Daisy Zhe Wang

Anthony Colas, Haodi Ma, Xuanli He et al.

Knowledge Graph (KG)-to-Text Generation has seen recent improvements in generating fluent and informative sentences which describe a given KG. As KGs are widespread across multiple domains and contain important entity-relation information, and as text simplification aims to reduce the complexity of a text while preserving the meaning of the original text, we propose KGSimple, a novel approach to unsupervised text simplification which infuses KG-established techniques in order to construct a simplified KG path and generate a concise text which preserves the original input's meaning. Through an iterative and sampling KG-first approach, our model is capable of simplifying text when starting from a KG by learning to keep important information while harnessing KG-to-text generation to output fluent and descriptive sentences. We evaluate various settings of the KGSimple model on currently-available KG-to-text datasets, demonstrating its effectiveness compared to unsupervised text simplification models which start with a given complex text. Our code is available on GitHub.

Yang Bai, Anthony Colas, Daisy Zhe Wang

Check-worthy claim detection aims at providing plausible misinformation to downstream fact-checking systems or human experts to check. This is a crucial step toward accelerating the fact-checking process. Many efforts have been put into how to identify check-worthy claims from a small scale of pre-collected claims, but how to efficiently detect check-worthy claims directly from a large-scale information source, such as Twitter, remains underexplored. To fill this gap, we introduce MythQA, a new multi-answer open-domain question answering(QA) task that involves contradictory stance mining for query-based large-scale check-worthy claim detection. The idea behind this is that contradictory claims are a strong indicator of misinformation that merits scrutiny by the appropriate authorities. To study this task, we construct TweetMythQA, an evaluation dataset containing 522 factoid multi-answer questions based on controversial topics. Each question is annotated with multiple answers. Moreover, we collect relevant tweets for each distinct answer, then classify them into three categories: "Supporting", "Refuting", and "Neutral". In total, we annotated 5.3K tweets. Contradictory evidence is collected for all answers in the dataset. Finally, we present a baseline system for MythQA and evaluate existing NLP models for each system component using the TweetMythQA dataset. We provide initial benchmarks and identify key challenges for future models to improve upon. Code and data are available at: https://github.com/TonyBY/Myth-QA

Jayetri Bardhan, Anthony Colas, Kirk Roberts et al.

This paper develops the first question answering dataset (DrugEHRQA) containing question-answer pairs from both structured tables and unstructured notes from a publicly available Electronic Health Record (EHR). EHRs contain patient records, stored in structured tables and unstructured clinical notes. The information in structured and unstructured EHRs is not strictly disjoint: information may be duplicated, contradictory, or provide additional context between these sources. Our dataset has medication-related queries, containing over 70,000 question-answer pairs. To provide a baseline model and help analyze the dataset, we have used a simple model (MultimodalEHRQA) which uses the predictions of a modality selection network to choose between EHR tables and clinical notes to answer the questions. This is used to direct the questions to the table-based or text-based state-of-the-art QA model. In order to address the problem arising from complex, nested queries, this is the first time Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers (RAT-SQL) has been used to test the structure of query templates in EHR data. Our goal is to provide a benchmark dataset for multi-modal QA systems, and to open up new avenues of research in improving question answering over EHR structured data by using context from unstructured clinical data.

Anthony Colas, Mehrdad Alvandipour, Daisy Zhe Wang

Recent improvements in KG-to-text generation are due to additional auxiliary pre-training tasks designed to give the fine-tune task a boost in performance. These tasks require extensive computational resources while only suggesting marginal improvements. Here, we demonstrate that by fusing graph-aware elements into existing pre-trained language models, we are able to outperform state-of-the-art models and close the gap imposed by additional pre-training tasks. We do so by proposing a mask structure to capture neighborhood information and a novel type encoder that adds a bias to the graph-attention weights depending on the connection type. Experiments on two KG-to-text benchmark datasets show our models are competitive while involving fewer parameters and no additional pre-training tasks. By formulating the problem as a framework, we can interchange the various proposed components and begin interpreting KG-to-text generative models based on the topological and type information found in a graph.

