arXiv subjects

Huan Sun

Publications and source records attributed to Huan Sun.

At least 109 records · Page 6Linked to original sources

arXiv2022

Towards Transparent Interactive Semantic Parsing via Step-by-Step Correction

Existing studies on semantic parsing focus primarily on mapping a natural-language utterance to a corresponding logical form in one turn. However, because natural language can contain a great deal of ambiguity and variability, this is a difficult challenge. In this work, we investigate an interactive semantic parsing framework that explains the predicted logical form step by step in natural language and enables the user to make corrections through natural-language feedback for individual steps. We focus on question answering over knowledge bases (KBQA) as an instantiation of our framework, aiming to increase the transparency of the parsing process and help the user appropriately trust the final answer. To do so, we construct INSPIRED, a crowdsourced dialogue dataset derived from the ComplexWebQuestions dataset. Our experiments show that the interactive framework with human feedback has the potential to greatly improve overall parse accuracy. Furthermore, we develop a pipeline for dialogue simulation to evaluate our framework w.r.t. a variety of state-of-the-art KBQA models without involving further crowdsourcing effort. The results demonstrate that our interactive semantic parsing framework promises to be effective across such models.

cs.CL↗

arXiv2022

Synthetic Question Value Estimation for Domain Adaptation of Question Answering

Synthesizing QA pairs with a question generator (QG) on the target domain has become a popular approach for domain adaptation of question answering (QA) models. Since synthetic questions are often noisy in practice, existing work adapts scores from a pretrained QA (or QG) model as criteria to select high-quality questions. However, these scores do not directly serve the ultimate goal of improving QA performance on the target domain. In this paper, we introduce a novel idea of training a question value estimator (QVE) that directly estimates the usefulness of synthetic questions for improving the target-domain QA performance. By conducting comprehensive experiments, we show that the synthetic questions selected by QVE can help achieve better target-domain QA performance, in comparison with existing techniques. We additionally show that by using such questions and only around 15% of the human annotations on the target domain, we can achieve comparable performance to the fully-supervised baselines.

cs.CL↗

arXiv2022

DOM-LM: Learning Generalizable Representations for HTML Documents

HTML documents are an important medium for disseminating information on the Web for human consumption. An HTML document presents information in multiple text formats including unstructured text, structured key-value pairs, and tables. Effective representation of these documents is essential for machine understanding to enable a wide range of applications, such as Question Answering, Web Search, and Personalization. Existing work has either represented these documents using visual features extracted by rendering them in a browser, which is typically computationally expensive, or has simply treated them as plain text documents, thereby failing to capture useful information presented in their HTML structure. We argue that the text and HTML structure together convey important semantics of the content and therefore warrant a special treatment for their representation learning. In this paper, we introduce a novel representation learning approach for web pages, dubbed DOM-LM, which addresses the limitations of existing approaches by encoding both text and DOM tree structure with a transformer-based encoder and learning generalizable representations for HTML documents via self-supervised pre-training. We evaluate DOM-LM on a variety of webpage understanding tasks, including Attribute Extraction, Open Information Extraction, and Question Answering. Our extensive experiments show that DOM-LM consistently outperforms all baselines designed for these tasks. In particular, DOM-LM demonstrates better generalization performance both in few-shot and zero-shot settings, making it attractive for making it suitable for real-world application settings with limited labeled data.

cs.CL↗

arXiv2021

TopNet: Learning from Neural Topic Model to Generate Long Stories

Long story generation (LSG) is one of the coveted goals in natural language processing. Different from most text generation tasks, LSG requires to output a long story of rich content based on a much shorter text input, and often suffers from information sparsity. In this paper, we propose \emph{TopNet} to alleviate this problem, by leveraging the recent advances in neural topic modeling to obtain high-quality skeleton words to complement the short input. In particular, instead of directly generating a story, we first learn to map the short text input to a low-dimensional topic distribution (which is pre-assigned by a topic model). Based on this latent topic distribution, we can use the reconstruction decoder of the topic model to sample a sequence of inter-related words as a skeleton for the story. Experiments on two benchmark datasets show that our proposed framework is highly effective in skeleton word selection and significantly outperforms the state-of-the-art models in both automatic evaluation and human evaluation.

cs.LG↗

arXiv2021

CliniQG4QA: Generating Diverse Questions for Domain Adaptation of Clinical Question Answering

