Partly automated creation of interlinear glossed text (IGT) has the potential to assist in linguistic documentation. We argue that LLMs can make this process more accessible to linguists because of their capacity to follow natural-language instructions. We investigate the effectiveness of a retrieval-based LLM prompting approach to glossing, applied to the seven languages from the SIGMORPHON 2023 shared task. Our system beats the BERT-based shared task baseline for every language in the morpheme-level score category, and we show that a simple 3-best oracle has higher word-level scores than the challenge winner (a tuned sequence model) in five languages. In a case study on Tsez, we ask the LLM to automatically create and follow linguistic instructions, reducing errors on a confusing grammatical feature. Our results thus demonstrate the potential contributions which LLMs can make in interactive systems for glossing, both in making suggestions to human annotators and following directions.

Negation has been a long-standing challenge for language models. Previous studies have shown that they struggle with negation in many natural language understanding tasks. In this work, we propose a self-supervised method to make language models more robust against negation. We introduce a novel task, Next Sentence Polarity Prediction (NSPP), and a variation of the Next Sentence Prediction (NSP) task. We show that BERT and RoBERTa further pre-trained on our tasks outperform the off-the-shelf versions on nine negation-related benchmarks. Most notably, our pre-training tasks yield between 1.8% and 9.1% improvement on CondaQA, a large question-answering corpus requiring reasoning over negation.

First, effectiveness of their utterances when faced with we extend the investigation to a much larger corpus multiple options for structuring a message. The of informal written English collected from social UID hypothesis (Frank and Jaeger, 2008; Collins, media. Second, we use contemporary large language 2014; Hahn et al., 2020) suggests that speakers models (LLMs) to estimate the operationalizations tend to spread information evenly throughout an of information uniformity in syntactic reduction, utterance, avoiding large fluctuations in the perunit suggesting the robustness of our findings.

Dependency length minimization is a universally observed quantitative property of natural languages. However, the extent of dependency length minimization, and the cognitive mechanisms through which the language processor achieves this minimization remain unclear. This research offers mechanistic insights by postulating that moving a short preverbal constituent next to the main verb explains preverbal constituent ordering decisions better than global minimization of dependency length in SOV languages. This approach constitutes a least-effort strategy because it's just one operation but simultaneously reduces the length of all preverbal dependencies linked to the main verb. We corroborate this strategy using large-scale corpus evidence across all seven SOV languages that are prominently represented in the Universal Dependency Treebank. These findings align with the concept of bounded rationality, where decision-making is influenced by 'quick-yet-economical' heuristics rather than exhaustive searches for optimal solutions. Overall, this work sheds light on the role of bounded rationality in linguistic decision-making and language evolution.

The principle of DEPENDENCY LENGTH MINIMIZATION, which seeks to keep syntactically related words close in a sentence, is thought to universally shape the structure of human languages for effective communication. However, the extent to which dependency length minimization is applied in human language systems is not yet fully understood. Preverbally, the placement of long-before-short constituents and postverbally, short-before-long constituents are known to minimize overall dependency length of a sentence. In this study, we test the hypothesis that placing only the shortest preverbal constituent next to the main-verb explains word order preferences in Hindi (a SOV language) as opposed to the global minimization of dependency length. We characterize this approach as a least-effort strategy because it is a cost-effective way to shorten all dependencies between the verb and its preverbal dependencies. As such, this approach is consistent with the bounded-rationality perspective according to which decision making is governed by "fast but frugal" heuristics rather than by a search for optimal solutions. Consistent with this idea, our results indicate that actual corpus sentences in the Hindi-Urdu Treebank corpus are better explained by the least effort strategy than by global minimization of dependency lengths. Additionally, for the task of distinguishing corpus sentences from counterfactual variants, we find that the dependency length and constituent length of the constituent closest to the main verb are much better predictors of whether a sentence appeared in the corpus than total dependency length. Overall, our findings suggest that cognitive resource constraints play a crucial role in shaping natural languages.

While multilingual language models can improve NLP performance on low-resource languages by leveraging higher-resource languages, they also reduce average performance on all languages (the 'curse of multilinguality'). Here we show another problem with multilingual models: grammatical structures in higher-resource languages bleed into lower-resource languages, a phenomenon we call grammatical structure bias. We show this bias via a novel method for comparing the fluency of multilingual models to the fluency of monolingual Spanish and Greek models: testing their preference for two carefully-chosen variable grammatical structures (optional pronoun-drop in Spanish and optional Subject-Verb ordering in Greek). We find that multilingual BERT is biased toward the English-like setting (explicit pronouns and Subject-Verb-Object ordering) as compared to our monolingual control language model. With our case studies, we hope to bring to light the fine-grained ways in which multilingual models can be biased,and encourage more linguistically-aware fluency evaluation.

Community question answering aims at choosing the most appropriate answer for a given question, which is important in many NLP applications. Previous neural network-based methods consider several different aspects of information through calculating attentions. These different kinds of attentions are always simply summed up and can be seen as a ``single view", causing severe information loss. To overcome this problem, we propose a Multi-View Fusion Neural Network, where each attention component generates a ``view'' of the QA pair and a fusion RNN integrates the generated views to form a more holistic representation.    In this fusion RNN method, a filter gate  collects  important information of  input and directly adds it to the output, which borrows the idea of residual networks.    Experimental results on the WikiQA and SemEval-2016 CQA datasets demonstrate that our proposed model outperforms the state-of-the-art methods.

Next, through the Question Type Selection ArikIturri (Aldabe et al. 2006) is a system developed for the process, the question type is selected. Finally, by means automatic generation of different types of exercise. One of of the Question Construction step, the surface form of the the aims of ArikIturri is to generate items that could form question is created based on the previous steps. As regards part of real scenarios; this is why their creation is based our QG system, the sentence retriever module is responsible on topics that are part of the curriculum. Thus, the system for the Target Selection task and the item generator module is able to automatically generate tests from texts, to be included performs the Question Type Selection and Question Construction in testing tasks. The system is able to produce fill-inthe-blank processes.