Bushi Xiao, Sarvesh Soni, Daisy Zhe Wang

As large language models are increasingly deployed for clinical text, ensuring they can reliably signal their own uncertainty becomes critical. Most existing uncertainty quantification (UQ) methods are designed for open-domain generation and cannot localize uncertainty at the token or span level in long clinical text. We propose Reverse Probing, the first UQ framework specialized for clinical summarization, which estimates token-level uncertainty directly from pre-existing labeled summaries. Rather than sampling new outputs, Reverse Probing treats the text as a probe into the model's internal state, extracting uncertainty signals from four categories of internal activations. We evaluate on two expert-annotated clinical datasets and outperform eight adapted baselines on all metrics, achieving up to 4 times higher AUPRC while reducing inference time and computational costs. Feature analysis reveals that delta energy and neighborhood context are the most consistent predictors across all models. This study offers interpretable insights into how models internally respond to unsupported clinical content.

Haodi Ma, Daisy Zhe Wang

Knowledge graphs (KG) have served as the key component of various natural language processing applications. Commonsense knowledge graphs (CKG) are a special type of KG, where entities and relations are composed of free-form text. However, previous works in KG completion and CKG completion suffer from long-tail relations and newly-added relations which do not have many know triples for training. In light of this, few-shot KG completion (FKGC), which requires the strengths of graph representation learning and few-shot learning, has been proposed to challenge the problem of limited annotated data. In this paper, we comprehensively survey previous attempts on such tasks in the form of a series of methods and applications. Specifically, we first introduce FKGC challenges, commonly used KGs, and CKGs. Then we systematically categorize and summarize existing works in terms of the type of KGs and the methods. Finally, we present applications of FKGC models on prediction tasks in different areas and share our thoughts on future research directions of FKGC.

Yang Peng, Daisy Zhe Wang

Over the past few years, large knowledge bases have been constructed to store massive amounts of knowledge. However, these knowledge bases are highly incomplete. To solve this problem, we propose a web-based question answering system system with multimodal fusion of unstructured and structured information, to fill in missing information for knowledge bases. To utilize unstructured information from the Web for knowledge base completion, we design a web-based question answering system using multimodal features and question templates to extract missing facts, which can achieve good performance with very few questions. To help improve extraction quality, the question answering system employs structured information from knowledge bases, such as entity types and entity-to-entity relatedness.

Yifan Wang, Daisy Zhe Wang

Density-based clustering is a commonly used tool in data science. Today many data science works are utilizing high-dimensional neural embeddings. However, traditional density-based clustering techniques like DBSCAN have a degraded performance on high-dimensional data. In this paper, we propose LAF, a generic learned accelerator framework to speed up the original DBSCAN and the sampling-based variants of DBSCAN on high-dimensional data with angular distance metric. This framework consists of a learned cardinality estimator and a post-processing module. The cardinality estimator can fast predict whether a data point is core or not to skip unnecessary range queries, while the post-processing module detects the false negative predictions and merges the falsely separated clusters. The evaluation shows our LAF-enhanced DBSCAN method outperforms the state-of-the-art efficient DBSCAN variants on both efficiency and quality.

Yang Peng, Daisy Zhe Wang

Over the past few years, large knowledge bases have been constructed to store massive amounts of knowledge. However, these knowledge bases are highly incomplete, for example, over 70% of people in Freebase have no known place of birth. To solve this problem, we propose a query-driven knowledge base completion system with multimodal fusion of unstructured and structured information. To effectively fuse unstructured information from the Web and structured information in knowledge bases to achieve good performance, our system builds multimodal knowledge graphs based on question answering and rule inference. We propose a multimodal path fusion algorithm to rank candidate answers based on different paths in the multimodal knowledge graphs, achieving much better performance than question answering, rule inference and a baseline fusion algorithm. To improve system efficiency, query-driven techniques are utilized to reduce the runtime of our system, providing fast responses to user queries. Extensive experiments have been conducted to demonstrate the effectiveness and efficiency of our system.

Haodi Ma, Anthony Colas, Yuejie Wang et al.

Recent works in neural knowledge graph inference attempt to combine logic rules with knowledge graph embeddings to benefit from prior knowledge. However, they usually cannot avoid rule grounding, and injecting a diverse set of rules has still not been thoroughly explored. In this work, we propose InjEx, a mechanism to inject multiple types of rules through simple constraints, which capture definite Horn rules. To start, we theoretically prove that InjEx can inject such rules. Next, to demonstrate that InjEx infuses interpretable prior knowledge into the embedding space, we evaluate InjEx on both the knowledge graph completion (KGC) and few-shot knowledge graph completion (FKGC) settings. Our experimental results reveal that InjEx outperforms both baseline KGC models as well as specialized few-shot models while maintaining its scalability and efficiency.

