Educational dialogue systems have been used to support students and teachers for decades. Such systems rely on explicit pedagogically motivated dialogue rules. With the ease of integrating large language models (LLMs) into dialogue systems, applications have been arising that directly use model responses without the use of human-written rules, raising concerns about their use in classroom settings. Here, we explore how to constrain LLM outputs to generate appropriate and supportive teacher-like responses. We present results comparing the effectiveness of different constraint variations in a zero-shot prompting setting on a large mathematics classroom corpus. Generated outputs are evaluated with human annotation for Fluency, Relevance, Helpfulness, and Adherence to the provided constraints. Including all constraints in the prompt led to the highest values for Fluency and Helpfulness, and the second highest value for Relevance. The annotation results also demonstrate that the prompts that result in the highest adherence to constraints do not necessarily indicate higher perceived scores for Fluency, Relevance, or Helpfulness. In a direct comparison, all of the non-baseline LLM responses were ranked higher than the actual teacher responses in the corpus over 50% of the time.

This paper presents a novel Cross-document Abstract Meaning Representation (X-AMR) annotation tool designed for annotating key corpus-level event semantics. Leveraging machine assistance through the Prodigy Annotation Tool, we enhance the user experience, ensuring ease and efficiency in the annotation process. Through empirical analyses, we demonstrate the effectiveness of our tool in augmenting an existing event corpus, highlighting its advantages when integrated with GPT-4. Code and annotations: href{https://anonymous.4open.science/r/xamr-9ED0}{anonymous.4open.science/r/xamr-9ED0} footnote Demo: {href{https://youtu.be/TuirftxciNE}{https://youtu.be/TuirftxciNE}} footnote Live Link: {href{https://tinyurl.com/mrxmafwh}{https://tinyurl.com/mrxmafwh}}

This paper is dedicated to the design and evaluation of the first AMR parser tailored for clinical notes. Our objective was to facilitate the precise transformation of the clinical notes into structured AMR expressions, thereby enhancing the interpretability and usability of clinical text data at scale. Leveraging the colon cancer dataset from the Temporal Histories of Your Medical Events (THYME) corpus, we adapted a state-of-the-art AMR parser utilizing continuous training. Our approach incorporates data augmentation techniques to enhance the accuracy of AMR structure predictions. Notably, through this learning strategy, our parser achieved an impressive F1 score of 88% on the THYME corpus’s colon cancer dataset. Moreover, our research delved into the efficacy of data required for domain adaptation within the realm of clinical notes, presenting domain adaptation data requirements for AMR parsing. This exploration not only underscores the parser’s robust performance but also highlights its potential in facilitating a deeper understanding of clinical narratives through structured semantic representations.

The most popular Cross-Document Event Coreference Resolution (CDEC) datasets fail to convey the true difficulty of the task, due to the lack of lexical diversity between coreferring event triggers (words or phrases that refer to an event). Furthermore, there is a dearth of event datasets for figurative language, limiting a crucial avenue of research in event comprehension. We address these two issues by introducing ECB+META, a lexically rich variant of Event Coref Bank Plus (ECB+) for CDEC on symbolic and metaphoric language. We use ChatGPT as a tool for the metaphoric transformation of sentences in the documents of ECB+, then tag the original event triggers in the transformed sentences in a semi-automated manner. In this way, we avoid the re-annotation of expensive coreference links. We present results that show existing methods that work well on ECB+ struggle with ECB+META, thereby paving the way for CDEC research on a much more challenging dataset. Code/data: https://github.com/ahmeshaf/llms_coref

This paper reports the first release of the UMR (Uniform Meaning Representation) data set. UMR is a graph-based meaning representation formalism consisting of a sentence-level graph and a document-level graph. The sentence-level graph represents predicate-argument structures, named entities, word senses, aspectuality of events, as well as person and number information for entities. The document-level graph represents coreferential, temporal, and modal relations that go beyond sentence boundaries. UMR is designed to capture the commonalities and variations across languages and this is done through the use of a common set of abstract concepts, relations, and attributes as well as concrete concepts derived from words from invidual languages. This UMR release includes annotations for six languages (Arapaho, Chinese, English, Kukama, Navajo, Sanapana) that vary greatly in terms of their linguistic properties and resource availability. We also describe on-going efforts to enlarge this data set and extend it to other genres and modalities. We also briefly describe the available infrastructure (UMR annotation guidelines and tools) that others can use to create similar data sets.

