Jeff Dalton

Mirelle Bueno, Carlos Gemmell, Jeffrey Dalton et al.

The ability to extrapolate, i.e., to make predictions on sequences that are longer than those presented as training examples, is a challenging problem for current deep learning models. Recent work shows that this limitation persists in state-of-the-art Transformer-based models. Most solutions to this problem use specific architectures or training methods that do not generalize to other tasks. We demonstrate that large language models can succeed in extrapolation without modifying their architecture or training procedure. Our experimental results show that generating step-by-step rationales and introducing marker tokens are both required for effective extrapolation. First, we induce a language model to produce step-by-step rationales before outputting the answer to effectively communicate the task to the model. However, as sequences become longer, we find that current models struggle to keep track of token positions. To address this issue, we interleave output tokens with markup tokens that act as explicit positional and counting symbols. Our findings show how these two complementary approaches enable remarkable sequence extrapolation and highlight a limitation of current architectures to effectively generalize without explicit surface form guidance. Code available at https://github.com/MirelleB/induced-rationales-markup-tokens

Francesco Dalla Serra, Chaoyang Wang, Fani Deligianni et al.

Radiology reports are detailed text descriptions of the content of medical scans. Each report describes the presence/absence and location of relevant clinical findings, commonly including comparison with prior exams of the same patient to describe how they evolved. Radiology reporting is a time-consuming process, and scan results are often subject to delays. One strategy to speed up reporting is to integrate automated reporting systems, however clinical deployment requires high accuracy and interpretability. Previous approaches to automated radiology reporting generally do not provide the prior study as input, precluding comparison which is required for clinical accuracy in some types of scans, and offer only unreliable methods of interpretability. Therefore, leveraging an existing visual input format of anatomical tokens, we introduce two novel aspects: (1) longitudinal representation learning -- we input the prior scan as an additional input, proposing a method to align, concatenate and fuse the current and prior visual information into a joint longitudinal representation which can be provided to the multimodal report generation model; (2) sentence-anatomy dropout -- a training strategy for controllability in which the report generator model is trained to predict only sentences from the original report which correspond to the subset of anatomical regions given as input. We show through in-depth experiments on the MIMIC-CXR dataset how the proposed approach achieves state-of-the-art results while enabling anatomy-wise controllable report generation.

Francesco Dalla Serra, Chaoyang Wang, Fani Deligianni et al.

The task of radiology reporting comprises describing and interpreting the medical findings in radiographic images, including description of their location and appearance. Automated approaches to radiology reporting require the image to be encoded into a suitable token representation for input to the language model. Previous methods commonly use convolutional neural networks to encode an image into a series of image-level feature map representations. However, the generated reports often exhibit realistic style but imperfect accuracy. Inspired by recent works for image captioning in the general domain in which each visual token corresponds to an object detected in an image, we investigate whether using local tokens corresponding to anatomical structures can improve the quality of the generated reports. We introduce a novel adaptation of Faster R-CNN in which finding detection is performed for the candidate bounding boxes extracted during anatomical structure localisation. We use the resulting bounding box feature representations as our set of finding-aware anatomical tokens. This encourages the extracted anatomical tokens to be informative about the findings they contain (required for the final task of radiology reporting). Evaluating on the MIMIC-CXR dataset of chest X-Ray images, we show that task-aware anatomical tokens give state-of-the-art performance when integrated into an automated reporting pipeline, yielding generated reports with improved clinical accuracy.

Carlos Gemmell, Jeffrey Dalton

Tabular question answering (TQA) presents a challenging setting for neural systems by requiring joint reasoning of natural language with large amounts of semi-structured data. Unlike humans who use programmatic tools like filters to transform data before processing, language models in TQA process tables directly, resulting in information loss as table size increases. In this paper we propose ToolWriter to generate query specific programs and detect when to apply them to transform tables and align them with the TQA model's capabilities. Focusing ToolWriter to generate row-filtering tools improves the state-of-the-art for WikiTableQuestions and WikiSQL with the most performance gained on long tables. By investigating headroom, our work highlights the broader potential for programmatic tools combined with neural components to manipulate large amounts of structured data.

Carlos Gemmell, Iain Mackie, Paul Owoicho et al.

GRILLBot is the winning system in the 2022 Alexa Prize TaskBot Challenge, moving towards the next generation of multimodal task assistants. It is a voice assistant to guide users through complex real-world tasks in the domains of cooking and home improvement. These are long-running and complex tasks that require flexible adjustment and adaptation. The demo highlights the core aspects, including a novel Neural Decision Parser for contextualized semantic parsing, a new "TaskGraph" state representation that supports conditional execution, knowledge-grounded chit-chat, and automatic enrichment of tasks with images and videos.