Aashish Dhawan, Christopher Driggers-Ellis, Dzmitry Kasinets et al.

We present the University of Florida Gators submission to the AmericasNLP 2026 shared task on cultural image captioning for Indigenous languages. Our two-stage pipeline generates a Spanish intermediate caption with Qwen2.5-VL, then produces the target-language caption using retrieval-augmented many-shot prompting with Gemini 2.5 Flash. We achieve 164.1%, 131.7%, and 122.6% improvements over the shared task baseline for Bribri, Guaraní, and Orizaba Nahuatl captioning, respectively, in our dev set evaluation and maintain >150% improvements for the Bribri and Orizaba Nahuatl languages in the test set evaluation. We find retrieval is highly language-dependent, beneficial only for large, in-domain corpora, and that synthetic data augmentation accounts for around 28 chrF++ of the dev set Guaraní performance gain. Our submission is the overall winner of the shared task, placing second out of five finalist submissions in human evaluations of target-language captions.

Jayetri Bardhan, Kirk Roberts, Daisy Zhe Wang

Question Answering (QA) systems on patient-related data can assist both clinicians and patients. They can, for example, assist clinicians in decision-making and enable patients to have a better understanding of their medical history. Significant amounts of patient data are stored in Electronic Health Records (EHRs), making EHR QA an important research area. In EHR QA, the answer is obtained from the medical record of the patient. Because of the differences in data format and modality, this differs greatly from other medical QA tasks that employ medical websites or scientific papers to retrieve answers, making it critical to research EHR question answering. This study aimed to provide a methodological review of existing works on QA over EHRs. We searched for articles from January 1st, 2005 to September 30th, 2023 in four digital sources including Google Scholar, ACL Anthology, ACM Digital Library, and PubMed to collect relevant publications on EHR QA. 4111 papers were identified for our study, and after screening based on our inclusion criteria, we obtained a total of 47 papers for further study. Out of the 47 papers, 25 papers were about EHR QA datasets, and 37 papers were about EHR QA models. It was observed that QA on EHRs is relatively new and unexplored. Most of the works are fairly recent. Also, it was observed that emrQA is by far the most popular EHR QA dataset, both in terms of citations and usage in other papers. Furthermore, we identified the different models used in EHR QA along with the evaluation metrics used for these models.

Yang Bai, Christan Earl Grant, Daisy Zhe Wang

Multi-modal retrieval-augmented Question Answering (MRAQA), integrating text and images, has gained significant attention in information retrieval (IR) and natural language processing (NLP). Traditional ranking methods rely on small encoder-based language models, which are incompatible with modern decoder-based generative large language models (LLMs) that have advanced various NLP tasks. To bridge this gap, we propose RAMQA, a unified framework combining learning-to-rank methods with generative permutation-enhanced ranking techniques. We first train a pointwise multi-modal ranker using LLaVA as the backbone. Then, we apply instruction tuning to train a LLaMA model for re-ranking the top-k documents using an innovative autoregressive multi-task learning approach. Our generative ranking model generates re-ranked document IDs and specific answers from document candidates in various permutations. Experiments on two MRAQA benchmarks, WebQA and MultiModalQA, show significant improvements over strong baselines, highlighting the effectiveness of our approach. Code and data are available at: https://github.com/TonyBY/RAMQA

Yang Bai, Anthony Colas, Christan Grant et al.

In recent research, contrastive learning has proven to be a highly effective method for representation learning and is widely used for dense retrieval. However, we identify that relying solely on contrastive learning can lead to suboptimal retrieval performance. On the other hand, despite many retrieval datasets supporting various learning objectives beyond contrastive learning, combining them efficiently in multi-task learning scenarios can be challenging. In this paper, we introduce M3, an advanced recursive Multi-hop dense sentence retrieval system built upon a novel Multi-task Mixed-objective approach for dense text representation learning, addressing the aforementioned challenges. Our approach yields state-of-the-art performance on a large-scale open-domain fact verification benchmark dataset, FEVER. Code and data are available at: https://github.com/TonyBY/M3