Event Coreference Resolution (ECR) as a pairwise mention classification task is expensive both for automated systems and manual annotations. The task’s quadratic difficulty is exacerbated when using Large Language Models (LLMs), making prompt engineering for ECR prohibitively costly. In this work, we propose a graphical representation of events, X-AMR, anchored around individual mentions using a cross-document version of Abstract Meaning Representation. We then linearize the ECR with a novel multi-hop coreference algorithm over the event graphs. The event graphs simplify ECR, making it a) LLM cost-effective, b) compositional and interpretable, and c) easily annotated. For a fair assessment, we first enrich an existing ECR benchmark dataset with these event graphs using an annotator-friendly tool we introduce. Then, we employ GPT-4, the newest LLM by OpenAI, for these annotations. Finally, using the ECR algorithm, we assess GPT-4 against humans and analyze its limitations. Through this research, we aim to advance the state-of-the-art for efficient ECR and shed light on the potential shortcomings of current LLMs at this task. Code and annotations: https://github.com/ahmeshaf/gpt_coref

Event coreference resolution (ECR) is the task of determining whether distinct mentions of events within a multi-document corpus are actually linked to the same underlying occurrence. Images of the events can help facilitate resolution when language is ambiguous. Here, we propose a multimodal cross-document event coreference resolution method that integrates visual and textual cues with a simple linear map between vision and language models. As existing ECR benchmark datasets rarely provide images for all event mentions, we augment the popular ECB+ dataset with event-centric images scraped from the internet and generated using image diffusion models. We establish three methods that incorporate images and text for coreference: 1) a standard fused model with finetuning, 2) a novel linear mapping method without finetuning and 3) an ensembling approach based on splitting mention pairs by semantic and discourse-level difficulty. We evaluate on 2 datasets: the augmented ECB+, and AIDA Phase 1. Our ensemble systems using cross-modal linear mapping establish an upper limit (91.9 CoNLL F1) on ECB+ ECR performance given the preprocessing assumptions used, and establish a novel baseline on AIDA Phase 1. Our results demonstrate the utility of multimodal information in ECR for certain challenging coreference problems, and highlight a need for more multimodal resources in the coreference resolution space.

Recent large language models (LLMs) can solve puzzles that require creativity and lateral thinking. To advance this front of research, we tackle SemEval-2024 Task 9: BRAINTEASER: A Novel Task Defying Common Sense. We approach this task by introducing a technique that we call Fine-tuned Generated Chain-of-Thought (FtG-CoT). It is a novel few-shot prompting method that combines a fine-tuned BERT classifier encoder with zero-shot chain-of-thought generation and a fine-tuned LLM. The fine-tuned BERT classifier provides a context-rich encoding of each example question and choice list. Zero-shot chain-of-thought generation leverages the benefits of chain-of-thought prompting without requiring manual creation of the reasoning chains. We fine-tune the LLM on the generated chains-of-thought and include a set of generated reasoning chains in the final few-shot LLM prompt to maximize the relevance and correctness of the final generated response. In this paper, we show that FtG-CoT outperforms the zero-shot prompting baseline presented in the task paper and is highly effective at solving challenging sentence puzzles achieving a perfect score on the practice set and a 0.9 score on the evaluation set.

Event Coreference Resolution (ECR) is the task of linking mentions of the same event either within or across documents. Most mention pairs are not coreferent, yet many that are coreferent can be identified through simple techniques such as lemma matching of the event triggers or the sentences in which they appear. Existing methods for training coreference systems sample from a largely skewed distribution, making it difficult for the algorithm to learn coreference beyond surface matching. Additionally, these methods are intractable because of the quadratic operations needed. To address these challenges, we break the problem of ECR into two parts: a) a heuristic to efficiently filter out a large number of non-coreferent pairs, and b) a training approach on a balanced set of coreferent and non-coreferent mention pairs. By following this approach, we show that we get comparable results to the state of the art on two popular ECR datasets while significantly reducing compute requirements. We also analyze the mention pairs that are “hard” to accurately classify as coreferent or non-coreferentcode repo: github.com/ahmeshaf/lemma_ce_coref.