Clinical question answering (QA) aims to automatically answer questions from medical professionals based on clinical texts. Studies show that neural QA models trained on one corpus may not generalize well to new clinical texts from a different institute or a different patient group, where large-scale QA pairs are not readily available for model retraining. To address this challenge, we propose a simple yet effective framework, CliniQG4QA, which leverages question generation (QG) to synthesize QA pairs on new clinical contexts and boosts QA models without requiring manual annotations. In order to generate diverse types of questions that are essential for training QA models, we further introduce a seq2seq-based question phrase prediction (QPP) module that can be used together with most existing QG models to diversify the generation. Our comprehensive experiment results show that the QA corpus generated by our framework can improve QA models on the new contexts (up to 8% absolute gain in terms of Exact Match), and that the QPP module plays a crucial role in achieving the gain.

cs.CL↗

arXiv2021

COUGH: A Challenge Dataset and Models for COVID-19 FAQ Retrieval

We present a large, challenging dataset, COUGH, for COVID-19 FAQ retrieval. Similar to a standard FAQ dataset, COUGH consists of three parts: FAQ Bank, Query Bank and Relevance Set. The FAQ Bank contains ~16K FAQ items scraped from 55 credible websites (e.g., CDC and WHO). For evaluation, we introduce Query Bank and Relevance Set, where the former contains 1,236 human-paraphrased queries while the latter contains ~32 human-annotated FAQ items for each query. We analyze COUGH by testing different FAQ retrieval models built on top of BM25 and BERT, among which the best model achieves 48.8 under P@5, indicating a great challenge presented by COUGH and encouraging future research for further improvement. Our COUGH dataset is available at https://github.com/sunlab-osu/covid-faq.

cs.CL↗

arXiv2021

ReasonBERT: Pre-trained to Reason with Distant Supervision

We present ReasonBert, a pre-training method that augments language models with the ability to reason over long-range relations and multiple, possibly hybrid contexts. Unlike existing pre-training methods that only harvest learning signals from local contexts of naturally occurring texts, we propose a generalized notion of distant supervision to automatically connect multiple pieces of text and tables to create pre-training examples that require long-range reasoning. Different types of reasoning are simulated, including intersecting multiple pieces of evidence, bridging from one piece of evidence to another, and detecting unanswerable cases. We conduct a comprehensive evaluation on a variety of extractive question answering datasets ranging from single-hop to multi-hop and from text-only to table-only to hybrid that require various reasoning capabilities and show that ReasonBert achieves remarkable improvement over an array of strong baselines. Few-shot experiments further demonstrate that our pre-training method substantially improves sample efficiency.

cs.CL↗

arXiv2021

Cooperative Reflection Design with Timing Offsets in Distributed Multi-RIS Communications

This letter investigates a wireless communication system deploying distributed reconfigurable intelligent surfaces (RISs). Existing works have assumed that perfect timing synchronization is available among all the cooperative RISs. For practical considerations, we first study cooperative reflection design for multi-RIS-aided communication systems taking into account timing synchronization errors. We aim to minimize the mean-squared error of the recovered data in the presence of timing offsets subject to the unit modulus constraints imposed on the phase shifts of the RISs. In order to handle this sophisticated nonconvex problem, we develop a computationally-efficient algorithm based on the majorization-minimization framework where the RIS phase shift matrices and the timing offset equalizer are respectively developed in a semi-closed form and a closed form. Numerical examples validate the improved performance of our proposed design compared with various schemes.

cs.IT↗

arXiv2021

Differential Privacy for Text Analytics via Natural Text Sanitization

Texts convey sophisticated knowledge. However, texts also convey sensitive information. Despite the success of general-purpose language models and domain-specific mechanisms with differential privacy (DP), existing text sanitization mechanisms still provide low utility, as cursed by the high-dimensional text representation. The companion issue of utilizing sanitized texts for downstream analytics is also under-explored. This paper takes a direct approach to text sanitization. Our insight is to consider both sensitivity and similarity via our new local DP notion. The sanitized texts also contribute to our sanitization-aware pretraining and fine-tuning, enabling privacy-preserving natural language processing over the BERT language model with promising utility. Surprisingly, the high utility does not boost up the success rate of inference attacks.