Jayetri Bardhan, Bushi Xiao, Daisy Zhe Wang

Question answering (QA) over tables and text has gained much popularity over the years. Multi-hop table-text QA requires multiple hops between the table and text, making it a challenging QA task. Although several works have attempted to solve the table-text QA task, most involve training the models and requiring labeled data. In this paper, we have proposed a Retrieval Augmented Generation (RAG) based model - TTQA-RS: A break-down prompting approach for Multi-hop Table-Text Question Answering with Reasoning and Summarization. Our model uses an enhanced retriever for table-text information retrieval and uses augmented knowledge, including table-text summary with decomposed sub-questions with answers for a reasoning-based table-text QA. Using open-source language models, our model outperformed all existing prompting methods for table-text QA tasks on existing table-text QA datasets, such as HybridQA and OTT-QA's development set. Our experiments demonstrate the potential of prompt-based approaches using open-source LLMs. Additionally, by using LLaMA3-70B, our model achieved state-of-the-art performance for prompting-based methods on multi-hop table-text QA.

Ruyu Li, Tinghui Zhang, Haodi Ma et al.

With the increasingly use of multi-modal data, semantic query has become more and more demanded in data management systems, which is an important way to access and analyze multi-modal data. As unstructured data, most information of multi-modal data (text, image, video, etc) hides in the semantics, which cannot be accessed by the traditional database queries like SQL. Given the power of Large Language Model (LLM) in understanding semantics and processing natural language, in recent years several LLM-based semantic query systems have been proposed, to support semantic querying over unstructured data. However, this rapid growth has produced a fragmented ecosystem. Applications face significant integration challenges due to (1) disparate APIs of different semantic query systems and (2) a fundamental trade-off between specialization and generality. Many semantic query systems are highly specialized, offering state-of-the-art performance within a single modality but struggling with multi-modal data. Conversely, some "all-in-one" systems handle multiple modalities but often exhibit suboptimal performance compared to their specialized counterparts in specific modalities. This paper introduces Meta Engine, a novel "query system on query systems", designed to resolve those aforementioned challenges. Meta Engine is a unified semantic query engine that integrates heterogeneous, specialized LLM-based query systems. Its architecture comprises five key components: (1) a Natural Language (NL) Query Parser, (2) an Operator Generator, (3) a Query Router, (4) a set of Adapters, and (5) a Result Aggregator. In the evaluation, Meta Engine consistently outperforms all baselines, yielding 3-6x higher F1 in most cases and up to 24x on specific datasets.

Aashish Dhawan, Christopher Driggers-Ellis, Christan Grant et al.

Low-resource indigenous languages often lack the parallel corpora required for effective neural machine translation (NMT). Synthetic data generation offers a practical strategy for mitigating this limitation in data-scarce settings. In this work, we augment curated parallel datasets for indigenous languages of the Americas with synthetic sentence pairs generated using a high-capacity multilingual translation model. We fine-tune a multilingual mBART model on curated-only and synthetically augmented data and evaluate translation quality using chrF++, the primary metric used in recent AmericasNLP shared tasks for agglutinative languages. We further apply language-specific preprocessing, including orthographic normalization and noise-aware filtering, to reduce corpus artifacts. Experiments on Guarani-Spanish and Quechua-Spanish translation show consistent chrF++ improvements from synthetic data augmentation, while diagnostic experiments on Aymara highlight the limitations of generic preprocessing for highly agglutinative languages.

Tinghui Zhang, Yifan Wang, Daisy Zhe Wang

Prompt compression methods enhance the efficiency of Large Language Models (LLMs) and minimize the cost by reducing the length of input context. The goal of prompt compression is to shorten the LLM prompt while maintaining a high generation quality. However, existing solutions, mainly based on token removal, face challenges such as information loss and structural incoherence, like missing grammar elements in a sentence, or incomplete word phrases after token removal. Such challenges limit the final generation quality of LLM. To overcome these limitations, we present a novel generative prompt compression method. Unlike the existing token removal methods, our method centers at a chunking-and-summarization mechanism. Specifically, our method splits prompt into semantically coherent chunks and rewrites the chunks to be more concise. The chunks are reconstructed into meaningful prompt finally. We design several optimization techniques for the mechanism, including optimized semantic chunking, outlier chunk handling, dynamic compression ratio, compression prioritization, and keyword maintaining. These techniques effectively improve the identifying and preserving of critical information and coherence among texts, as well as providing finer grind control of the compression ratio. We conduct extensive evaluation on question-answering and summarization tasks, with datasets covering multiple different domain. The evaluation shows our method achieves a significantly better compression quality, and higher stability than the state-of-the-art methods, especially under high compression ratio, which proves the effectiveness and practicality of our method.

