Xiangjue Dong, Cong Wang, Maria Teleki, Millennium Bismay, Ruihong Huang, James Caverlee. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Media outlets are becoming more partisan and polarized nowadays.Most previous work focused on detecting media bias.In this paper, we aim to mitigate media bias by generating a neutralized summary given multiple articles presenting different ideological views.Motivated by the critical role of events and event relations in media bias detection, we propose to increase awareness of bias in LLMs via multi-document events reasoning and use a multi-document event relation graph to guide the summarization process.This graph contains rich event information useful to reveal bias: four common types of in-doc event relations to reflect content framing bias, cross-doc event coreference relation to reveal content selection bias, and event-level moral opinions to highlight opinionated framing bias.We further develop two strategies to incorporate the multi-document event relation graph for neutralized summarization.Firstly, we convert a graph into natural language descriptions and feed the textualized graph into LLMs as a part of a hard text prompt.Secondly, we encode the graph with graph attention network and insert the graph embedding into LLMs as a soft prompt.Both automatic evaluation and human evaluation confirm that our approach effectively mitigates both lexical and informational media bias, and meanwhile improves content preservation 1 .

Detecting deception in an increasingly digital world is both a critical and challenging task.In this study, we present a comprehensive evaluation of the automated deception detection capabilities of Large Language Models (LLMs) and Large Multimodal Models (LMMs) across diverse domains.We assess the performance of both open-source and proprietary LLMs on three distinct datasets-real-life trial interviews (RLTD), instructed deception in interpersonal scenarios (MU3D), and deceptive reviews (OpSpam).We systematically analyze the effectiveness of different experimental setups for deception detection, including zeroshot and few-shot approaches with random or similarity-based in-context example selection.Our findings indicate that fine-tuned LLMs achieve state-of-the-art performance on textual deception detection, whereas LMMs struggle to fully leverage multimodal cues, particularly in real-world settings.Additionally, we analyze the impact of auxiliary features, such as non-verbal gestures, video summaries, and evaluate the effectiveness of different prompting strategies, such as direct label generation and post-hoc reasoning generation.Experiments unfold that reasoning-based predictions do not consistently improve performance over direct classification, contrary to the expectations.

Recent advancements in NLP, largely driven by Large Language Models (LLMs), have significantly improved performance on an array of tasks.However, Dialogue Act (DA) classification remains challenging, particularly in the fine-grained 50-class, multiparty setting.This paper investigates the root causes of LLMs' poor performance in DA classification through a linguistically motivated analysis.We identify three key pre-tasks essential for accurate DA prediction: Turn Management, Communicative Function Identification, and Dialogue Structure Prediction.Our experiments reveal that LLMs struggle with these fundamental tasks, often failing to outperform simple rulebased baselines.Additionally, we establish a strong empirical correlation between errors in these pre-tasks and DA classification failures.A human study further highlights the significant gap between LLM and human-level dialogue understanding.These findings indicate that LLMs' shortcomings in dialogue comprehension hinder their ability to accurately predict DAs, highlighting the need for improved dialogue-aware training approaches. InstructionYou are an intelligent annotator capable of classifying the intention behind each speaker's utterance.You will be provided with a list of possible Dialogue Acts and their definitions.You will be given an utterance surrounded by '#'.Your task is to predict the correct label for that utterance.You will also be given a snapshot of the conversation to provide context for your prediction.Return the answer in the format: 'label:predicted label'.The Dialogue Acts and their definitions are as follows: Statement: General statements.Accept: A short utterance indicating acceptance of a previous speaker's statement.Disruption: Indecipherable or disrupted speech.Defending/Explanation: The speaker defends their opinion or provides an explanation.

Localizing unusual activities in videos, such as abnormal behaviors or traffic incidents, holds practical significance. However, pretrained foundation models struggle with localizing diverse unusual events likely because of their insufficient representation in the models' pretraining datasets. To explore foundation models' capability in localizing unusual activities, we introduce UALBench, a comprehensive benchmark for unusual activity localization, featuring three video datasets (UAG-OOPS, UAG-SSBD, and UAG-FunQA), and an instruction-tuning dataset (OOPS-UAG-Instruct), to improve model capabilities. We also introduce a new metric, R@1, TD ≤ p, as an auxiliary metric to reasonably consider detections as true positive if their starting and ending timestamps are within a threshold. On UAL-Bench, we evaluate three approaches: Video-Language Models (VidLLMs), instruction-tuned Vid-LLMs, and a novel integration of Vision-Language Models and Large Language Models (VLM-LLM). Our results show the VLM-LLM approach excels in localizing short-span unusual events and predicting their onset (start time) more accurately than VidLLMs. Our findings highlight the challenges posed by longduration videos, particularly in autism diagnosis scenarios, and the need for further advancements in localization techniques. Our work not only provides a benchmark for unusual activity localization but also outlines the key challenges for existing foundation models, suggesting future research directions on this important task.

