With the rapid development of LLMs, it is natural to ask how to harness their capabilities efficiently. In this paper, we explore whether it is feasible to direct each input query to a single most suitable LLM. To this end, we propose LLM routing for challenging reasoning tasks. Our extensive experiments suggest that such routing shows promise but is not feasible in all scenarios, so more robust approaches should be investigated to fill this gap.

This paper investigates the question of what makes math word problems (MWPs) in English challenging for large language models (LLMs). We conduct an in-depth analysis of the key linguistic and mathematical characteristics of MWPs. In addition, we train feature-based classifiers to better understand the impact of each feature on the overall difficulty of MWPs for prominent LLMs and investigate whether this helps predict how well LLMs fare against specific categories of MWPs.

Text simplification lacks a universal standard of quality, and annotated reference simplifications are scarce and costly. We propose to alleviate such limitations by introducing REFeREE, a reference-free model-based metric with a 3-stage curriculum. REFeREE leverages an arbitrarily scalable pretraining stage and can be applied to any quality standard as long as a small number of human annotations are available. Our experiments show that our metric outperforms existing reference-based metrics in predicting overall ratings and reaches competitive and consistent performance in predicting specific ratings while requiring no reference simplifications at inference time.

In this paper, we present our submission to the SemEval-2023 Task 3 “The Competition of Multimodal Emotion Cause Analysis in Conversations”, focusing on extracting emotion-cause pairs from dialogs. Specifically, our approach relies on combining fine-tuned GPT-3.5 for emotion classification and using a BiLSTM-based neural network to detect causes. We score 2nd in the ranking for Subtask 1, demonstrating the effectiveness of our approach through one of the highest weighted-average proportional F1 scores recorded at 0.264.

In this paper, we present our submission to the SemEval-2024 Task 8 “Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection”, focusing on the detection of machine-generated texts (MGTs) in English. Specifically, our approach relies on combining embeddings from the RoBERTa-base with diversity features and uses a resampled training set. We score 16th from 139 in the ranking for Subtask A, and our results show that our approach is generalizable across unseen models and domains, achieving an accuracy of 0.91.

In recent years, the main focus of research on automatic readability assessment (ARA) has shifted towards using expensive deep learning-based methods with the primary goal of increasing models’ accuracy. This, however, is rarely applicable for low-resource languages where traditional handcrafted features are still widely used due to the lack of existing NLP tools to extract deeper linguistic representations. In this work, we take a step back from the technical component and focus on how linguistic aspects such as mutual intelligibility or degree of language relatedness can improve ARA in a low-resource setting. We collect short stories written in three languages in the Philippines—Tagalog, Bikol, and Cebuano—to train readability assessment models and explore the interaction of data and features in various cross-lingual setups. Our results show that the inclusion of CrossNGO, a novel specialized feature exploiting n-gram overlap applied to languages with high mutual intelligibility, significantly improves the performance of ARA models compared to the use of off-the-shelf large multilingual language models alone. Consequently, when both linguistic representations are combined, we achieve state-of-the-art results for Tagalog and Cebuano, and baseline scores for ARA in Bikol.

Current research on automatic readability assessment (ARA) has focused on improving the performance of models in high-resource languages such as English. In this work, we introduce and release BasahaCorpus as part of an initiative aimed at expanding available corpora and baseline models for readability assessment in lower resource languages in the Philippines. We compiled a corpus of short fictional narratives written in Hiligaynon, Minasbate, Karay-a, and Rinconada—languages belonging to the Central Philippine family tree subgroup—to train ARA models using surface-level, syllable-pattern, and n-gram overlap features. We also propose a new hierarchical cross-lingual modeling approach that takes advantage of a language’s placement in the family tree to increase the amount of available training data. Our study yields encouraging results that support previous work showcasing the efficacy of cross-lingual models in low-resource settings, as well as similarities in highly informative linguistic features for mutually intelligible languages.

This paper describes the results of the first shared task on generation of teacher responses in educational dialogues. The goal of the task was to benchmark the ability of generative language models to act as AI teachers, replying to a student in a teacher-student dialogue. Eight teams participated in the competition hosted on CodaLab and experimented with a wide variety of state-of-the-art models, including Alpaca, Bloom, DialoGPT, DistilGPT-2, Flan-T5, GPT- 2, GPT-3, GPT-4, LLaMA, OPT-2.7B, and T5- base. Their submissions were automatically scored using BERTScore and DialogRPT metrics, and the top three among them were further manually evaluated in terms of pedagogical ability based on Tack and Piech (2022). The NAISTeacher system, which ranked first in both automated and human evaluation, generated responses with GPT-3.5 Turbo using an ensemble of prompts and DialogRPT-based ranking of responses for given dialogue contexts. Despite promising achievements of the participating teams, the results also highlight the need for evaluation metrics better suited to educational contexts.

Lexical complexity is a highly subjective notion, yet this factor is often neglected in lexical simplification and readability systems which use a ”one-size-fits-all” approach. In this paper, we investigate which aspects contribute to the notion of lexical complexity in various groups of readers, focusing on native and non-native speakers of English, and how the notion of complexity changes depending on the proficiency level of a non-native reader. To facilitate reproducibility of our approach and foster further research into these aspects, we release a dataset of complex words annotated by readers with different backgrounds.