Yifan Wang, Haodi Ma, Daisy Zhe Wang

Large language model (LLM) has marked a pivotal moment in the field of machine learning and deep learning. Recently its capability for query planning has been investigated, including both single-modal and multi-modal queries. However, there is no work on the query optimization capability of LLM. As a critical (or could even be the most important) step that significantly impacts the execution performance of the query plan, such analysis and attempts should not be missed. From another aspect, existing query optimizers are usually rule-based or rule-based + cost-based, i.e., they are dependent on manually created rules to complete the query plan rewrite/transformation. Given the fact that modern optimizers include hundreds to thousands of rules, designing a multi-modal query optimizer following a similar way is significantly time-consuming since we will have to enumerate as many multi-modal optimization rules as possible, which has not been well addressed today. In this paper, we investigate the query optimization ability of LLM and use LLM to design LaPuda, a novel LLM and Policy based multi-modal query optimizer. Instead of enumerating specific and detailed rules, LaPuda only needs a few abstract policies to guide LLM in the optimization, by which much time and human effort are saved. Furthermore, to prevent LLM from making mistakes or negative optimization, we borrow the idea of gradient descent and propose a guided cost descent (GCD) algorithm to perform the optimization, such that the optimization can be kept in the correct direction. In our evaluation, our methods consistently outperform the baselines in most cases. For example, the optimized plans generated by our methods result in 1~3x higher execution speed than those by the baselines.

Yifan Wang, Vyom Pathak, Daisy Zhe Wang

Similarity join finds all pairs of close points within a given distance threshold. Many similarity join methods have been proposed, but they are usually not efficient on high-dimensional space due to the curse of dimensionality and data-unawareness. We investigate the possibility of using metric space Bloom filter (MSBF), a family of data structures checking if a query point has neighbors in a multi-dimensional space, to speed up similarity join. However, there are several challenges when applying MSBF to similarity join, including excessive information loss, data-unawareness and hard constraint on the distance metric. In this paper, we propose Xling, a generic framework to build a learning-based metric space filter with any existing regression model, aiming at accurately predicting whether a query point has enough number of neighbors. The framework provides a suite of optimization strategies to further improve the prediction quality based on the learning model, which has demonstrated significantly higher prediction quality than existing MSBF. We also propose XJoin, one of the first filter-based similarity join methods, based on Xling. By predicting and skipping those queries without enough neighbors, XJoin can effectively reduce unnecessary neighbor searching and therefore it achieves a remarkable acceleration. Benefiting from the generalization capability of deep learning models, XJoin can be easily transferred onto new dataset (in similar distribution) without re-training. Furthermore, Xling is not limited to being applied in XJoin, instead, it acts as a flexible plugin that can be inserted to any loop-based similarity join methods for a speedup.

Christopher Driggers-Ellis, Detravious Brinkley, Ray Chen et al.

Multi30k is frequently cited in the multimodal machine translation (MMT) literature, offering parallel text data for training and fine-tuning deep learning models. However, it is limited to four languages: Czech, English, French, and German. This restriction has led many researchers to focus their investigations only on these languages. As a result, MMT research on diverse languages has been stalled because the official Multi30k dataset only represents European languages in Latin scripts. Previous efforts to extend Multi30k exist, but the list of supported languages, represented language families, and scripts is still very short. To address these issues, we propose MultiScript30k, a new Multi30k dataset extension for global languages in various scripts, created by translating the English version of Multi30k (Multi30k-En) using NLLB200-3.3B. The dataset consists of over \(30000\) sentences and provides translations of all sentences in Multi30k-En into Ar, Es, Uk, Zh\_Hans and Zh\_Hant. Similarity analysis shows that Multi30k extension consistently achieves greater than \(0.8\) cosine similarity and symmetric KL divergence less than \(0.000251\) for all languages supported except Zh\_Hant which is comparable to the previous Multi30k extensions ArEnMulti30k and Multi30k-Uk. COMETKiwi scores reveal mixed assessments of MultiScript30k as a translation of Multi30k-En in comparison to the related work. ArEnMulti30k scores nearly equal MultiScript30k-Ar, but Multi30k-Uk scores $6.4\%$ greater than MultiScript30k-Uk per split.

