Large Language Models (LLMs) have been increasingly studied as neural knowledge bases for supporting knowledge-intensive applications such as question answering and fact checking. However, the structural organization of their knowledge remains unexplored. Inspired by cognitive neuroscience findings, such as semantic clustering and priming, where knowing one fact increases the likelihood of recalling related facts, we investigate an analogous knowledge homophily pattern in LLMs. To this end, we map LLM knowledge into a graph representation through knowledge checking at both the triplet and entity levels. After that, we analyze the knowledgeability relationship between an entity and its neighbors, discovering that LLMs tend to possess a similar level of knowledge about entities positioned closer in the graph. Motivated by this homophily principle, we propose a Graph Neural Network (GNN) regression model to estimate entity-level knowledgeability scores for triplets by leveraging their neighborhood scores. The predicted knowledgeability enables us to prioritize checking less well-known triplets, thereby maximizing knowledge coverage under the same labeling budget. This not only improves the efficiency of active labeling for fine-tuning to inject knowledge into LLMs but also enhances multi-hop path retrieval in reasoning-intensive question answering.

Recent advancements in large language models (LLMs) based embedding models have established new state-of-the-art benchmarks for text embedding tasks, particularly in dense vector-based retrieval. However, these models predominantly focus on English, leaving multilingual embedding capabilities largely unexplored. To address this limitation, we present LUSIFER, a novel zero-shot approach that adapts LLM-based embedding models for multilingual tasks without requiring multilingual supervision. LUSIFER's architecture combines a multilingual encoder, serving as a language-universal learner, with an LLM-based embedding model optimized for embedding-specific tasks. These components are seamlessly integrated through a minimal set of trainable parameters that act as a connector, effectively transferring the multilingual encoder's language understanding capabilities to the specialized embedding model. Additionally, to comprehensively evaluate multilingual embedding performance, we introduce a new benchmark encompassing 5 primary embedding tasks, 123 diverse datasets, and coverage across 14 languages. Extensive experimental results demonstrate that LUSIFER significantly enhances the multilingual performance across various embedding tasks, particularly for medium and low-resource languages, without requiring explicit multilingual training data. The code and dataset for training are available at: https://github.com/hieum98/lusifer

Multilingual semantic retrieval involves retrieving semantically relevant content to a query irrespective of the language. Compared to monolingual and bilingual semantic retrieval, multilingual semantic retrieval requires a stronger alignment approach to pull the contents to be retrieved close to the representation of their corresponding queries, no matter their language combinations. Traditionally, this is achieved through more supervision in the form of multilingual parallel resources, which are expensive to obtain, especially for low-resource languages. In this work, on top of an optimization-based Model-Agnostic Meta-Learner (MAML), we propose a data-efficient meta-distillation approach: MAML-Align,1 specifically for low-resource multilingual semantic retrieval. Our approach simulates a gradual feedback loop from monolingual to bilingual and from bilingual to multilingual semantic retrieval. We systematically compare multilingual meta-distillation learning to different baselines and conduct ablation studies on the role of different sampling approaches in the meta-task construction. We show that MAML-Align's gradual feedback loop boosts the generalization to different languages, including zero-shot ones, better than naive fine-tuning and vanilla MAML.

Graph Neural Networks (GNNs) have become increasingly important due to their representational power and state-of-the-art predictive performance on many fundamental learning tasks. Despite this success, GNNs suffer from fairness issues that arise as a result of the underlying graph data and the fundamental aggregation mechanism that lies at the heart of the large class of GNN models. In this article, we examine and categorize fairness techniques for improving the fairness of GNNs. We categorize these techniques by whether they focus on improving fairness in the pre-processing, in-processing (during training), or post-processing phases. We discuss how such techniques can be used together whenever appropriate and highlight the advantages and intuition as well. We also introduce an intuitive taxonomy for fairness evaluation metrics, including graph-level fairness, neighborhood-level fairness, embedding-level fairness, and prediction-level fairness metrics. In addition, graph datasets that are useful for benchmarking the fairness of GNN models are summarized succinctly. Finally, we highlight key open problems and challenges that remain to be addressed.

Aspect-based meeting transcript summarization aims to produce multiple summaries, each focusing on one aspect of content in a meeting transcript. It is challenging as sentences related to different aspects can mingle together, and those relevant to a specific aspect can be scattered throughout the long transcript of a meeting. The traditional summarization methods produce one summary mixing information of all aspects, which cannot deal with the above challenges of aspect-based meeting transcript summarization. In this paper, we propose a two-stage method for aspect-based meeting transcript summarization. To select the input content related to specific aspects, we train a sentence classifier on a dataset constructed from the AMI corpus with pseudo-labeling. Then we merge the sentences selected for a specific aspect as the input for the summarizer to produce the aspect-based summary. Experimental results on the AMI corpus outperform many strong baselines, which verifies the effectiveness of our proposed method.

In this paper, we introduce a novel concept of user-entity differential privacy (UeDP) to provide formal privacy protection simultaneously to both sensitive entities in textual data and data owners in learning natural language models (NLMs). To preserve UeDP, we developed a novel algorithm, called UeDP-Alg, optimizing the trade-off between privacy loss and model utility with a tight sensitivity bound derived from seamlessly combining user and sensitive entity sampling processes. An extensive theoretical analysis and evaluation show that our UeDP-Alg outperforms baseline approaches in model utility under the same privacy budget consumption on several NLM tasks, using benchmark datasets.

Search engines for domain-specific media collections often rely on rich metadata being available for the content items. The annotations may not be complete or rich enough to support an adequate retrieval effectiveness. As a result, some search queries receive only a small result set (low recall) and others might suffer from reduced relevance (low precision). To alleviate this, we present a framework that exploits external knowledge to provide entity-oriented reformulation suggestions for queries that contain entities. We propose that queries be added as surrogate nodes to an external Knowledge Graph (KG) via the use of state-of-the-art entity linking algorithms. Embedding methods are invoked on the augmented graph, which contains additional edges between surrogate nodes and KG entities. We introduce a new evaluation setting to evaluate the quality of these embeddings. Experimental results on seven datasets confirm the effectiveness of the approach.

We address the poor generalization of few-shot learning models for event detection (ED) using transfer learning and representation regularization. In particular, we propose to transfer knowledge from open-domain word sense disambiguation into few-shot learning models for ED to improve their generalization to new event types. We also propose a novel training signal derived from dependency graphs to regularize the representation learning for ED. Moreover, we evaluate few-shot learning models for ED with a large-scale human-annotated ED dataset to obtain more reliable insights for this problem. Our comprehensive experiments demonstrate that the proposed model outperforms state-of-the-art baseline models in the few-shot learning and supervised learning settings for ED. Code and data splits are available at https://github.com/laiviet/ed-fsl.

In this paper, we propose a novel method for a sentence-level answer-selection task that is a fundamental problem in natural language processing. First, we explore the effect of additional information by adopting a pretrained language model to compute the vector representation of the input text and by applying transfer learning from a large-scale corpus. Second, we enhance the compare-aggregate model by proposing a novel latent clustering method to compute additional information within the target corpus and by changing the objective function from listwise to pointwise. To evaluate the performance of the proposed approaches, experiments are performed with the WikiQA and TREC-QA datasets. The empirical results demonstrate the superiority of our proposed approach, which achieve state-of-the-art performance for both datasets.