Within the domain of intelligent education, also known as smart education, the recommender system propelled by deep learning strives to attain exemplary model performance. However, deep learning models invariably involve the processing of voluminous user privacy data during training phases, and they subjugate themselves to substantial risks of privacy breaches concerning both students and educators. Traditional approaches involving retraining the entire dataset and classical privacy protection methods such as differential privacy and homomorphic encryption struggle to balance model performance and training time expenditure. This presents significant challenges for individuals and enterprises in managing privacy concerns. While balancing personal and corporate interests in privacy protection, Machine Unlearning reveals its potential as a productive strategy to navigate these challenges. This study compares the time cost and performance of model retraining using Machine Unlearning with those of retraining using conventional approaches. The experiment results show that the use of Machine Unlearning algorithms not only effectively protects privacy but also significantly reduces the time required for model retraining. Furthermore, the performance of models employing Machine Unlearning is essentially congruent with that of retraining a model with an entire dataset.

Logical fallacy uses invalid or faulty reasoning in the construction of a statement.Despite the prevalence and harmfulness of logical fallacies, detecting and classifying logical fallacies still remains a challenging task.We observe that logical fallacies often use connective words to indicate an intended logical relation between two arguments, while the argument semantics does not actually support the logical relation.Inspired by this observation, we propose to build a logical structure tree to explicitly represent and track the hierarchical logic flow among relation connectives and their arguments in a statement.Specifically, this logical structure tree is constructed in an unsupervised manner guided by the constituency tree and a taxonomy of connectives for ten common logical relations, with relation connectives as non-terminal nodes and textual arguments as terminal nodes, and the latter are mostly elementary discourse units.We further develop two strategies to incorporate the logical structure tree into LLMs for fallacy reasoning.Firstly, we transform the tree into natural language descriptions and feed the textualized tree into LLMs as a part of the hard text prompt.Secondly, we derive a relation-aware tree embedding and insert the tree embedding into LLMs as a soft prompt.Experiments on benchmark datasets demonstrate that our approach based on logical structure tree significantly improves precision and recall for both fallacy detection and fallacy classification 1 .

Jiangshu Du, Yibo Wang, Wenting Zhao, Zhongfen Deng, Shuaiqi Liu, Renze Lou, Henry Peng Zou, Pranav Narayanan Venkit, Nan Zhang, Mukund Srinath, Haoran Ranran Zhang, Vipul Gupta, Yinghui Li, Tao Li, Fei Wang, Qin Liu, Tianlin Liu, Pengzhi Gao, Congying Xia, Chen Xing, Cheng Jiayang, Zhaowei Wang, Ying Su, Raj Sanjay Shah, Ruohao Guo, Jing Gu, Haoran Li, Kangda Wei, Zihao Wang, Lu Cheng, Surangika Ranathunga, Meng Fang, Jie Fu, Fei Liu, Ruihong Huang, Eduardo Blanco, Yixin Cao, Rui Zhang, Philip S. Yu, Wenpeng Yin. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. 2024.

Yuanyuan Lei, Ruihong Huang. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024.

Yuanyuan Lei, Md Messal Monem Miah, Ayesha Qamar, Sai Ramana Reddy, Jonathan Tong, Haotian Xu, Ruihong Huang. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024.

Yuanyuan Lei, Kaiqiang Song, Sangwoo Cho, Xiaoyang Wang, Ruihong Huang, Dong Yu. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024.

Vector representations of natural language are ubiquitous in search applications.Recently, various methods based on contrastive learning have been proposed to learn textual representations from unlabelled data; by maximizing alignment between minimally-perturbed embeddings of the same text, and encouraging a uniform distribution of embeddings across a broader corpus.Differently, we propose maximizing alignment between texts and a composition of their phrasal constituents.We consider several realizations of this objective and elaborate the impact on representations in each case.Experimental results on semantic textual similarity tasks show improvements over baselines that are comparable with state-of-the-art approaches.Moreover, this work is the first to do so without incurring costs in auxiliary training objectives or additional network parameters. 1

Propaganda is a form of deceptive narratives that instigate or mislead the public, usually with a political purpose.In this paper, we aim to identify propaganda in political news at two fine-grained levels: sentence-level and tokenlevel.We observe that propaganda content is more likely to be embedded in sentences that attribute causality or assert contrast to nearby sentences, as well as seen in opinionated evaluation, speculation and discussions of future expectation.Hence, we propose to incorporate both local and global discourse structures for propaganda discovery and construct two teacher models for identifying PDTB-style discourse relations between nearby sentences and common discourse roles of sentences in a news article respectively.We further devise two methods to incorporate the two types of discourse structures for propaganda identification by either using teacher predicted probabilities as additional features or soliciting guidance in a knowledge distillation framework.Experiments on the benchmark dataset demonstrate that leveraging guidance from discourse structures can significantly improve both precision and recall of propaganda content identification. 1

Public opinion is shaped by the information news media provide, and that information in turn may be shaped by the ideological preferences of media outlets.But while much attention has been devoted to media bias via overt ideological language or topic selection, a more unobtrusive way in which the media shape opinion is via the strategic inclusion or omission of partisan events that may support one side or the other.We develop a latent variable-based framework to predict the ideology of news articles by comparing multiple articles on the same story and identifying partisan events whose inclusion or omission reveals ideology.Our experiments first validate the existence of partisan event selection, and then show that article alignment and cross-document comparison detect partisan events and article ideology better than competitive baselines.Our results reveal the high-level form of media bias, which is present even among mainstream media with strong norms of objectivity and nonpartisanship.