Haodi Ma, Vyom Pathak, Daisy Zhe Wang

Video Question Answering (VQA) requires models to reason over spatial, temporal, and causal cues in videos. Recent vision language models (VLMs) achieve strong results but often rely on shallow correlations, leading to weak temporal grounding and limited interpretability. We study symbolic scene graphs (SGs) as intermediate grounding signals for VQA. SGs provide structured object-relation representations that complement VLMs holistic reasoning. We introduce SG-VLM, a modular framework that integrates frozen VLMs with scene graph grounding via prompting and visual localization. Across three benchmarks (NExT-QA, iVQA, ActivityNet-QA) and multiple VLMs (QwenVL, InternVL), SG-VLM improves causal and temporal reasoning and outperforms prior baselines, though gains over strong VLMs are limited. These findings highlight both the promise and current limitations of symbolic grounding, and offer guidance for future hybrid VLM-symbolic approaches in video understanding.

Bushi Xiao, Michael Bennie, Jayetri Bardhan et al.

Structural priming is a cognitive phenomenon where exposure to a particular syntactic structure increases the likelihood of producing the same structure in subsequent utterances. While humans consistently demonstrate structural priming effects across various linguistic contexts, it remains unclear whether multimodal large language models (MLLMs) exhibit similar syntactic preservation behaviors. We introduce PRISMATIC, the first multimodal structural priming dataset, which advances computational linguistics by providing a standardized benchmark for investigating syntax-vision interactions. We propose the Syntactic Preservation Index (SPI), a novel reference-free evaluation metric designed specifically to assess structural priming effects in sentence level. Using this metric, we constructed and tested models with two different multimodal encoding architectures to investigate their structural preservation capabilities. Our experimental results demonstrate that models with both encoding methods show comparable syntactic priming effects. However, only fusion-encoded models exhibit robust positive correlations between priming effects and visual similarity, suggesting a cognitive process more aligned with human psycholinguistic patterns. This work provides new insights into evaluating and understanding how syntactic information is processed in multimodal language models.

Haodi Ma, Dzmitry Kasinets, Daisy Zhe Wang

Multimodal knowledge graph completion (MMKGC) aims to predict missing links in multimodal knowledge graphs (MMKGs) by leveraging information from various modalities alongside structural data. Existing MMKGC approaches primarily extend traditional knowledge graph embedding (KGE) models, which often require creating an embedding for every entity. This results in large model sizes and inefficiencies in integrating multimodal information, particularly for real-world graphs. Meanwhile, Transformer-based models have demonstrated competitive performance in knowledge graph completion (KGC). However, their focus on single-modal knowledge limits their capacity to utilize cross-modal information. Recently, Large vision-language models (VLMs) have shown potential in cross-modal tasks but are constrained by the high cost of training. In this work, we propose a novel approach that integrates Transformer-based KGE models with cross-modal context generated by pre-trained VLMs, thereby extending their applicability to MMKGC. Specifically, we employ a pre-trained VLM to transform relevant visual information from entities and their neighbors into textual sequences. We then frame KGC as a sequence-to-sequence task, fine-tuning the model with the generated cross-modal context. This simple yet effective method significantly reduces model size compared to traditional KGE approaches while achieving competitive performance across multiple large-scale datasets with minimal hyperparameter tuning.

Shibo Hao, Yi Gu, Haodi Ma et al.

Large language models (LLMs) have shown remarkable reasoning capabilities, especially when prompted to generate intermediate reasoning steps (e.g., Chain-of-Thought, CoT). However, LLMs can still struggle with problems that are easy for humans, such as generating action plans for executing tasks in a given environment, or performing complex math, logical, and commonsense reasoning. The deficiency stems from the key fact that LLMs lack an internal $\textit{world model}$ to predict the world $\textit{state}$ (e.g., environment status, intermediate variable values) and simulate long-term outcomes of actions. This prevents LLMs from performing deliberate planning akin to human brains, which involves exploring alternative reasoning paths, anticipating future states and rewards, and iteratively refining existing reasoning steps. To overcome the limitations, we propose a new LLM reasoning framework, $\underline{R}$easoning vi$\underline{a}$ $\underline{P}$lanning $\textbf{(RAP)}$. RAP repurposes the LLM as both a world model and a reasoning agent, and incorporates a principled planning algorithm (based on Monto Carlo Tree Search) for strategic exploration in the vast reasoning space. During reasoning, the LLM (as agent) incrementally builds a reasoning tree under the guidance of the LLM (as world model) and task-specific rewards, and obtains a high-reward reasoning path efficiently with a proper balance between exploration $\textit{vs.}$ exploitation. We apply RAP to a variety of challenging reasoning problems including plan generation, math reasoning, and logical inference. Empirical results on these tasks demonstrate the superiority of RAP over various strong baselines, including CoT and least-to-most prompting with self-consistency. RAP on LLAMA-33B surpasses CoT on GPT-4 with 33% relative improvement in a plan generation setting.