In this paper, we introduce CAMRA (Copilot for AMR Annotatations), a cutting-edge web-based tool designed for constructing Abstract Meaning Representation (AMR) from natural language text. CAMRA offers a novel approach to deep lexical semantics annotation such as AMR, treating AMR annotation akin to coding in programming languages. Leveraging the familiarity of programming paradigms, CAMRA encompasses all essential features of existing AMR editors, including example lookup, while going a step further by integrating Propbank roleset lookup as an autocomplete feature within the tool. Notably, CAMRA incorporates AMR parser models as coding co-pilots, greatly enhancing the efficiency and accuracy of AMR annotators.

This paper presents detailed mappings between the structures used in Abstract Meaning Representation (AMR) and those used in Uniform Meaning Representation (UMR). These structures include general semantic roles, rolesets, and concepts that are largely shared between AMR and UMR, but with crucial differences. While UMR annotation of new low-resource languages is ongoing, AMR-annotated corpora already exist for many languages, and these AMR corpora are ripe for conversion to UMR format. Rather than focusing on semantic coverage that is new to UMR (which will likely need to be dealt with manually), this paper serves as a resource (with illustrated mappings) for users looking to understand the fine-grained adjustments that have been made to the representation techniques for semantic categoriespresent in both AMR and UMR.

Annotating cross-document event coreference links is a time-consuming and cognitively demanding task that can compromise annotation quality and efficiency. To address this, we propose a model-in-the-loop annotation approach for event coreference resolution, where a machine learning model suggests likely corefering event pairs only. We evaluate the effectiveness of this approach by first simulating the annotation process and then, using a novel annotator-centric Recall-Annotation effort trade-off metric, we compare the results of various underlying models and datasets. We finally present a method for obtaining 97% recall while substantially reducing the workload required by a fully manual annotation process.

Transcripts of teaching episodes can be effective tools to understand discourse patterns in classroom instruction. According to most educational experts, sustained classroom discourse is a critical component of equitable, engaging, and rich learning environments for students. This paper describes the TalkMoves dataset, composed of 567 human-annotated K-12 mathematics lesson transcripts (including entire lessons or portions of lessons) derived from video recordings. The set of transcripts primarily includes in-person lessons with whole-class discussions and/or small group work, as well as some online lessons. All of the transcripts are human-transcribed, segmented by the speaker (teacher or student), and annotated at the sentence level for ten discursive moves based on accountable talk theory. In addition, the transcripts include utterance-level information in the form of dialogue act labels based on the Switchboard Dialog Act Corpus. The dataset can be used by educators, policymakers, and researchers to understand the nature of teacher and student discourse in K-12 math classrooms. Portions of this dataset have been used to develop the TalkMoves application, which provides teachers with automated, immediate, and actionable feedback about their mathematics instruction.

“Talk moves” are specific discursive strategies used by teachers and students to facilitate conversations in which students share their thinking, and actively consider the ideas of others, and engage in rich discussions. Experts in instructional practices often rely on cues to identify and document these strategies, for example by annotating classroom transcripts. Prior efforts to develop automated systems to classify teacher talk moves using transformers achieved a performance of 76.32% F1. In this paper, we investigate the feasibility of using enriched contextual cues to improve model performance. We applied state-of-the-art deep learning approaches for Natural Language Processing (NLP), including Robustly optimized bidirectional encoder representations from transformers (Roberta) with a special input representation that supports previous and subsequent utterances as context for talk moves classification. We worked with the publically available TalkMoves dataset, which contains utterances sourced from real-world classroom sessions (human- transcribed and annotated). Through a series of experimentations, we found that a combination of previous and subsequent utterances improved the transformers’ ability to differentiate talk moves (by 2.6% F1). These results constitute a new state of the art over previously published results and provide actionable insights to those in the broader NLP community who are working to develop similar transformer-based classification models.