cs.CL↗

arXiv2021

Learning Structural Edits via Incremental Tree Transformations

While most neural generative models generate outputs in a single pass, the human creative process is usually one of iterative building and refinement. Recent work has proposed models of editing processes, but these mostly focus on editing sequential data and/or only model a single editing pass. In this paper, we present a generic model for incremental editing of structured data (i.e., "structural edits"). Particularly, we focus on tree-structured data, taking abstract syntax trees of computer programs as our canonical example. Our editor learns to iteratively generate tree edits (e.g., deleting or adding a subtree) and applies them to the partially edited data, thereby the entire editing process can be formulated as consecutive, incremental tree transformations. To show the unique benefits of modeling tree edits directly, we further propose a novel edit encoder for learning to represent edits, as well as an imitation learning method that allows the editor to be more robust. We evaluate our proposed editor on two source code edit datasets, where results show that, with the proposed edit encoder, our editor significantly improves accuracy over previous approaches that generate the edited program directly in one pass. Finally, we demonstrate that training our editor to imitate experts and correct its mistakes dynamically can further improve its performance.

cs.LG↗

arXiv2020

TURL: Table Understanding through Representation Learning

Relational tables on the Web store a vast amount of knowledge. Owing to the wealth of such tables, there has been tremendous progress on a variety of tasks in the area of table understanding. However, existing work generally relies on heavily-engineered task-specific features and model architectures. In this paper, we present TURL, a novel framework that introduces the pre-training/fine-tuning paradigm to relational Web tables. During pre-training, our framework learns deep contextualized representations on relational tables in an unsupervised manner. Its universal model design with pre-trained representations can be applied to a wide range of tasks with minimal task-specific fine-tuning. Specifically, we propose a structure-aware Transformer encoder to model the row-column structure of relational tables, and present a new Masked Entity Recovery (MER) objective for pre-training to capture the semantics and knowledge in large-scale unlabeled data. We systematically evaluate TURL with a benchmark consisting of 6 different tasks for table understanding (e.g., relation extraction, cell filling). We show that TURL generalizes well to all tasks and substantially outperforms existing methods in almost all instances.

cs.IR↗

arXiv2020

Adversarial Training for Code Retrieval with Question-Description Relevance Regularization

Code retrieval is a key task aiming to match natural and programming languages. In this work, we propose adversarial learning for code retrieval, that is regularized by question-description relevance. First, we adapt a simple adversarial learning technique to generate difficult code snippets given the input question, which can help the learning of code retrieval that faces bi-modal and data-scarce challenges. Second, we propose to leverage question-description relevance to regularize adversarial learning, such that a generated code snippet should contribute more to the code retrieval training loss, only if its paired natural language description is predicted to be less relevant to the user given question. Experiments on large-scale code retrieval datasets of two programming languages show that our adversarial learning method is able to improve the performance of state-of-the-art models. Moreover, using an additional duplicate question prediction model to regularize adversarial learning further improves the performance, and this is more effective than using the duplicated questions in strong multi-task learning baselines

cs.CL↗

arXiv2020

Learning a Cost-Effective Annotation Policy for Question Answering

State-of-the-art question answering (QA) relies upon large amounts of training data for which labeling is time consuming and thus expensive. For this reason, customizing QA systems is challenging. As a remedy, we propose a novel framework for annotating QA datasets that entails learning a cost-effective annotation policy and a semi-supervised annotation scheme. The latter reduces the human effort: it leverages the underlying QA system to suggest potential candidate annotations. Human annotators then simply provide binary feedback on these candidates. Our system is designed such that past annotations continuously improve the future performance and thus overall annotation cost. To the best of our knowledge, this is the first paper to address the problem of annotating questions with minimal annotation cost. We compare our framework against traditional manual annotations in an extensive set of experiments. We find that our approach can reduce up to 21.1% of the annotation cost.

cs.CL↗

arXiv2020

Joint Power Allocation and User Association Optimization for IRS-Assisted mmWave Systems

