduce chain-of-thought to mimic human cognitive process and propose the Decomposed Emotion-Cause Chain (DECC) framework. Combining inducing inference and logical pruning, DECC guides LLMs to tackle ECPE task. We further enhance the framework by incorporating in-context learning. Experiment results demonstrate the strength of DECC compared to state-of-the-art supervised fine-tuning methods. Finally, we analyze the effectiveness of each component and the robustness of the method in various scenarios, including different LLM bases, rebalanced datasets, and multi-pair extraction.
Biomedical entity linking (BioEL) has achieved remarkable progress with the help of pre-trained language models. However, existing BioEL methods usually struggle to handle rare and difficult entities due to long-tailed distribution. To address this limitation, we introduce a new scheme $k$NN-BioEL, which provides a BioEL model with the ability to reference similar instances from the entire training corpus as clues for prediction, thus improving the generalization capabilities. Moreover, we design a contrastive learning objective with dynamic hard negative sampling (DHNS) that improves the quality of the retrieved neighbors during inference. Extensive experimental results show that $k$NN-BioEL outperforms state-of-the-art baselines on several datasets.
Accurately predicting drug-drug interactions (DDI) for emerging drugs, which offer possibilities for treating and alleviating diseases, with computational methods can improve patient care and contribute to efficient drug development. However, many existing computational methods require large amounts of known DDI information, which is scarce for emerging drugs. In this paper, we propose EmerGNN, a graph neural network (GNN) that can effectively predict interactions for emerging drugs by leveraging the rich information in biomedical networks. EmerGNN learns pairwise representations of drugs by extracting the paths between drug pairs, propagating information from one drug to the other, and incorporating the relevant biomedical concepts on the paths. The different edges on the biomedical network are weighted to indicate the relevance for the target DDI prediction. Overall, EmerGNN has higher accuracy than existing approaches in predicting interactions for emerging drugs and can identify the most relevant information on the biomedical network.
Knowledge graph (KG) embedding is a fundamental task in natural language processing, and various methods have been proposed to explore semantic patterns in distinctive ways. In this paper, we propose to learn an ensemble by leveraging existing methods in a relation-aware manner. However, exploring these semantics using relation-aware ensemble leads to a much larger search space than general ensemble methods. To address this issue, we propose a divide-search-combine algorithm RelEns-DSC that searches the relation-wise ensemble weights independently. This algorithm has the same computation cost as general ensemble methods but with much better performance. Experimental results on benchmark datasets demonstrate the effectiveness of the proposed method in efficiently searching relation-aware ensemble weights and achieving state-of-the-art embedding performance. The code is public at https://github.com/LARS-research/RelEns.
In this paper, we investigate a state-of-the-art reconfigurable intelligent surface (RIS)-assisted spatial scattering modulation (SSM) scheme for millimeter-wave (mmWave) systems, where a more practical scenario that the RIS is near the transmitter while the receiver is far from RIS is considered. To this end, the line-of-sight (LoS) and non-LoS links are utilized in the transmitter-RIS and RIS-receiver channels, respectively. By employing the maximum likelihood detector at the receiver, the conditional pairwise error probability (CPEP) expression for the RIS-SSM scheme is derived under the two scenarios that the received beam demodulation is correct or not. Furthermore, the union upper bound of average bit error probability (ABEP) is obtained based on the CPEP expression. Finally, the derivation results are exhaustively validated by the Monte Carlo simulations.
This paper investigates the reconfigurable intelligent surface (RIS) assisted spatial scattering modulation (SSM) scheme for millimeter-wave (mmWave) multiple-input multiple-output (MIMO) systems, in which line-of-sight (LoS) and non-line-of-sight (NLoS) paths are respectively considered in the transmitter-RIS and RIS-receiver channels. Based on the maximum likelihood detector, the conditional pairwise error probability (CPEP) expression for the RIS-SSM scheme is derived under the two cases of received beam correct and demodulation error. Furthermore, we derive the closed-form expressions of the unconditional pairwise error probability (UPEP) by employing two different methods: the probability density function and the moment-generating function expressions with a descending order of scatterer gains. To provide more useful insights, we derive the asymptotic UPEP and the diversity gain of the RIS-SSM scheme in the high SNR region. Depending on UPEP and the corresponding Euclidean distance, we get the union upper bound of the average bit error probability (ABEP). A new framework for ergodic capacity analysis is also provided to acquire the proposed system's effective capacity. Finally, all derivation results are validated via extensive Monte Carlo simulations, revealing that the proposed RIS-SSM scheme outperforms the benchmarks in terms of reliability.
Drawing inspiration from the advantages of intelligent reflecting surfaces (IRS) in wireless networks,this paper presents a novel design for intelligent omni surface (IOS) enabled integrated sensing and communications (ISAC). By harnessing the power of multi antennas and a multitude of elements, the dual-function base station (BS) and IOS collaborate to realize joint active and passive beamforming, enabling seamless 360-degree ISAC coverage. The objective is to maximize the minimum signal-tointerference-plus-noise ratio (SINR) of multi-target sensing, while ensuring the multi-user multi-stream communications. To achieve this, a comprehensive optimization approach is employed, encompassing the design of radar receive vector, transmit beamforming matrix, and IOS transmissive and reflective coefficients. Due to the non-convex nature of the formulated problem, an auxiliary variable is introduced to transform it into a more tractable form. Consequently, the problem is decomposed into three subproblems based on the block coordinate descent algorithm. Semidefinite relaxation and successive convex approximation methods are leveraged to convert the sub-problem into a convex problem, while the iterative rank minimization algorithm and penalty function method ensure the equivalence. Furthermore,the scenario is extended to mode switching and time switching protocols. Simulation results validate the convergence and superior performance of the proposed algorithm compared to other benchmark algorithms.