arXiv subjects

Cornelia Caragea

Publications and source records attributed to Cornelia Caragea.

70 records · Page 4Linked to original sources

arXiv2022

On the Evaluation of Answer-Agnostic Paragraph-level Multi-Question Generation

We study the task of predicting a set of salient questions from a given paragraph without any prior knowledge of the precise answer. We make two main contributions. First, we propose a new method to evaluate a set of predicted questions against the set of references by using the Hungarian algorithm to assign predicted questions to references before scoring the assigned pairs. We show that our proposed evaluation strategy has better theoretical and practical properties compared to prior methods because it can properly account for the coverage of references. Second, we compare different strategies to utilize a pre-trained seq2seq model to generate and select a set of questions related to a given paragraph. The code is available.

cs.CL↗

arXiv2021

Generating Summaries for Scientific Paper Review

The review process is essential to ensure the quality of publications. Recently, the increase of submissions for top venues in machine learning and NLP has caused a problem of excessive burden on reviewers and has often caused concerns regarding how this may not only overload reviewers, but also may affect the quality of the reviews. An automatic system for assisting with the reviewing process could be a solution for ameliorating the problem. In this paper, we explore automatic review summary generation for scientific papers. We posit that neural language models have the potential to be valuable candidates for this task. In order to test this hypothesis, we release a new dataset of scientific papers and their reviews, collected from papers published in the NeurIPS conference from 2013 to 2020. We evaluate state of the art neural summarization models, present initial results on the feasibility of automatic review summary generation, and propose directions for the future.

cs.CL↗

arXiv2021

DeepZensols: Deep Natural Language Processing Framework

Reproducing results in publications by distributing publicly available source code is becoming ever more popular. Given the difficulty of reproducing machine learning (ML) experiments, there have been significant efforts in reducing the variance of these results. As in any science, the ability to consistently reproduce results effectively strengthens the underlying hypothesis of the work, and thus, should be regarded as important as the novel aspect of the research itself. The contribution of this work is a framework that is able to reproduce consistent results and provides a means of easily creating, training, and evaluating natural language processing (NLP) deep learning (DL) models.

cs.CL↗

arXiv2021

Modeling Hierarchical Structures with Continuous Recursive Neural Networks

Recursive Neural Networks (RvNNs), which compose sequences according to their underlying hierarchical syntactic structure, have performed well in several natural language processing tasks compared to similar models without structural biases. However, traditional RvNNs are incapable of inducing the latent structure in a plain text sequence on their own. Several extensions have been proposed to overcome this limitation. Nevertheless, these extensions tend to rely on surrogate gradients or reinforcement learning at the cost of higher bias or variance. In this work, we propose Continuous Recursive Neural Network (CRvNN) as a backpropagation-friendly alternative to address the aforementioned limitations. This is done by incorporating a continuous relaxation to the induced structure. We demonstrate that CRvNN achieves strong performance in challenging synthetic tasks such as logical inference and ListOps. We also show that CRvNN performs comparably or better than prior latent structure models on real-world tasks such as sentiment analysis and natural language inference.

cs.CL↗

arXiv2020

Identifying Documents In-Scope of a Collection from Web Archives

Web archive data usually contains high-quality documents that are very useful for creating specialized collections of documents, e.g., scientific digital libraries and repositories of technical reports. In doing so, there is a substantial need for automatic approaches that can distinguish the documents of interest for a collection out of the huge number of documents collected by web archiving institutions. In this paper, we explore different learning models and feature representations to determine the best performing ones for identifying the documents of interest from the web archived data. Specifically, we study both machine learning and deep learning models and "bag of words" (BoW) features extracted from the entire document or from specific portions of the document, as well as structural features that capture the structure of documents. We focus our evaluation on three datasets that we created from three different Web archives. Our experimental results show that the BoW classifiers that focus only on specific portions of the documents (rather than the full text) outperform all compared methods on all three datasets.

cs.IR↗

arXiv2020

Interpretable Multi-Step Reasoning with Knowledge Extraction on Complex Healthcare Question Answering

Healthcare question answering assistance aims to provide customer healthcare information, which widely appears in both Web and mobile Internet. The questions usually require the assistance to have proficient healthcare background knowledge as well as the reasoning ability on the knowledge. Recently a challenge involving complex healthcare reasoning, HeadQA dataset, has been proposed, which contains multiple-choice questions authorized for the public healthcare specialization exam. Unlike most other QA tasks that focus on linguistic understanding, HeadQA requires deeper reasoning involving not only knowledge extraction, but also complex reasoning with healthcare knowledge. These questions are the most challenging for current QA systems, and the current performance of the state-of-the-art method is slightly better than a random guess. In order to solve this challenging task, we present a Multi-step reasoning with Knowledge extraction framework (MurKe). The proposed framework first extracts the healthcare knowledge as supporting documents from the large corpus. In order to find the reasoning chain and choose the correct answer, MurKe iterates between selecting the supporting documents, reformulating the query representation using the supporting documents and getting entailment score for each choice using the entailment model. The reformulation module leverages selected documents for missing evidence, which maintains interpretability. Moreover, we are striving to make full use of off-the-shelf pre-trained models. With less trainable weight, the pre-trained model can easily adapt to healthcare tasks with limited training samples. From the experimental results and ablation study, our system is able to outperform several strong baselines on the HeadQA dataset.