Anthony Colas, Ali Sadeghian, Yue Wang et al.

We introduce EventNarrative, a knowledge graph-to-text dataset from publicly available open-world knowledge graphs. Given the recent advances in event-driven Information Extraction (IE), and that prior research on graph-to-text only focused on entity-driven KGs, this paper focuses on event-centric data. However, our data generation system can still be adapted to other other types of KG data. Existing large-scale datasets in the graph-to-text area are non-parallel, meaning there is a large disconnect between the KGs and text. The datasets that have a paired KG and text, are small scale and manually generated or generated without a rich ontology, making the corresponding graphs sparse. Furthermore, these datasets contain many unlinked entities between their KG and text pairs. EventNarrative consists of approximately 230,000 graphs and their corresponding natural language text, 6 times larger than the current largest parallel dataset. It makes use of a rich ontology, all of the KGs entities are linked to the text, and our manual annotations confirm a high data quality. Our aim is two-fold: help break new ground in event-centric research where data is lacking, and to give researchers a well-defined, large-scale dataset in order to better evaluate existing and future knowledge graph-to-text models. We also evaluate two types of baseline on EventNarrative: a graph-to-text specific model and two state-of-the-art language models, which previous work has shown to be adaptable to the knowledge graph-to-text domain.

Yang Bai, Daisy Zhe Wang

Textual Question Answering (QA) aims to provide precise answers to user's questions in natural language using unstructured data. One of the most popular approaches to this goal is machine reading comprehension(MRC). In recent years, many novel datasets and evaluation metrics based on classical MRC tasks have been proposed for broader textual QA tasks. In this paper, we survey 47 recent textual QA benchmark datasets and propose a new taxonomy from an application point of view. In addition, We summarize 8 evaluation metrics of textual QA tasks. Finally, we discuss current trends in constructing textual QA benchmarks and suggest directions for future work.

Ali Sadeghian, Mohammadreza Armandpour, Anthony Colas et al.

Despite the importance and abundance of temporal knowledge graphs, most of the current research has been focused on reasoning on static graphs. In this paper, we study the challenging problem of inference over temporal knowledge graphs. In particular, the task of temporal link prediction. In general, this is a difficult task due to data non-stationarity, data heterogeneity, and its complex temporal dependencies. We propose Chronological Rotation embedding (ChronoR), a novel model for learning representations for entities, relations, and time. Learning dense representations is frequently used as an efficient and versatile method to perform reasoning on knowledge graphs. The proposed model learns a k-dimensional rotation transformation parametrized by relation and time, such that after each fact's head entity is transformed using the rotation, it falls near its corresponding tail entity. By using high dimensional rotation as its transformation operator, ChronoR captures rich interaction between the temporal and multi-relational characteristics of a Temporal Knowledge Graph. Experimentally, we show that ChronoR is able to outperform many of the state-of-the-art methods on the benchmark datasets for temporal knowledge graph link prediction.

Anthony Colas, Seokhwan Kim, Franck Dernoncourt et al.

Despite the number of currently available datasets on video question answering, there still remains a need for a dataset involving multi-step and non-factoid answers. Moreover, relying on video transcripts remains an under-explored topic. To adequately address this, We propose a new question answering task on instructional videos, because of their verbose and narrative nature. While previous studies on video question answering have focused on generating a short text as an answer, given a question and video clip, our task aims to identify a span of a video segment as an answer which contains instructional details with various granularities. This work focuses on screencast tutorial videos pertaining to an image editing program. We introduce a dataset, TutorialVQA, consisting of about 6,000manually collected triples of (video, question, answer span). We also provide experimental results with several baselines algorithms using the video transcripts. The results indicate that the task is challenging and call for the investigation of new algorithms.