To address the variety and veracity issues of big data, data dependencies have been extended as data quality rules to adapt to various data types, ranging from (1) categorical data with equality relationships to (2) heterogeneous data with similarity relationships, and (3) numerical data with order relationships. In this survey, we briefly review the recent proposals on data dependencies categorized into the aforesaid types of data. In addition to (a) the concepts of these data dependency notations, we investigate (b) the extension relationships between data dependencies. It forms a family tree of extensions, mostly rooted in FDs. Moreover, we summarize (c) the discovery of dependencies from data, and (d) the applications of the extended data dependencies. Finally, we conclude with several directions of future studies on the emerging data.

Ming Li, Ruihong Huang. Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers). 2023.

Language models pretrained on large collections of tabular data have demonstrated their effectiveness in several downstream tasks. However, many of these models do not take into account the row/column permutation invariances, hierarchical structure, etc. that exist in tabular data. To alleviate these limitations, we propose HyTrel, a tabular language model, that captures the permutation invariances and three more structural properties of tabular data by using hypergraphs--where the table cells make up the nodes and the cells occurring jointly together in each row, column, and the entire table are used to form three different types of hyperedges. We show that HyTrel is maximally invariant under certain conditions for tabular data, i.e., two tables obtain the same representations via HyTrel iff the two tables are identical up to permutation. Our empirical results demonstrate that HyTrel consistently outperforms other competitive baselines on four downstream tasks with minimal pretraining, illustrating the advantages of incorporating inductive biases associated with tabular data into the representations. Finally, our qualitative analyses showcase that HyTrel can assimilate the table structure to generate robust representations for the cells, rows, columns, and the entire table.

Outsourced cloud computing can be considered as an effective way to overcome the data island among users and relieve the pressure of limited resources. However, due to the concerns about trust in cloud servers, outsourcing the users’ data and model training task has considerable privacy disclosure risks. This article presents a PriKPM scheme by using additive secret sharing (ASS), so as to implement the privacy-preserving${k}$-prototype clustering for mixed data (i.e., including numerical and categorical attributes). In PriKPM, data samples are randomly split into two shares and delivered offline to two collaborative servers. We design a secure initialization method for determining the location and number of cluster centers. Then, both servers securely calculate the mixed distance between samples and cluster centers, and execute the samples partion and cluster updating operations. An efficient and secure comparison protocol is developed to offer flexibly the “less than or equal” and “equal” functions during the entire clustering process. Furthermore, theoretical analysis proves the effectiveness and security of PriKPM. Sufficient experiments demonstrate that PriKPM is computationally more efficient than existing secure clustering works. PriKPM can achieve the approximate accuracy of the plaintext${k}$-prototype clustering scheme.

Event data are often dirty owing to various recording conventions or simply system errors. These errors may cause serious damage to real applications, such as inaccurate provenance answers, poor profiling results, or concealing interesting patterns from event data. Cleaning dirty event data is strongly demanded. While existing event data cleaning techniques view event logs as sequences, structural information does exist among events, such as the task passing relationships between staffs in workflow or the invocation relationships among different micro-services in monitoring application performance. We argue that such structural information enhances not only the accuracy of repairing inconsistent events but also the computation efficiency. It is notable that both the structure and the names (labeling) of events could be inconsistent. In real applications, while an unsound structure is not repaired automatically (which requires manual effort from business actors to handle the structure error), it is highly desirable to repair the inconsistent event names introduced by recording mistakes. In this article, we first prove that the inconsistent label repairing problem is NP-complete. Then, we propose a graph repair approach for (1) detecting unsound structures, and (2) repairing inconsistent event names. Efficient pruning techniques together with two heuristic solutions are also presented. Extensive experiments over real and synthetic datasets demonstrate both the effectiveness and efficiency of our proposal.

As polarization continues to rise among both the public and the news media, increasing attention has been devoted to detecting media bias.Most recent work in the NLP community, however, identify bias at the level of individual articles.However, each article itself comprises multiple sentences, which vary in their ideological bias.In this paper, we aim to identify sentences within an article that can illuminate and explain the overall bias of the entire article.We show that understanding the discourse role of a sentence in telling a news story, as well as its relation with nearby sentences, can reveal the ideological leanings of an author even when the sentence itself appears merely neutral.In particular, we consider using a functional news discourse structure and PDTB discourse relations to inform bias sentence identification, and distill the auxiliary knowledge from the two types of discourse structure into our bias sentence identification system.Experimental results on benchmark datasets show that incorporating both the global functional discourse structure and local rhetorical discourse relations can effectively increase the recall of bias sentence identification by 8.27% -8.62%, as well as increase the precision by 2.82% -3.48% 1 .