Iain Mackie, Jeffrey Dalton

Across the financial domain, researchers answer complex questions by extensively "searching" for relevant information to generate long-form reports. This workshop paper discusses automating the construction of query-specific document and entity knowledge graphs (KGs) for complex research topics. We focus on the CODEC dataset, where domain experts (1) create challenging questions, (2) construct long natural language narratives, and (3) iteratively search and assess the relevance of documents and entities. For the construction of query-specific KGs, we show that state-of-the-art ranking systems have headroom for improvement, with specific failings due to a lack of context or explicit knowledge representation. We demonstrate that entity and document relevance are positively correlated, and that entity-based query feedback improves document ranking effectiveness. Furthermore, we construct query-specific KGs using retrieval and evaluate using CODEC's "ground-truth graphs", showing the precision and recall trade-offs. Lastly, we point to future work, including adaptive KG retrieval algorithms and GNN-based weighting methods, while highlighting key challenges such as high-quality data, information extraction recall, and the size and sparsity of complex topic graphs.

Elena Soare, Iain Mackie, Jeffrey Dalton

Current SQL generators based on pre-trained language models struggle to answer complex questions requiring domain context or understanding fine-grained table structure. Humans would deal with these unknowns by reasoning over the documentation of the tables. Based on this hypothesis, we propose DocuT5, which uses off-the-shelf language model architecture and injects knowledge from external `documentation' to improve domain generalization. We perform experiments on the Spider family of datasets that contain complex questions that are cross-domain and multi-table. Specifically, we develop a new text-to-SQL failure taxonomy and find that 19.6% of errors are due to foreign key mistakes, and 49.2% are due to a lack of domain knowledge. We proposed DocuT5, a method that captures knowledge from (1) table structure context of foreign keys and (2) domain knowledge through contextualizing tables and columns. Both types of knowledge improve over state-of-the-art T5 with constrained decoding on Spider, and domain knowledge produces state-of-the-art comparable effectiveness on Spider-DK and Spider-SYN datasets.

Ekaterina Taktasheva, Jeff Dalton

Recent advancements in tool-augmented large language models have enabled them to interact with external tools, enhancing their ability to perform complex user tasks. However, existing approaches overlook the role of personalisation in guiding tool use. This work investigates how user preferences can be effectively integrated into goal-oriented dialogue agents. Through extensive analysis, we identify key weaknesses in the ability of LLMs to personalise tool use. To this end, we introduce TAPS, a novel solution that enhances personalised tool use by leveraging a structured tagging tool and an uncertainty-based tool detector. TAPS significantly improves the ability of LLMs to incorporate user preferences, achieving the new state-of-the-art for open source models on the NLSI task.

Mohammad Aliannejadi, Zahra Abbasiantaeb, Shubham Chatterjee et al.

Conversational information seeking has evolved rapidly in the last few years with the development of Large Language Models (LLMs), providing the basis for interpreting and responding in a naturalistic manner to user requests. The extended TREC Interactive Knowledge Assistance Track (iKAT) collection aims to enable researchers to test and evaluate their Conversational Search Agents (CSA). The collection contains a set of 36 personalized dialogues over 20 different topics each coupled with a Personal Text Knowledge Base (PTKB) that defines the bespoke user personas. A total of 344 turns with approximately 26,000 passages are provided as assessments on relevance, as well as additional assessments on generated responses over four key dimensions: relevance, completeness, groundedness, and naturalness. The collection challenges CSA to efficiently navigate diverse personal contexts, elicit pertinent persona information, and employ context for relevant conversations. The integration of a PTKB and the emphasis on decisional search tasks contribute to the uniqueness of this test collection, making it an essential benchmark for advancing research in conversational and interactive knowledge assistants.

Shubham Chatterjee, Iain Mackie, Jeff Dalton

While entity-oriented neural IR models have advanced significantly, they often overlook a key nuance: the varying degrees of influence individual entities within a document have on its overall relevance. Addressing this gap, we present DREQ, an entity-oriented dense document re-ranking model. Uniquely, we emphasize the query-relevant entities within a document's representation while simultaneously attenuating the less relevant ones, thus obtaining a query-specific entity-centric document representation. We then combine this entity-centric document representation with the text-centric representation of the document to obtain a "hybrid" representation of the document. We learn a relevance score for the document using this hybrid representation. Using four large-scale benchmarks, we show that DREQ outperforms state-of-the-art neural and non-neural re-ranking methods, highlighting the effectiveness of our entity-oriented representation approach.