Intelligent reflect surface (IRS) is a potential technology to build programmable wireless environment in future communication systems. In this paper, we consider an IRS-assisted multi-base station (multi-BS) multi-user millimeter wave (mmWave) downlink communication system, exploiting IRS to extend mmWave signal coverage to blind spots. Considering the impact of IRS on user association in multi-BS mmWave systems, we formulate a sum rate maximization problem by jointly optimizing passive beamforming at IRS, power allocation and user association. This leads to an intractable nonconvex problem, for which to tackle we propose a computationally affordable iterative algorithm, capitalizing on alternating optimization, sequential fractional programming (SFP) and forward-reverse auction (FRA). In particular, passive beamforming at IRS is optimized by utilizing the SFP method, power allocation is solved through means of standard convex optimization method, and user association is handled by the network optimization based FRA algorithm. Simulation results demonstrate that the proposed algorithm can achieve significant performance gains, e.g., it can provide up to 175% higher sum rate compared with the benchmark and 140% higher energy efficiency compared with amplify-and-forward relay.

cs.IT↗

arXiv2020

Joint Passive Beamforming and User Association Optimization for IRS-assisted mmWave Systems

In this paper, we investigate an intelligent reflect surface (IRS) assisted multi-user millimeter wave (mmWave) downlink communication system, exploiting IRS to alleviate the blockage effect and enhance the performance of the mmWave system. Considering the impact of IRS on user association, we formulate a sum rate maximization problem by jointly optimizing the passive beamforming at IRS and user association, which is an intractable non-convex problem. Then an alternating optimization algorithm is proposed to solve the problem efficiently. In the proposed algorithm, passive beamforming at IRS is optimized by utilizing the fractional programming method and user association is solved through the network optimization based auction algorithm. We provide numerical comparisons between the proposed algorithm and different reference algorithms. Simulation results demonstrate that the proposed algorithm can achieve significant gains in the sum rate of all users.

eess.SP↗

arXiv2020

An Imitation Game for Learning Semantic Parsers from User Interaction

Despite the widely successful applications, bootstrapping and fine-tuning semantic parsers are still a tedious process with challenges such as costly data annotation and privacy risks. In this paper, we suggest an alternative, human-in-the-loop methodology for learning semantic parsers directly from users. A semantic parser should be introspective of its uncertainties and prompt for user demonstration when uncertain. In doing so it also gets to imitate the user behavior and continue improving itself autonomously with the hope that eventually it may become as good as the user in interpreting their questions. To combat the sparsity of demonstration, we propose a novel annotation-efficient imitation learning algorithm, which iteratively collects new datasets by mixing demonstrated states and confident predictions and re-trains the semantic parser in a Dataset Aggregation fashion (Ross et al., 2011). We provide a theoretical analysis of its cost bound and also empirically demonstrate its promising performance on the text-to-SQL problem. Code will be available at https://github.com/sunlab-osu/MISP.

cs.CL↗

arXiv2020

Cascaded Channel Estimation for IRS-assisted Mmwave Multi-antenna with Quantized Beamforming

In this letter, we optimize the channel estimator of the cascaded channel in an intelligent reflecting surface (IRS)-assisted millimeter wave (mmWave) multi-antenna system. In this system, the receiver is equipped with a hybrid architecture adopting quantized beamforming. Different from traditional multiple-input multiple-output (MIMO) systems, the design of channel estimation is challenging since the IRS is usually a passive array with limited signal processing capability. We derive the optimized channel estimator in a closed form by reformulating the problem of cascaded channel estimation in this system, leveraging the typical mean-squared error (MSE) criterion. Considering the presence of possible channel sparsity in mmWave channels, we generalize the proposed method by exploiting the channel sparsity for further performance enhancement and computational complexity reduction. Simulation results verify that the proposed estimator significantly outperforms the existing ones.

cs.IT↗

arXiv2020

Energy Efficiency Optimization in IRS-Enhanced mmWave Systems with Lens Antenna Array

In millimeter wave (mmWave) systems, the advanced lens antenna array can effectively reduce the radio frequency chains cost. However, the mmWave signal is still vulnerable to blocking obstacles and suffers from severe path loss. To address this problem, we propose an intelligent reflect surface (IRS) enhanced multi-user mmWave communication system with lens antenna array. Moreover, we attempt to optimize energy efficiency in the proposed system. An energy efficiency maximization problem is formulated where the transmit beamforming at base station and the reflect beamforming at IRSs are jointly considered. To solve this non-convex problem, we propose an algorithm based on the alternating optimization technique. In the proposed algorithm, the transmit beamforming is handled by the sequential convex approximation method and the reflect beamforming is optimized based on the quadratic transform method. Our simulation results show that the proposed algorithm can achieve significant energy efficiency improvement under various scenarios.

eess.SP↗