General domain Named Entity Recognition (NER) datasets like CoNLL-2003 mostly annotate coarse-grained location entities such as a country or a city.But many applications require identifying fne-grained locations from texts and mapping them precisely to geographic sites, e.g., a crossroad, an apartment building, or a grocery store.In this paper, we introduce a new dataset HarveyNER with fne-grained locations annotated in tweets.This dataset presents unique challenges and characterizes many complex and long location mentions in informal descriptions.We built strong baseline models using Curriculum Learning and experimented with different heuristic curricula to better recognize diffcult location mentions.Experimental results show that the simple curricula can improve the system's performance on hard cases and its overall performance, and outperform several other baseline systems.The dataset and the baseline models can be found at https: //github.com/brickee/HarveyNER.

Besides the conventional schema-oriented tasks, data dependencies are recently revisited for data quality applications, such as violation detection, data repairing and record matching. To address the variety and veracity issues of big data, data dependencies have been extended as data quality rules to adapt to various data types, ranging from (1) categorical data with equality relationships to (2) heterogeneous data with similarity relationships, and (3) numerical data with order relationships. In this survey, we briefly review the recent proposals on data dependencies categorized into the aforesaid types of data. In addition to (a) the concepts of these data dependency notations, we investigate (b) the extension relationships between data dependencies, e.g., conditional functional dependencies (CFDs) extend the conventional functional dependencies (FDs). It forms a family tree of extensions, mostly rooted in FDs, helping us understand the expressive power of various data dependencies. Moreover, we summarize (c) the discovery of dependencies from data, since data dependencies are often unlikely to be manually specified in a traditional way, given the huge volume and high variety of big data. We further outline (d) the applications of the extended data dependencies, in particular in data quality practice. It guides users to select proper data dependencies with sufficient expressive power and reasonable discovery cost. Finally, we conclude with several directions of future studies on the emerging data.

Multiscale design is the widely practiced use of space and scale to visually explore and articulate relationships. Free-form web curation (FFWC) is an approach to supporting multiscale design, involving creative strategies of collecting content, assembling it to juxtapose and organize, sketching, writing, shifting perspective to navigate, and exhibiting to share and collaborate. Our long term goal is to support design students with automatic, on demand feedback.

We propose to leverage lexical paraphrases and high precision rules informed by news discourse structure to automatically collect coreferential and non-coreferential event pairs from unlabeled English news articles.We perform both manual validation and empirical evaluation on multiple evaluation datasets with different event domains and text genres to assess the quality of our acquired event pairs.We found that a model trained on our acquired event pairs performs comparably as the supervised model when applied to new data out of the training data domains.Further, augmenting human-annotated data with the acquired event pairs provides empirical performance gains on both in-domain and out-of-domain evaluation datasets.

Clustering is often distracted by errors, frequently observed in almost all areas, ranging from online questionnaire to sensor reading in IoT. The dirty data values not only make themselves (the corresponding tuples) outlying, but also mislead the clustering of remaining tuples, e.g., mistakenly splitting a cluster into two or distorting the cluster center. The reason is that the traditional clustering methods either simply ignore the outliers such as DBSCAN or assign them to the closest clusters anyway, e.g., in K-Means. In this paper, we propose to save the outliers for better clustering. The idea is to adjust the erroneous values (often minimally) of the outlier in order to make it appear normally. That is, the tuples after adjusting values are no longer outlying, and thus will be clustered without distracting others. The outlier saving by value adjustment is designed to work with any clustering methods (e.g., DBSCAN or K-Means). Our technical contributions include: (1) showing NPhardness of the outlier saving problem for clustering, (2) deriving lower and upper bounds of the optimal solutions, and (3) devising approximation algorithm with performance guarantees referring to the aforesaid bounds. Experiments on datasets with real-world outliers demonstrate the higher accuracy of our proposal, compared to the state-of-the-art approaches. Remarkably, we show that the adjusted data with outlier saving indeed improve significantly clustering, as well as other applications such as classification and record matching.

Queries over event data are posed in a form of event patterns, for example, to retrieve the flights from IAH to LGA without a stopover. If the expected answer is not returned, one may ask why not, also known as explanations of non-answers. Analogous to the relational data, the explanations over event data lie in two aspects. (1) The pattern consistency explanation indicates that the patterns specified in the query are wrong (inconsistent), that is, there exists no tuple of events that can match the query. (2) The timestamp modification explanation speculates that the instance of event tuple is incorrect, for example, the timestamps of some events are imprecise and need modification. To the best of our knowledge, this is the first study on explaining non-answers over event data. We prove that both explanation problems are NP-complete. By encoding event patterns as a novel notation, we identify the special cases that can be efficiently solved or approximated. General cases are addressed by utilizing the solutions of special cases. Extensive experiments over real and synthetic datasets demonstrate both effectiveness and efficiency of our proposal.

People increasingly use social media to report emergencies, seek help or share information during disasters, which makes social networks an important tool for disaster management. To meet these time-critical needs, we present a weakly supervised approach for rapidly building high-quality classifiers that label each individual Twitter message with fine-grained event categories. Most importantly, we propose a novel method to create high-quality labeled data in a timely manner that automatically clusters tweets containing an event keyword and asks a domain expert to disambiguate event word senses and label clusters quickly. In addition, to process extremely noisy and often rather short user-generated messages, we enrich tweet representations using preceding context tweets and reply tweets in building event recognition classifiers. The evaluation on two hurricanes, Harvey and Florence, shows that using only 1-2 person-hours of human supervision, the rapidly trained weakly supervised classifiers outperform supervised classifiers trained using more than ten thousand annotated tweets created in over 50 person-hours.

