Ruwan Wickramarachchi

Making analogies is fundamental to cognition. Proportional analogies, which consist of four terms, are often used to assess linguistic and cognitive abilities. For instance, completing analogies like “Oxygen is to Gas as \is to \” requires identifying the semantic relationship (e.g., “type of”) between the first pair of terms (“Oxygen” and “Gas”) and finding a second pair that shares the same relationship (e.g., “Aluminum” and “Metal”). In this work, we introduce a 15K Multiple-Choice Question Answering (MCQA) dataset for proportional analogy completion and evaluate the performance of contemporary Large Language Models (LLMs) in various knowledge-enhanced prompt settings. Specifically, we augment prompts with three types of knowledge: exemplar, structured, and targeted. Our results show that despite extensive training data, solving proportional analogies remains challenging for current LLMs, with the best model achieving an accuracy of 55%. Notably, we find that providing targeted knowledge can better assist models in completing proportional analogies compared to providing exemplars or collections of structured knowledge.

Knowledge graph completion (KGC) is a problem of significant importance due to the inherent incompleteness in knowledge graphs (KGs). The current approaches for KGC using link prediction (LP) mostly rely on a common set of benchmark datasets that are quite different from real-world industrial KGs. Therefore, the adaptability of current LP methods for real-world KGs and domain-specific applications is questionable. To support the evaluation of current and future LP and KGC methods for industrial KGs, we introduce DSceneKG, a suite of real-world driving scene knowledge graphs that are currently being used across various industrial applications. The DSceneKG is publicly available at: https://github.com/ruwantw/DSceneKG.

The ability of Large Language Models (LLMs) to encode syntactic and semantic structures of language is well examined in NLP. Additionally, analogy identification, in the form of word analogies are extensively studied in the last decade of language modeling literature. In this work we specifically look at how LLMs’ abilities to capture sentence analogies (sentences that convey analogous meaning to each other) vary with LLMs’ abilities to encode syntactic and semantic structures of sentences. Through our analysis, we find that LLMs’ ability to identify sentence analogies is positively correlated with their ability to encode syntactic and semantic structures of sentences. Specifically, we find that the LLMs which capture syntactic structures better, also have higher abilities in identifying sentence analogies.

Over the past decade, analogies, in the form of word-level analogies, have played a significant role as an intrinsic measure of evaluating the quality of word embedding methods such as word2vec. Modern large language models (LLMs), however, are primarily evaluated on extrinsic measures based on benchmarks such as GLUE and SuperGLUE, and there are only a few investigations on whether LLMs can draw analogies between long texts. In this paper, we present ANALOGICAL, a new benchmark to intrinsically evaluate LLMs across a taxonomy of analogies of long text with six levels of complexity – (i) word, (ii) word vs. sentence, (iii) syntactic, (iv) negation, (v) entailment, and (vi) metaphor. Using thirteen datasets and three different distance measures, we evaluate the abilities of eight LLMs in identifying analogical pairs in the semantic vector space. Our evaluation finds that it is increasingly challenging for LLMs to identify analogies when going up the analogy taxonomy.

Generating high-quality annotations for object detection and recognition is a challenging and important task, especially in relation to safety-critical applications such as autonomous driving (AD). Due to the difficulty of perception in challenging situations such as occlusion, degraded weather, and sensor failure, objects can go unobserved and unlabeled. In this paper, we present CLUE-AD, a general-purpose method for detecting and labeling unobserved entities by leveraging the object continuity assumption within the context of a scene. This method is dataset-agnostic, supporting any existing and future AD datasets. Using a real-world dataset representing complex urban driving scenes, we demonstrate the applicability of CLUE-AD for detecting unobserved entities and augmenting the scene data with new labels.

Scene understanding is a key technical challenge within the autonomous driving domain. It requires a deep semantic understanding of the entities and relations found within complex physical and social environments that is both accurate and complete. In practice, this can be accomplished by representing entities in a scene and their relations as a knowledge graph (KG). This scene knowledge graph may then be utilized for the task of entity prediction, leading to improved scene understanding. In this paper, we will define and formalize this problem as Knowledge-based Entity Prediction (KEP). KEP aims to improve scene understanding by predicting potentially unrecognized entities by leveraging heterogeneous, high-level semantic knowledge of driving scenes. An innovative neuro-symbolic solution for KEP is presented, based on knowledge-infused learning, which 1) introduces a dataset agnostic ontology to describe driving scenes, 2) uses an expressive, holistic representation of scenes with knowledge graphs, and 3) proposes an effective, non-standard mapping of the KEP problem to the problem of link prediction (LP) using knowledge-graph embeddings (KGE). Using real, complex and high-quality data from urban driving scenes, we demonstrate its effectiveness by showing that the missing entities may be predicted with high precision (0.87 Hits@1) while significantly outperforming the non-semantic/rule-based baselines.