Pointing out the Shortcomings of Relation Extraction Models with Semantically Motivated Adversarials

In recent years, large language models have achieved state-of-the-art performance across various NLP tasks. However, investigations have shown that these models tend to rely on shortcut features, leading to inaccurate predictions and causing the models to be unreliable at generalization to out-of-distribution (OOD) samples. For instance, in the context of relation extraction (RE), we would expect a model to identify the same relation independently of the entities involved in it. For example, consider the sentence "Leonardo da Vinci painted the Mona Lisa" expressing the created(Leonardo_da_Vinci, Mona_Lisa) relation. If we substiute "Leonardo da Vinci" with "Barack Obama", then the sentence still expresses the created relation. A robust model is supposed to detect the same relation in both cases. In this work, we describe several semantically-motivated strategies to generate adversarial examples by replacing entity mentions and investigate how state-of-the-art RE models perform under pressure. Our analyses show that the performance of these models significantly deteriorates on the modified datasets (avg. of -48.5% in F1), which indicates that these models rely to a great extent on shortcuts, such as surface forms (or patterns therein) of entities, without making full use of the information present in the sentences.

Insights from an OTTR-centric Ontology Engineering Methodology

OTTR is a language for representing ontology modeling patterns, which enables to build ontologies or knowledge bases by instantiating templates. Thereby, particularities of the ontological representation language are hidden from the domain experts, and it enables ontology engineers to, to some extent, separate the processes of deciding about what information to model from deciding about how to model the information, e.g., which design patterns to use. Certain decisions can thus be postponed for the benefit of focusing on one of these processes. To date, only few works on ontology engineering where ontology templates are applied are described in the literature. In this paper, we outline our methodology and report findings from our ontology engineering activities in the domain of Material Science. In these activities, OTTR templates play a key role. Our ontology engineering process is bottom-up, as we begin modeling activities from existing data that is then, via templates, fed into a knowledge graph, and it is top-down, as we first focus on which data to model and postpone the decision of how to model the data. We find, among other things, that OTTR templates are especially useful as a means of communication with domain experts. Furthermore, we find that because OTTR templates encapsulate modeling decisions, the engineering process becomes flexible, meaning that design decisions can be changed at little cost.

Large Language Models on Wikipedia-Style Survey Generation: an Evaluation in NLP Concepts

Large Language Models (LLMs) have achieved significant success across various natural language processing (NLP) tasks, encompassing question-answering, summarization, and machine translation, among others. While LLMs excel in general tasks, their efficacy in domain-specific applications remains under exploration. Additionally, LLM-generated text sometimes exhibits issues like hallucination and disinformation. In this study, we assess LLMs' capability of producing concise survey articles within the computer science-NLP domain, focusing on 20 chosen topics. Automated evaluations indicate that GPT-4 outperforms GPT-3.5 when benchmarked against the ground truth. Furthermore, four human evaluators provide insights from six perspectives across four model configurations. Through case studies, we demonstrate that while GPT often yields commendable results, there are instances of shortcomings, such as incomplete information and the exhibition of lapses in factual accuracy.