Understanding discourse structures of news articles is vital to effectively contextualize the occurrence of a news event.To enable computational modeling of news structures, we apply an existing theory of functional discourse structure for news articles that revolves around the main event and create a human-annotated corpus of 802 documents spanning over four domains and three media sources.Next, we propose several documentlevel neural-network models to automatically construct news content structures.Finally, we demonstrate that incorporating system predicted news structures yields new state-of-theart performance for event coreference resolution.The news documents we annotated are openly available and the annotations are publicly released for future research 1 .

Pattern matching is an important task in the field of Complex Event Processing (CEP). However, exact event pattern matching methods could suffer from low hit rate and loss for meaningful events identification due to the heterogeneous and dirty sources in the big data era. Since both events and patterns could be imprecise, the actual event trace may have different event names as well as structures from the pre-defined pattern. The low-quality data even intensifies the difficulty of matching. In this work, we propose to learn embedding representations for patterns and event traces separately and calculate their similarity as the scores for approximate matching.

We present a new benchmark dataset called PARADE for paraphrase identification that requires specialized domain knowledge.PA-RADE contains paraphrases that overlap very little at the lexical and syntactic level but are semantically equivalent based on computer science domain knowledge, as well as nonparaphrases that overlap greatly at the lexical and syntactic level but are not semantically equivalent based on this domain knowledge.Experiments show that both state-of-the-art neural models and non-expert human annotators have poor performance on PARADE.For example, BERT after fine-tuning achieves an F1 score of 0.709, which is much lower than its performance on other paraphrase identification datasets.PARADE can serve as a resource for researchers interested in testing models that incorporate domain knowledge.We make our data and code freely available.1

Subevents elaborate an event and widely exist in event descriptions.Subevent knowledge is useful for discourse analysis and event-centric applications.Acknowledging the scarcity of subevent knowledge, we propose a weakly supervised approach to extract subevent relation tuples from text and build the first large scale subevent knowledge base.We first obtain the initial set of event pairs that are likely to have the subevent relation, by exploiting two observations that 1) subevents are temporally contained by the parent event, and 2) the definitions of the parent event can be used to further guide the identification of subevents.Then, we collect rich weak supervision using the initial seed subevent pairs to train a contextual classifier using BERT and apply the classifier to identify new subevent pairs.The evaluation showed that the acquired subevent tuples (239K) are of high quality (90.1% accuracy) and cover a wide range of event types.The acquired subevent knowledge has been shown useful for discourse analysis and identifying a range of event-event relations 1 .

Most supervised word sense disambiguation (WSD) systems build word-specific classifiers by leveraging labeled data. However, when using word-specific classifiers, the sparseness of annotations leads to inferior sense disambiguation performance on less frequently seen words. To combat data sparsity, we propose to learn a single model that derives sense representations and meanwhile enforces congruence between a word instance and its right sense by using both sense-annotated data and lexical resources. The model is shared across words that allows utilizing sense correlations across words, and therefore helps to transfer common disambiguation rules from annotation-rich words to annotation-lean words. Empirical evaluation on benchmark datasets shows that the proposed shared model outperforms the equivalent classifier-based models by 1.7%, 2.5% and 3.8% in F1-score when using GloVe, ELMo and BERT word embeddings respectively.

Online reviews play a critical role in persuading or dissuading users when making purchase decisions. And yet very few users take the time to write helpful reviews. Encouragingly, recent advances in deep neural networks offer good potential to produce review-like natural language content. However, there is a lack of large, high-quality labeled data at both the aspect and sentiment level for training. Hence, toward enabling a writing assistant framework to help users post online reviews, this paper proposes a scalable labeling method for bootstrapping aspect and sentiment labels.Concretely, the proposed approach ? Aspect Dependent Online RE-views (ADORE) - leverages the underlying distribution of reviews and a small seed set of labeled data through carefully designed review segmentation and label assignment. We then show how these labels can inform a generative model to produce aspect and sentiment-aware reviews. We study the effectiveness of ADORE under various scenarios such as how end-users perceive the quality of the labels and aspect-aware generated reviews. Our experiments indicate that the proposed effective labeling process along with a regularized joint generative model lead to high quality reviews with 90% accuracy.