Shubham Chatterjee, Jeff Dalton

Neural IR has advanced through two distinct paths: entity-oriented approaches leveraging knowledge graphs and multi-vector models capturing fine-grained semantics. We introduce QDER, a neural re-ranking model that unifies these approaches by integrating knowledge graph semantics into a multi-vector model. QDER's key innovation lies in its modeling of query-document relationships: rather than computing similarity scores on aggregated embeddings, we maintain individual token and entity representations throughout the ranking process, performing aggregation only at the final scoring stage - an approach we call "late aggregation." We first transform these fine-grained representations through learned attention patterns, then apply carefully chosen mathematical operations for precise matches. Experiments across five standard benchmarks show that QDER achieves significant performance gains, with improvements of 36% in nDCG@20 over the strongest baseline on TREC Robust 2004 and similar improvements on other datasets. QDER particularly excels on difficult queries, achieving an nDCG@20 of 0.70 where traditional approaches fail completely (nDCG@20 = 0.0), setting a foundation for future work in entity-aware retrieval.

Hamed Zamani, Johanne R. Trippas, Jeff Dalton et al.

Conversational information seeking (CIS) is concerned with a sequence of interactions between one or more users and an information system. Interactions in CIS are primarily based on natural language dialogue, while they may include other types of interactions, such as click, touch, and body gestures. This monograph provides a thorough overview of CIS definitions, applications, interactions, interfaces, design, implementation, and evaluation. This monograph views CIS applications as including conversational search, conversational question answering, and conversational recommendation. Our aim is to provide an overview of past research related to CIS, introduce the current state-of-the-art in CIS, highlight the challenges still being faced in the community. and suggest future directions.

Yola Jones, Fani Deligianni, Jeff Dalton

Cardiovascular disease is a large worldwide healthcare issue; symptoms often present suddenly with minimal warning. The electrocardiogram (ECG) is a fast, simple and reliable method of evaluating the health of the heart, by measuring electrical activity recorded through electrodes placed on the skin. ECGs often need to be analyzed by a cardiologist, taking time which could be spent on improving patient care and outcomes. Because of this, automatic ECG classification systems using machine learning have been proposed, which can learn complex interactions between ECG features and use this to detect abnormalities. However, algorithms built for this purpose often fail to generalize well to unseen data, reporting initially impressive results which drop dramatically when applied to new environments. Additionally, machine learning algorithms suffer a "black-box" issue, in which it is difficult to determine how a decision has been made. This is vital for applications in healthcare, as clinicians need to be able to verify the process of evaluation in order to trust the algorithm. This paper proposes a method for visualizing model decisions across each class in the MIT-BIH arrhythmia dataset, using adapted saliency maps averaged across complete classes to determine what patterns are being learned. We do this by building two algorithms based on state-of-the-art models. This paper highlights how these maps can be used to find problems in the model which could be affecting generalizability and model performance. Comparing saliency maps across complete classes gives an overall impression of confounding variables or other biases in the model, unlike what would be highlighted when comparing saliency maps on an ECG-by-ECG basis.

Mohammad Aliannejadi, Julia Kiseleva, Aleksandr Chuklin et al.

Enabling open-domain dialogue systems to ask clarifying questions when appropriate is an important direction for improving the quality of the system response. Namely, for cases when a user request is not specific enough for a conversation system to provide an answer right away, it is desirable to ask a clarifying question to increase the chances of retrieving a satisfying answer. To address the problem of 'asking clarifying questions in open-domain dialogues': (1) we collect and release a new dataset focused on open-domain single- and multi-turn conversations, (2) we benchmark several state-of-the-art neural baselines, and (3) we propose a pipeline consisting of offline and online steps for evaluating the quality of clarifying questions in various dialogues. These contributions are suitable as a foundation for further research.

Shahrzad Naseri, Jeffrey Dalton, Andrew Yates et al.

In this work we leverage recent advances in context-sensitive language models to improve the task of query expansion. Contextualized word representation models, such as ELMo and BERT, are rapidly replacing static embedding models. We propose a new model, Contextualized Embeddings for Query Expansion (CEQE), that utilizes query-focused contextualized embedding vectors. We study the behavior of contextual representations generated for query expansion in ad-hoc document retrieval. We conduct our experiments on probabilistic retrieval models as well as in combination with neural ranking models. We evaluate CEQE on two standard TREC collections: Robust and Deep Learning. We find that CEQE outperforms static embedding-based expansion methods on multiple collections (by up to 18% on Robust and 31% on Deep Learning on average precision) and also improves over proven probabilistic pseudo-relevance feedback (PRF) models. We further find that multiple passes of expansion and reranking result in continued gains in effectiveness with CEQE-based approaches outperforming other approaches. The final model incorporating neural and CEQE-based expansion score achieves gains of up to 5% in P@20 and 2% in AP on Robust over the state-of-the-art transformer-based re-ranking model, Birch.