Ali Sadeghian, Mohammadreza Armandpour, Patrick Ding et al.

In this paper, we study the problem of learning probabilistic logical rules for inductive and interpretable link prediction. Despite the importance of inductive link prediction, most previous works focused on transductive link prediction and cannot manage previously unseen entities. Moreover, they are black-box models that are not easily explainable for humans. We propose DRUM, a scalable and differentiable approach for mining first-order logical rules from knowledge graphs which resolves these problems. We motivate our method by making a connection between learning confidence scores for each rule and low-rank tensor approximation. DRUM uses bidirectional RNNs to share useful information across the tasks of learning rules for different relations. We also empirically demonstrate the efficiency of DRUM over existing rule mining methods for inductive link prediction on a variety of benchmark datasets.

Ali Sadeghian, Shervin Minaee, Ioannis Partalas et al.

We propose a neural network architecture for learning vector representations of hotels. Unlike previous works, which typically only use user click information for learning item embeddings, we propose a framework that combines several sources of data, including user clicks, hotel attributes (e.g., property type, star rating, average user rating), amenity information (e.g., the hotel has free Wi-Fi or free breakfast), and geographic information. During model training, a joint embedding is learned from all of the above information. We show that including structured attributes about hotels enables us to make better predictions in a downstream task than when we rely exclusively on click data. We train our embedding model on more than 40 million user click sessions from a leading online travel platform and learn embeddings for more than one million hotels. Our final learned embeddings integrate distinct sub-embeddings for user clicks, hotel attributes, and geographic information, providing an interpretable representation that can be used flexibly depending on the application. We show empirically that our model generates high-quality representations that boost the performance of a hotel recommendation system in addition to other applications. An important advantage of the proposed neural model is that it addresses the cold-start problem for hotels with insufficient historical click information by incorporating additional hotel attributes which are available for all hotels.

Xiaofeng Zhou, Ali Sadeghian, Daisy Zhe Wang

Multiple web-scale Knowledge Bases, e.g., Freebase, YAGO, NELL, have been constructed using semi-supervised or unsupervised information extraction techniques and many of them, despite their large sizes, are continuously growing. Much research effort has been put into mining inference rules from knowledge bases. To address the task of rule mining over evolving web-scale knowledge bases, we propose a parallel incremental rule mining framework. Our approach is able to efficiently mine rules based on the relational model and apply updates to large knowledge bases; we propose an alternative metric that reduces computation complexity without compromising quality; we apply multiple optimization techniques that reduce runtime by more than 2 orders of magnitude. Experiments show that our approach efficiently scales to web-scale knowledge bases and saves over 90% time compared to the state-of-the-art batch rule mining system. We also apply our optimization techniques to the batch rule mining algorithm, reducing runtime by more than half compared to the state-of-the-art. To the best of our knowledge, our incremental rule mining system is the first that handles updates to web-scale knowledge bases.

Meghan Brennan, Sahil Puri, Tezcan Ozrazgat-Baslanti et al.

Background: Major postoperative complications are associated with increased short and long-term mortality, increased healthcare cost, and adverse long-term consequences. The large amount of data contained in the electronic health record (EHR) creates barriers for physicians to recognize patients most at risk. We hypothesize, if presented in an optimal format, information from data-driven predictive risk algorithms for postoperative complications can improve physician risk assessment. Methods: Prospective, non-randomized, interventional pilot study of twenty perioperative physicians at a quarterly academic medical center. Using 150 clinical cases we compared physicians' risk assessment before and after interaction with MySurgeryRisk, a validated machine-learning algorithm predicting preoperative risk for six major postoperative complications using EHR data. Results: The area under the curve (AUC) of MySurgeryRisk algorithm ranged between 0.73 and 0.85 and was significantly higher than physicians' risk assessments (AUC between 0.47 and 0.69) for all postoperative complications except cardiovascular complications. The AUC for repeated physician's risk assessment improved by 2% to 5% for all complications with the exception of thirty-day mortality. Physicians' risk assessment for acute kidney injury and intensive care unit admission longer than 48 hours significantly improved after knowledge exchange, resulting in net reclassification improvement of 12.4% and 16%, respectively. Conclusions: The validated MySurgeryRisk algorithm predicted postoperative complications with equal or higher accuracy than pilot cohort of physicians using available clinical preoperative data. The interaction with algorithm significantly improved physicians' risk assessment.