Structured entities are commonly abstracted, such as from XML, RDF or hidden-web databases. Direct retrieval of various structured entities is highly demanded in data lakes, e.g., given a JSON object, to find the XML entities that denote the same real-world object. Existing approaches on evaluating structured entity similarity emphasize too much the structural inconsistency. Indeed, entities from heterogeneous sources could have very distinct structures, owing to various information representation conventions. We argue that the retrieval could be more tolerant to structural differences and focus more on the contents of the entities. In this paper, we first identify the unique challenge of parent-child (containment) relationships among structured entities, which unfortunately prevent the retrieval of proper entities (returning parents or children). To solve the problem, a novel hierarchy smooth function is proposed to combine the term scores in different nodes of a structured entity. Entities sharing the same structure, namely an entity family, are employed to learn the coefficient in aggregating the scores, and thus distinguish/prune the parent or child entities. Remarkably, the proposed method could cooperate with both the bag-of-words (BOW) and word embedding models, successful in retrieving unstructured documents, for querying structured entities. Extensive experiments on real datasets demonstrate that our proposal is effective and efficient.

IoT data with timestamps are often found with outliers, such as GPS trajectories or sensor readings. While existing systems mostly focus on detecting temporal outliers without explanations, a decision maker may be more interested in the cause of the outlier appearance such that subsequent actions would be taken, e.g., cleaning unreliable readings or repairing broken devices. Such outlier detection and explanation are expected to be performed in either offline (batch) or online modes (over streaming IoT data with timestamps). In this work, we present TsOutlier, a new prototype system for detecting outliers with explanations over IoT data. The framework defines uniform profiles to explain the outliers detected by various algorithms, including the outliers with variant time intervals. Both batch and streaming processing are supported in a uniform framework. In particular, by varying the block size, it provides a tradeoff between computing the accurate results and approximating with efficient incremental computation. In this paper, we present several case studies of applying TsOutlier in industry, e.g., how this framework works in detecting outliers over the operation data of Shanghai Subway, and how to get reasonable explanations for the detected outliers in tracking excavators.

Zeyu Dai, Ruihong Huang. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Lisa Fan, Marshall White, Eva Sharma, Ruisi Su, Prafulla Kumar Choubey, Ruihong Huang, Lu Wang. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

This paper proposes a novel approach for event coreference resolution that models correlations between event coreference chains and document topical structures through an Integer Linear Programming formulation.We explicitly model correlations between the main event chains of a document with topic transition sentences, inter-coreference chain correlations, event mention distributional characteristics and sub-event structure, and use them with scores obtained from a local coreference relation classifier for jointly resolving multiple event chains in a document.Our experiments across KBP 2016 and 2017 datasets suggest that each of the structures contribute to improving event coreference resolution performance.

Inspired by the double temporality characteristic of narrative texts, we propose a novel approach for acquiring rich temporal "before/after" event knowledge across sentences in narrative stories.The double temporality states that a narrative story often describes a sequence of events following the chronological order and therefore, the temporal order of events matches with their textual order.We explored narratology principles and built a weakly supervised approach that identifies 287k narrative paragraphs from three large text corpora.We then extracted rich temporal event knowledge from these narrative paragraphs.Such event knowledge is shown useful to improve temporal relation classification and outperform several recent neural network models on the narrative cloze task.

Capabilities to categorize a clause based on the type of situation entity (e.g., events, states and generic statements) the clause introduces to the discourse can benefit many NLP applications.Observing that the situation entity type of a clause depends on discourse functions the clause plays in a paragraph and the interpretation of discourse functions depends heavily on paragraph-wide contexts, we propose to build context-aware clause representations for predicting situation entity types of clauses.Specifically, we propose a hierarchical recurrent neural network model to read a whole paragraph at a time and jointly learn representations for all the clauses in the paragraph by extensively modeling context influences and inter-dependencies of clauses.Experimental results show that our model achieves the state-of-the-art performance for clause-level situation entity classification on the genrerich MASC+Wiki corpus, which approaches human-level performance.

We argue that semantic meanings of a sentence or clause can not be interpreted independently from the rest of a paragraph, or independently from all discourse relations and the overall paragraph-level discourse structure.With the goal of improving implicit discourse relation classification, we introduce a paragraph-level neural networks that model inter-dependencies between discourse units as well as discourse relation continuity and patterns, and predict a sequence of discourse relations in a paragraph.Experimental results show that our model outperforms the previous state-of-the-art systems on the benchmark corpus of PDTB.