We present refinements over existing temporal relation annotations in the Electronic Medical Record clinical narrative. We refined the THYME corpus annotations to more faithfully represent nuanced temporality and nuanced temporal-coreferential relations. The main contributions are in re-defining CONTAINS and OVERLAP relations into CONTAINS, CONTAINS-SUBEVENT, OVERLAP and NOTED-ON. We demonstrate that these refinements lead to substantial gains in learnability for state-of-the-art transformer models as compared to previously reported results on the original THYME corpus. We thus establish a baseline for the automatic extraction of these refined temporal relations. Although our study is done on clinical narrative, we believe it addresses far-reaching challenges that are corpus- and domain- agnostic.

We present a system which parses sentences into Abstract Meaning Representations, improving state-of-the-art results for this task by more than 5%. AMR graphs represent semantic content using linguistic properties such as semantic roles, coreference, negation, and more. The AMR parser does not rely on a syntactic pre-parse, or heavily engineered features, and uses five recurrent neural networks as the key architectural components for inferring AMR graphs.

We present an approach to creating corpora for use in detecting deception in text, including a discussion of the challenges peculiar to this task. Our approach is based on soliciting several types of reviews from writers and was implemented using Amazon Mechanical Turk. We describe the multi-dimensional corpus of reviews built using this approach, available free of charge from LDC as the Boulder Lies and Truth Corpus (BLT-C). Challenges for both corpus creation and the deception detection include the fact that human performance on the task is typically at chance, that the signal is faint, that paid writers such as turkers are sometimes deceptive, and that deception is a complex human behavior; manifestations of deception depend on details of domain, intrinsic properties of the deceiver (such as education, linguistic competence, and the nature of the intention), and specifics of the deceptive act (e.g., lying vs. fabricating.) To overcome the inherent lack of ground truth, we have developed a set of semi-automatic techniques to ensure corpus validity. We present some preliminary results on the task of deception detection which suggest that the BLT-C is an improvement in the quality of resources available for this task.

In times of mass emergency, vast amounts of data are generated via computer-mediated communication (CMC) that are difficult to manually collect and organize into a coherent picture. Yet valuable information is broadcast, and can provide useful insight into time- and safety-critical situations if captured and analyzed efficiently and effectively. We describe a natural language processing component of the EPIC (Empowering the Public with Information in Crisis) Project infrastructure, designed to extract linguistic and behavioral information from tweet text to aid in the task of information integration. The system incorporates linguistic annotation, in the form of Named Entity Tagging, as well as behavioral annotations to capture tweets contributing to situational awareness and analyze the information type of the tweet content. We show classification results and describe future integration of these classifiers in the larger EPIC infrastructure.

While recent corpus annotation efforts cover a wide variety of semantic structures, work on temporal and causal relations is still in its early stages. Annotation efforts have typically considered either temporal relations or causal relations, but not both, and no corpora currently exist that allow the relation between temporals and causals to be examined empirically. We have annotated a corpus of 1000 event pairs for both temporal and causal relations, focusing on a relatively frequent construction in which the events are conjoined by the word “and”. Temporal relations were annotated using an extension of the BEFORE and AFTER scheme used in the TempEval competition, and causal relations were annotated using a scheme based on connective phrases like “and as a result”. The annotators achieved 81.2% agreement on temporal relations and 77.8% agreement on causal relations. Analysis of the resulting corpus revealed some interesting findings, for example, that over 30% of CAUSAL relations do not have an underlying BEFORE relation. The corpus was also explored using machine learning methods, and while model performance exceeded all baselines, the results suggested that simple grammatical cues may be insufficient for identifying the more difficult temporal and causal relations.

This paper summarizes the annotation of fine-grained entailment relationships in the context of student answers to science assessment questions. We annotated a corpus of 15,357 answer pairs with 145,911 fine-grained entailment relationships. We provide the rationale for such fine-grained analysis and discuss its perceived benefits to an Intelligent Tutoring System. The corpus also has potential applications in other areas, such as question answering and multi-document summarization. Annotators achieved 86.2% inter-annotator agreement (Kappa=0.728, corresponding to substantial agreement) annotating the fine-grained facets of reference answers with regard to understanding expressed in student answers and labeling from one of five possible detailed relationship categories. The corpus described in this paper, which is the only one providing such detailed entailment annotations, is available as a public resource for the research community. The corpus is expected to enable application development, not only for intelligent tutoring systems, but also for general textual entailment applications, that is currently not practical.