Mohammad Aliannejadi, Julia Kiseleva, Aleksandr Chuklin et al.

This document presents a detailed description of the challenge on clarifying questions for dialogue systems (ClariQ). The challenge is organized as part of the Conversational AI challenge series (ConvAI3) at Search Oriented Conversational AI (SCAI) EMNLP workshop in 2020. The main aim of the conversational systems is to return an appropriate answer in response to the user requests. However, some user requests might be ambiguous. In IR settings such a situation is handled mainly thought the diversification of the search result page. It is however much more challenging in dialogue settings with limited bandwidth. Therefore, in this challenge, we provide a common evaluation framework to evaluate mixed-initiative conversations. Participants are asked to rank clarifying questions in an information-seeking conversations. The challenge is organized in two stages where in Stage 1 we evaluate the submissions in an offline setting and single-turn conversations. Top participants of Stage 1 get the chance to have their model tested by human annotators.

Carlos Gemmell, Federico Rossetto, Jeffrey Dalton

Tools capable of automatic code generation have the potential to augment programmer's capabilities. While straightforward code retrieval is incorporated into many IDEs, an emerging area is explicit code generation. Code generation is currently approached as a Machine Translation task, with Recurrent Neural Network (RNN) based encoder-decoder architectures trained on code-description pairs. In this work we introduce and study modern Transformer architectures for this task. We further propose a new model called the Relevance Transformer that incorporates external knowledge using pseudo-relevance feedback. The Relevance Transformer biases the decoding process to be similar to existing retrieved code while enforcing diversity. We perform experiments on multiple standard benchmark datasets for code generation including Django, Hearthstone, and CoNaLa. The results show improvements over state-of-the-art methods based on BLEU evaluation. The Relevance Transformer model shows the potential of Transformer-based architectures for code generation and introduces a method of incorporating pseudo-relevance feedback during inference.

Jeffrey Dalton, Chenyan Xiong, Jamie Callan

The Conversational Assistance Track (CAsT) is a new track for TREC 2019 to facilitate Conversational Information Seeking (CIS) research and to create a large-scale reusable test collection for conversational search systems. The document corpus is 38,426,252 passages from the TREC Complex Answer Retrieval (CAR) and Microsoft MAchine Reading COmprehension (MARCO) datasets. Eighty information seeking dialogues (30 train, 50 test) are an average of 9 to 10 questions long. Relevance assessments are provided for 30 training topics and 20 test topics. This year 21 groups submitted a total of 65 runs using varying methods for conversational query understanding and ranking. Methods include traditional retrieval based methods, feature based learning-to-rank, neural models, and knowledge enhanced methods. A common theme through the runs is the use of BERT-based neural reranking methods. Leading methods also employed document expansion, conversational query expansion, and generative language models for conversational query rewriting (GPT-2). The results show a gap between automatic systems and those using the manually resolved utterances, with a 35% relative improvement of manual rewrites over the best automatic system.

Laura Dietz, Bhaskar Mitra, Jeremy Pickens et al.

The vision of HIPstIR is that early stage information retrieval (IR) researchers get together to develop a future for non-mainstream ideas and research agendas in IR. The first iteration of this vision materialized in the form of a three day workshop in Portsmouth, New Hampshire attended by 24 researchers across academia and industry. Attendees pre-submitted one or more topics that they want to pitch at the meeting. Then over the three days during the workshop, we self-organized into groups and worked on six specific proposals of common interest. In this report, we present an overview of the workshop and brief summaries of the six proposals that resulted from the workshop.

Gilad Mishne, Jeff Dalton, Zhenghua Li et al.

We present the architecture behind Twitter's real-time related query suggestion and spelling correction service. Although these tasks have received much attention in the web search literature, the Twitter context introduces a real-time "twist": after significant breaking news events, we aim to provide relevant results within minutes. This paper provides a case study illustrating the challenges of real-time data processing in the era of "big data". We tell the story of how our system was built twice: our first implementation was built on a typical Hadoop-based analytics stack, but was later replaced because it did not meet the latency requirements necessary to generate meaningful real-time results. The second implementation, which is the system deployed in production, is a custom in-memory processing engine specifically designed for the task. This experience taught us that the current typical usage of Hadoop as a "big data" platform, while great for experimentation, is not well suited to low-latency processing, and points the way to future work on data analytics platforms that can handle "big" as well as "fast" data.