Understanding time as expressed in text is an important goal of natural language understanding and extremely important for many applications, including information extraction, information retrieval, and question answering. This book provides a comprehensive overview, the challenges, available data resources, and existing systems for the task, with a special emphasis on sensitivity of temporal tagging to domains. This is a well-written book with contents well structured and organized. The discussions on temporal tagging for different domains are valuable and inspirational not only to readers interested in the specific subject of temporal tagging, but also to a broader range of readers interested in Natural Language Processing (NLP) in general. Although it is well known that domain changes often cause significant performance reductions of various NLP systems, little work has been conducted to understand and further explain specific ways such influences have been applied. To a great extent, this book fills the gap in the context of temporal tagging.This book consists of six chapters, which I will group into three parts. The first three chapters give a clear introduction to temporal tagging, including subtasks, characteristics of time, realizations of temporal expressions, data annotation standards, data sets, and evaluation metrics. This initial part provides sufficent background knowledge for further discussion of domain influences on temporal tagging. The fourth chapter is the core of the book, and it identifies four major domains—news-style, narrative-style, colloquial-style, and autonomic-style documents—and elaborates on unique characteristics of each domain and their implications on temporal tagging. The last two chapters describe a list of temporal taggers (full-fledged or focusing on one stage of temporal tagging), compare their designs (rule-based vs. learning-based) and their capabilities of addressing multiple domains and even multiple languages, and conclude with future directions of temporal tagging.Chapter 1 specifies the two subtasks of temporal tagging, temporal expression extraction, and normalization, and explains that temporal tagging can be viewed as a specific type of named entity recognition and normalization. Chapter 1 also briefly describes several temporal tagging applications, including information extraction, information retrieval, and question answering.Chapter 2 clearly describes key characteristics of time. Specifically, time can be normalized and temporal information can be organized in a hierarchical structure based on their granularities. The chapter goes on to describe four categories of temporal expressions in real text (i.e., date, time, duration, and set). I found the discussion on differences between a point in time that may have a duration and a duration of time very interesting. The subsection on realizations of temporal expressions is at the core of this chapter, and clearly defines four types of temporal expressions—explicit, implicit, relative, and underspecified expressions. It is important to distinguish between these types before we examine differences of temporal tagging across domains. Note that recognizing and normalizing each type of temporal expression requires different strategies and is at a different difficulty level. Meanwhile, the authors discuss uncertainty or fuzziness of some temporal expressions. For instance, in “He visited Germany in 2010,” it is rather unlikely that the visit took place the whole year. The exact point or period in 2010 is not known.Chapter 3 surveys annotation standards, describes several evaluation metrics, and provides a comprehensive list of research competitions and annotated news-style corpora. Although I found the description of annotation standards to be generally well thought out and organized, I occasionally felt it was difficult to understand some of the description. For instance, it is difficult to immediately understand the major differences between TIMEX2 and TIMEX3, the description of TIMEX3, and its various tags and abstract tags with no extent in TIMEX3. More examples would be helpful. Instead of sequentially reading each section in this chapter, it may help to first read Section 3.4 on annotated corpora. This chapter also includes an extensive description of different metrics used for measuring temporal tagging performance. The list of research competitions covers all the recent major efforts that were indicated by their adopted corpora, including news-style corpora (MUC, ACE, and TempEval), biomedical texts, QA TempEval (news, wiki, and blogs), and multi-language annotated corpora.Chapter 4 is the core of the book, and defines four major types of domains, examines their unique characteristics, and discusses strategies of temporal tagging for each domain. The chapter starts by describing specific characteristics of news-style documents, which is the dominant type in most studies on temporal tagging. This is followed by a general discussion of genres or domains, covering news, Wikipedia, dialogs, short messages, and clinical reports. Specifically, this book defines a domain as a group of documents that have the same characteristics relevant for the task of temporal tagging. After providing a list of annotated corpora that includes non-news texts, the chapter identifies four broad types of domains—news-style, narrative-style, colloquial-style, and autonomic-style documents—which I found fascinating, especially the fourth domain that features unresolvable time expressions due to local time frames. The discussions of unique features for each domain with respect to temporal tagging are clear and well organized, and directly lead to strategies as suggested by the authors for addressing the task in each domain. Note that each domain is defined in a broad sense and covers texts created in several scenarios. For instance, the news-style documents include “not only news articles but also many other types of documents (e.g., letters and formal blog posts), which are written similarly and thus belong to the same domain from a temporal tagging point of view.”Chapter 5 describes several widely used temporal taggers, both rule-based and learning-based taggers, and compares their temporal tagging performance on documents in different domains. Clearly, taggers prepared or trained for a particular domain do not perform well when applied to a different domain. Consistent with the authors’ vision, this chapter emphasizes that temporal taggers developed with all the domains in mind are preferable. This chapter also argues for developing highly multilingual temporal taggers.Chapter 6 concludes the book by pointing out more directions for the future work of temporal tagging, in order to achieve accurate and complete temporal understanding.To summarize, this book provides timely references to most recent advances of temporal tagging, including both annotated corpora from different domains and systems, as well as insightful discussions and vision in terms of categorizing effects of domains and designing generalized temporal taggers. This book is recommended not only to students, researchers, and developers who work in the field of temporal tagging, but also to a wide range of readers interested in NLP in general.

We present a sequential model for temporal relation classification between intrasentence events.The key observation is that the overall syntactic structure and compositional meanings of the multi-word context between events are important for distinguishing among fine-grained temporal relations.Specifically, our approach first extracts a sequence of context words that indicates the temporal relation between two events, which well align with the dependency path between two event mentions.The context word sequence, together with a parts-of-speech tag sequence and a dependency relation sequence that are generated corresponding to the word sequence, are then provided as input to bidirectional recurrent neural network (LSTM) models.The neural nets learn compositional syntactic and semantic representations of contexts surrounding the two events and predict the temporal relation between them.Evaluation of the proposed approach on TimeBank corpus shows that sequential modeling is capable of accurately recognizing temporal relations between events, which outperforms a neural net model using various discrete features as input that imitates previous feature based models.