cs.AI↗

arXiv2020

Detecting Perceived Emotions in Hurricane Disasters

Natural disasters (e.g., hurricanes) affect millions of people each year, causing widespread destruction in their wake. People have recently taken to social media websites (e.g., Twitter) to share their sentiments and feelings with the larger community. Consequently, these platforms have become instrumental in understanding and perceiving emotions at scale. In this paper, we introduce HurricaneEmo, an emotion dataset of 15,000 English tweets spanning three hurricanes: Harvey, Irma, and Maria. We present a comprehensive study of fine-grained emotions and propose classification tasks to discriminate between coarse-grained emotion groups. Our best BERT model, even after task-guided pre-training which leverages unlabeled Twitter data, achieves only 68% accuracy (averaged across all groups). HurricaneEmo serves not only as a challenging benchmark for models but also as a valuable resource for analyzing emotions in disaster-centric domains.

cs.CL↗

arXiv2020

On Identifying Hashtags in Disaster Twitter Data

Tweet hashtags have the potential to improve the search for information during disaster events. However, there is a large number of disaster-related tweets that do not have any user-provided hashtags. Moreover, only a small number of tweets that contain actionable hashtags are useful for disaster response. To facilitate progress on automatic identification (or extraction) of disaster hashtags for Twitter data, we construct a unique dataset of disaster-related tweets annotated with hashtags useful for filtering actionable information. Using this dataset, we further investigate Long Short Term Memory-based models within a Multi-Task Learning framework. The best performing model achieves an F1-score as high as 92.22%. The dataset, code, and other resources are available on Github.

cs.IR↗

arXiv2019

Keyphrase Extraction from Disaster-related Tweets

While keyphrase extraction has received considerable attention in recent years, relatively few studies exist on extracting keyphrases from social media platforms such as Twitter, and even fewer for extracting disaster-related keyphrases from such sources. During a disaster, keyphrases can be extremely useful for filtering relevant tweets that can enhance situational awareness. Previously, joint training of two different layers of a stacked Recurrent Neural Network for keyword discovery and keyphrase extraction had been shown to be effective in extracting keyphrases from general Twitter data. We improve the model's performance on both general Twitter data and disaster-related Twitter data by incorporating contextual word embeddings, POS-tags, phonetics, and phonological features. Moreover, we discuss the shortcomings of the often used F1-measure for evaluating the quality of predicted keyphrases with respect to the ground truth annotations. Instead of the F1-measure, we propose the use of embedding-based metrics to better capture the correctness of the predicted keyphrases. In addition, we also present a novel extension of an embedding-based metric. The extension allows one to better control the penalty for the difference in the number of ground-truth and predicted keyphrases

cs.IR↗

arXiv2019

Cleaning Noisy and Heterogeneous Metadata for Record Linking Across Scholarly Big Datasets

Automatically extracted metadata from scholarly documents in PDF formats is usually noisy and heterogeneous, often containing incomplete fields and erroneous values. One common way of cleaning metadata is to use a bibliographic reference dataset. The challenge is to match records between corpora with high precision. The existing solution which is based on information retrieval and string similarity on titles works well only if the titles are cleaned. We introduce a system designed to match scholarly document entities with noisy metadata against a reference dataset. The blocking function uses the classic BM25 algorithm to find the matching candidates from the reference data that has been indexed by ElasticSearch. The core components use supervised methods which combine features extracted from all available metadata fields. The system also leverages available citation information to match entities. The combination of metadata and citation achieves high accuracy that significantly outperforms the baseline method on the same test dataset. We apply this system to match the database of CiteSeerX against Web of Science, PubMed, and DBLP. This method will be deployed in the CiteSeerX system to clean metadata and link records to other scholarly big datasets.

cs.DL↗

arXiv2019

Image Privacy Prediction Using Deep Neural Networks

Images today are increasingly shared online on social networking sites such as Facebook, Flickr, Foursquare, and Instagram. Despite that current social networking sites allow users to change their privacy preferences, this is often a cumbersome task for the vast majority of users on the Web, who face difficulties in assigning and managing privacy settings. Thus, automatically predicting images' privacy to warn users about private or sensitive content before uploading these images on social networking sites has become a necessity in our current interconnected world. In this paper, we explore learning models to automatically predict appropriate images' privacy as private or public using carefully identified image-specific features. We study deep visual semantic features that are derived from various layers of Convolutional Neural Networks (CNNs) as well as textual features such as user tags and deep tags generated from deep CNNs. Particularly, we extract deep (visual and tag) features from four pre-trained CNN architectures for object recognition, i.e., AlexNet, GoogLeNet, VGG-16, and ResNet, and compare their performance for image privacy prediction. Results of our experiments on a Flickr dataset of over thirty thousand images show that the learning models trained on features extracted from ResNet outperform the state-of-the-art models for image privacy prediction. We further investigate the combination of user tags and deep tags derived from CNN architectures using two settings: (1) SVM on the bag-of-tags features; and (2) text-based CNN. Our results show that even though the models trained on the visual features perform better than those trained on the tag features, the combination of deep visual features with image tags shows improvements in performance over the individual feature sets.