Multiword expressions (MWEs) represent lexemes that should be treated as single lexical units due to their idiosyncratic nature. Multiple NLP applications have been shown to benefit from MWE identification, however the research on lexical complexity of MWEs is still an under-explored area. In this work, we re-annotate the Complex Word Identification Shared Task 2018 dataset of Yimam et al. (2017), which provides complexity scores for a range of lexemes, with the types of MWEs. We release the MWE-annotated dataset with this paper, and we believe this dataset represents a valuable resource for the text simplification community. In addition, we investigate which types of expressions are most problematic for native and non-native readers. Finally, we show that a lexical complexity assessment system benefits from the information about MWE types.

The Sense Complexity Dataset (SeCoDa) provides a corpus that is annotated jointly for complexity and word senses. It thus provides a valuable resource for both word sense disambiguation and the task of complex word identification. The intention is that this dataset will be used to identify complexity at the level of word senses rather than word tokens. For word sense annotation SeCoDa uses a hierarchical scheme that is based on information available in the Cambridge Advanced Learner’s Dictionary. This way we can offer more coarse-grained senses than directly available in WordNet.

This paper describes a semi-supervised system that jointly learns verbal multiword expressions (VMWEs) and dependency parse trees as an auxiliary task. The model benefits from pre-trained multilingual BERT. BERT hidden layers are shared among the two tasks and we introduce an additional linear layer to retrieve VMWE tags. The dependency parse tree prediction is modelled by a linear layer and a bilinear one plus a tree CRF architecture on top of the shared BERT. The system has participated in the open track of the PARSEME shared task 2020 and ranked first in terms of F1-score in identifying unseen VMWEs as well as VMWEs in general, averaged across all 14 languages.

Multiword expressions (MWEs) were shown to be useful in a number of NLP tasks. However, research on the use of MWEs in lexical complexity assessment and simplification is still an under-explored area. In this paper, we propose a text complexity assessment system for English, which incorporates MWE identification. We show that detecting MWEs using state-of-the-art systems improves predicting complexity on an established lexical complexity dataset.

Complex Word Identification (CWI) is concerned with detection of words in need of simplification and is a crucial first step in a simplification pipeline. It has been shown that reliable CWI systems considerably improve text simplification. However, most CWI systems to date address the task on a word-by-word basis, not taking the context into account. In this paper, we present a novel approach to CWI based on sequence modelling. Our system is capable of performing CWI in context, does not require extensive feature engineering and outperforms state-of-the-art systems on this task.

Automating the assessment of learner summary provides a useful tool for assessing learner reading comprehension. We present a summarization task for evaluating non-native reading comprehension and propose three novel approaches to automatically assess the learner summaries. We evaluate our models on two datasets we created and show that our models outperform traditional approaches that rely on exact word match on this task. Our best model produces quality assessments close to professional examiners.

This paper presents a novel architecture for recursive context-aware lexical simplification, REC-LS, that is capable of (1) making use of the wider context when detecting the words in need of simplification and suggesting alternatives, and (2) taking previous simplification steps into account. We show that our system outputs lexical simplifications that are grammatically correct and semantically appropriate, and outperforms the current state-of-the-art systems in lexical simplification.

Lexical simplification systems replace complex words with simple ones based on a model of which words are complex in context. We explore how users can help train complex word identification models through labelling more efficiently and reliably. We show that using an interface where annotators make comparative rather than binary judgments leads to more reliable and consistent labels, and explore whether comparative judgments may provide a faster way for collecting labels.

This paper presents the winning systems we submitted to the Complex Word Identification Shared Task 2018. We describe our best performing systems’ implementations and discuss our key findings from this research. Our best-performing systems achieve an F1 score of 0.8792 on the NEWS, 0.8430 on the WIKINEWS and 0.8115 on the WIKIPEDIA test sets in the monolingual English binary classification track, and a mean absolute error of 0.0558 on the NEWS, 0.0674 on the WIKINEWS and 0.0739 on the WIKIPEDIA test sets in the probabilistic track.

Using methods of statistical analysis, we investigate how semantic knowledge is acquired in English as a second language and evaluate the pace of development across a number of predicate types and content word combinations, as well as across the levels of language proficiency and native languages. Our exploratory study helps identify the most problematic areas for language learners with different backgrounds and at different stages of learning.

In this paper we discuss three key points related to error detection (ED) in learners’ English. We focus on content word ED as one of the most challenging tasks in this area, illustrating our claims on adjective–noun (AN) combinations. In particular, we (1) investigate the role of context in accurately capturing semantic anomalies and implement a system based on distributional topic coherence, which achieves state-of-the-art accuracy on a standard test set; (2) thoroughly investigate our system’s performance across individual adjective classes, concluding that a class-dependent approach is beneficial to the task; (3) discuss the data size bottleneck in this area, and highlight the challenges of automatic error generation for content words.