We introduce a novel iterative approach for event coreference resolution that gradually builds event clusters by exploiting inter-dependencies among event mentions within the same chain as well as across event chains.Among event mentions in the same chain, we distinguish within-and cross-document event coreference links by using two distinct pairwise classifiers, trained separately to capture differences in feature distributions of within-and crossdocument event clusters.Our event coreference approach alternates between WD and CD clustering and combines arguments from both event clusters after every merge, continuing till no more merge can be made.And then it performs further merging between event chains that are both closely related to a set of other chains of events.Experiments on the ECB+ corpus show that our model outperforms state-of-the-art methods in joint task of WD and CD event coreference resolution.

In the wake of a polarizing election, social media is laden with hateful content. To address various limitations of supervised hate speech classification methods including corpus bias and huge cost of annotation, we propose a weakly supervised two-path bootstrapping approach for an online hate speech detection model leveraging large-scale unlabeled data. This system significantly outperforms hate speech detection systems that are trained in a supervised manner using manually annotated data. Applying this model on a large quantity of tweets collected before, after, and on election day reveals motivations and patterns of inflammatory language.

Identifying events of a specific type is a challenging task as events in texts are described in numerous and diverse ways. Aiming to resolve high complexities of event descriptions, previous work (Huang and Riloff, 2013) proposes multi-faceted event recognition and a bootstrapping method to automatically acquire both event facet phrases and event expressions from unannotated texts. However, to ensure high quality of learned phrases, this method is constrained to only learn phrases that match certain syntactic structures. In this paper, we propose a bilingual structure projection algorithm that explores linguistic divergences between two languages (Chinese and English) and mines new phrases with new syntactic structures, which have been ignored in the previous work. Experiments show that our approach can successfully find novel event phrases and structures, e.g., phrases headed by nouns. Furthermore, the newly mined phrases are capable of recognizing additional event descriptions and increasing the recall of event recognition.

We present our pilot research on automatically extracting subevents from a domain-specific corpus, focusing on the type of subevents that describe physical actions composing an event.We decompose the challenging problem and propose a two-phase approach that effectively captures sentential and local cues that describe subevents.We extracted a rich set of over 600 novel subevent phrases.Evaluation shows the automatically learned subevents help to discover 10% additional main events (of which the learned subevents are a part) and improve event detection performance.

Determining whether a major societal event has already happened, is still on-going, or may occur in the future is crucial for event prediction, timeline generation, and news summarization.We introduce a new task and a new corpus, EventStatus, which has 4500 English and Spanish articles about civil unrest events labeled as PAST, ON-GOING, or FU-TURE.We show that the temporal status of these events is difficult to classify because local tense and aspect cues are often lacking, time expressions are insufficient, and the linguistic contexts have rich semantic compositionality.We explore two approaches for event status classification: (1) a feature-based SVM classifier augmented with a novel induced lexicon of future-oriented verbs, such as "threatened" and "planned", and (2) a convolutional neural net.Both types of classifiers improve event status recognition over a state-of-the-art TempEval model, and our analysis offers linguistic insights into the semantic compositionality challenges for this new task.

We propose a general framework of risk-averse reinforcement learning for algorithmic trading. Our approach is tested in an experiment based on 1.5 years of millisecond time-scale limit order data from NASDAQ, which contain the data around the 2010 flash crash. The results show that our algorithm outperforms the risk-neutral reinforcement learning algorithm by 1) keeping the trading cost at a substantially low level at the spot when the flash crash happened, and 2) significantly reducing the risk over the whole test period.

The goal of our research is to distinguish veterinary message board posts that describe a case involving a specific patient from posts that ask a general question.We create a text classifier that incorporates automatically generated attribute lists for veterinary patients to tackle this problem.Using a small amount of annotated data, we train an information extraction (IE) system to identify veterinary patient attributes.We then apply the IE system to a large collection of unannotated texts to produce a lexicon of veterinary patient attribute terms.Our experimental results show that using the learned attribute lists to encode patient information in the text classifier yields improved performance on this task.

A common form of sarcasm on Twitter consists of a positive sentiment contrasted with a negative situation.For example, many sarcastic tweets include a positive sentiment, such as "love" or "enjoy", followed by an expression that describes an undesirable activity or state (e.g., "taking exams" or "being ignored").We have developed a sarcasm recognizer to identify this type of sarcasm in tweets.We present a novel bootstrapping algorithm that automatically learns lists of positive sentiment phrases and negative situation phrases from sarcastic tweets.We show that identifying contrasting contexts using the phrases learned through bootstrapping yields improved recall for sarcasm recognition.

Event extraction systems typically locate the role fillers for an event by analyzing sentences in isolation and identifying each role filler independently of the others. We argue that more accurate event extraction requires a view of the larger context to decide whether an entity is related to a relevant event. We propose a bottom-up approach to event extraction that initially identifies candidate role fillers independently and then uses that information as well as discourse properties to model textual cohesion. The novel component of the architecture is a sequentially structured sentence classifier that identifies event-related story contexts. The sentence classifier uses lexical associations and discourse relations across sentences, as well as domain-specific distributions of candidate role fillers within and across sentences. This approach yields state-of-the-art performance on the MUC-4 data set, achieving substantially higher precision than previous systems.