cs.CV↗

arXiv2019

Dynamic Deep Multi-modal Fusion for Image Privacy Prediction

With millions of images that are shared online on social networking sites, effective methods for image privacy prediction are highly needed. In this paper, we propose an approach for fusing object, scene context, and image tags modalities derived from convolutional neural networks for accurately predicting the privacy of images shared online. Specifically, our approach identifies the set of most competent modalities on the fly, according to each new target image whose privacy has to be predicted. The approach considers three stages to predict the privacy of a target image, wherein we first identify the neighborhood images that are visually similar and/or have similar sensitive content as the target image. Then, we estimate the competence of the modalities based on the neighborhood images. Finally, we fuse the decisions of the most competent modalities and predict the privacy label for the target image. Experimental results show that our approach predicts the sensitive (or private) content more accurately than the models trained on individual modalities (object, scene, and tags) and prior privacy prediction works. Also, our approach outperforms strong baselines, that train meta-classifiers to obtain an optimal combination of modalities.

cs.CV↗

arXiv2016

A Search/Crawl Framework for Automatically Acquiring Scientific Documents

Despite the advancements in search engine features, ranking methods, technologies, and the availability of programmable APIs, current-day open-access digital libraries still rely on crawl-based approaches for acquiring their underlying document collections. In this paper, we propose a novel search-driven framework for acquiring documents for scientific portals. Within our framework, publicly-available research paper titles and author names are used as queries to a Web search engine. Next, research papers and sources of research papers are identified from the search results using accurate classification modules. Our experiments highlight not only the performance of our individual classifiers but also the effectiveness of our overall Search/Crawl framework. Indeed, we were able to obtain approximately 0.665 million research documents through our fully-automated framework using about 0.076 million queries. These prolific results position Web search as an effective alternative to crawl methods for acquiring both the actual documents and seed URLs for future crawls.

cs.IR↗

arXiv2015

Privacy Prediction of Images Shared on Social Media Sites Using Deep Features

Online image sharing in social media sites such as Facebook, Flickr, and Instagram can lead to unwanted disclosure and privacy violations, when privacy settings are used inappropriately. With the exponential increase in the number of images that are shared online every day, the development of effective and efficient prediction methods for image privacy settings are highly needed. The performance of models critically depends on the choice of the feature representation. In this paper, we present an approach to image privacy prediction that uses deep features and deep image tags as feature representations. Specifically, we explore deep features at various neural network layers and use the top layer (probability) as an auto-annotation mechanism. The results of our experiments show that models trained on the proposed deep features and deep image tags substantially outperform baselines such as those based on SIFT and GIST as well as those that use "bag of tags" as features.

cs.CV↗

arXiv2015

Entity-Specific Sentiment Classification of Yahoo News Comments

Sentiment classification is widely used for product reviews and in online social media such as forums, Twitter, and blogs. However, the problem of classifying the sentiment of user comments on news sites has not been addressed yet. News sites cover a wide range of domains including politics, sports, technology, and entertainment, in contrast to other online social sites such as forums and review sites, which are specific to a particular domain. A user associated with a news site is likely to post comments on diverse topics (e.g., politics, smartphones, and sports) or diverse entities (e.g., Obama, iPhone, or Google). Classifying the sentiment of users tied to various entities may help obtain a holistic view of their personality, which could be useful in applications such as online advertising, content personalization, and political campaign planning. In this paper, we formulate the problem of entity-specific sentiment classification of comments posted on news articles in Yahoo News and propose novel features that are specific to news comments. Experimental results show that our models outperform state-of-the-art baselines.

cs.CL↗

arXiv2014

Keyword and Keyphrase Extraction Using Centrality Measures on Collocation Networks

Keyword and keyphrase extraction is an important problem in natural language processing, with applications ranging from summarization to semantic search to document clustering. Graph-based approaches to keyword and keyphrase extraction avoid the problem of acquiring a large in-domain training corpus by applying variants of PageRank algorithm on a network of words. Although graph-based approaches are knowledge-lean and easily adoptable in online systems, it remains largely open whether they can benefit from centrality measures other than PageRank. In this paper, we experiment with an array of centrality measures on word and noun phrase collocation networks, and analyze their performance on four benchmark datasets. Not only are there centrality measures that perform as well as or better than PageRank, but they are much simpler (e.g., degree, strength, and neighborhood size). Furthermore, centrality-based methods give results that are competitive with and, in some cases, better than two strong unsupervised baselines.

cs.CL↗