We present the results of the first Fact Extraction and VERification (FEVER) Shared Task. The task challenged participants to classify whether human-written factoid claims could be SUPPORTED or REFUTED using evidence retrieved from Wikipedia. We received entries from 23 competing teams, 19 of which scored higher than the previously published baseline. The best performing system achieved a FEVER score of 64.21%. In this paper, we present the results of the shared task and a summary of the systems, highlighting commonalities and innovations among participating systems.

Misinformation detection at the level of full news articles is a text classification problem. Reliably labeled data in this domain is rare. Previous work relied on news articles collected from so-called “reputable” and “suspicious” websites and labeled accordingly. We leverage fact-checking websites to collect individually-labeled news articles with regard to the veracity of their content and use this data to test the cross-domain generalization of a classifier trained on bigger text collections but labeled according to source reputation. Our results suggest that reputation-based classification is not sufficient for predicting the veracity level of the majority of news articles, and that the system performance on different test datasets depends on topic distribution. Therefore collecting well-balanced and carefully-assessed training data is a priority for developing robust misinformation detection systems.

Distant supervision is a popular method for performing relation extraction from text that is known to produce noisy labels. Most progress in relation extraction and classification has been made with crowdsourced corrections to distant-supervised labels, and there is evidence that indicates still more would be better. In this paper, we explore the problem of propagating human annotation signals gathered for open-domain relation classification through the CrowdTruth methodology for crowdsourcing, that captures ambiguity in annotations by measuring inter-annotator disagreement. Our approach propagates annotations to sentences that are similar in a low dimensional embedding space, expanding the number of labels by two orders of magnitude. Our experiments show significant improvement in a sentence-level multi-class relation classifier.

SimpleQuestions is a commonly used benchmark for single-factoid question answering (QA) over Knowledge Graphs (KG). Existing QA systems rely on various components to solve different sub-tasks of the problem (such as entity detection, entity linking, relation prediction and evidence integration). In this work, we propose a different approach to the problem and present an information retrieval style solution for it. We adopt a two-phase approach: candidate generation and candidate re-ranking to answer questions. We propose a Triplet-Siamese-Hybrid CNN (TSHCNN) to re-rank candidate answers. Our approach achieves an accuracy of 80% which sets a new state-of-the-art on the SimpleQuestions dataset.

In this paper we present a browser plugin NewsScan that assists online news readers in evaluating the quality of online content they read by providing information nutrition labels for online news articles. In analogy to groceries, where nutrition labels help consumers make choices that they consider best for themselves, information nutrition labels tag online news articles with data that help readers judge the articles they engage with. This paper discusses the choice of the labels, their implementation and visualization.

Information extraction about an event can be improved by incorporating external evidence. In this study, we propose a joint model for pseudo-relevance feedback based query expansion and information extraction with reinforcement learning. Our model generates an event-specific query to effectively retrieve documents relevant to the event. We demonstrate that our model is comparable or has better performance than the previous model in two publicly available datasets. Furthermore, we analyzed the influences of the retrieval effectiveness in our model on the extraction performance.

In this paper, we propose to adapt the four-staged pipeline proposed by Zubiaga et al. (2018) for the Rumor Verification task to the problem of Fake News Detection. We show that the recently released FNC-1 corpus covers two of its steps, namely the Tracking and the Stance Detection task. We identify asymmetry in length in the input to be a key characteristic of the latter step, when adapted to the framework of Fake News Detection, and propose to handle it as a specific type of Cross-Level Stance Detection. Inspired by theories from the field of Journalism Studies, we implement and test two architectures to successfully model the internal structure of an article and its interactions with a claim.

With the growth of the internet, the number of fake-news online has been proliferating every year. The consequences of such phenomena are manifold, ranging from lousy decision-making process to bullying and violence episodes. Therefore, fact-checking algorithms became a valuable asset. To this aim, an important step to detect fake-news is to have access to a credibility score for a given information source. However, most of the widely used Web indicators have either been shutdown to the public (e.g., Google PageRank) or are not free for use (Alexa Rank). Further existing databases are short-manually curated lists of online sources, which do not scale. Finally, most of the research on the topic is theoretical-based or explore confidential data in a restricted simulation environment. In this paper we explore current research, highlight the challenges and propose solutions to tackle the problem of classifying websites into a credibility scale. The proposed model automatically extracts source reputation cues and computes a credibility factor, providing valuable insights which can help in belittling dubious and confirming trustful unknown websites. Experimental results outperform state of the art in the 2-classes and 5-classes setting.

We present an automated approach to distinguish true, false, stretch, and dodge statements in questions and answers in the Canadian Parliament. We leverage the truthfulness annotations of a U.S. fact-checking corpus by training a neural net model and incorporating the prediction probabilities into our models. We find that in concert with other linguistic features, these probabilities can improve the multi-class classification results. We further show that dodge statements can be detected with an F1 measure as high as 82.57% in binary classification settings.

With the uncontrolled increasing of fake news and rumors over the Web, different approaches have been proposed to address the problem. In this paper, we present an approach that combines lexical, word embeddings and n-gram features to detect the stance in fake news. Our approach has been tested on the Fake News Challenge (FNC-1) dataset. Given a news title-article pair, the FNC-1 task aims at determining the relevance of the article and the title. Our proposed approach has achieved an accurate result (59.6 % Macro F1) that is close to the state-of-the-art result with 0.013 difference using a simple feature representation. Furthermore, we have investigated the importance of different lexicons in the detection of the classification labels.

We consider the task of relation classification, and pose this task as one of textual entailment. We show that this formulation leads to several advantages, including the ability to (i) perform zero-shot relation classification by exploiting relation descriptions, (ii) utilize existing textual entailment models, and (iii) leverage readily available textual entailment datasets, to enhance the performance of relation classification systems. Our experiments show that the proposed approach achieves 20.16% and 61.32% in F1 zero-shot classification performance on two datasets, which further improved to 22.80% and 64.78% respectively with the use of conditional encoding.

Existing entailment datasets mainly pose problems which can be answered without attention to grammar or word order. Learning syntax requires comparing examples where different grammar and word order change the desired classification. We introduce several datasets based on synthetic transformations of natural entailment examples in SNLI or FEVER, to teach aspects of grammar and word order. We show that without retraining, popular entailment models are unaware that these syntactic differences change meaning. With retraining, some but not all popular entailment models can learn to compare the syntax properly.

Fact-checking is a journalistic practice that compares a claim made publicly against trusted sources of facts. Wang (2017) introduced a large dataset of validated claims from the POLITIFACT.com website (LIAR dataset), enabling the development of machine learning approaches for fact-checking. However, approaches based on this dataset have focused primarily on modeling the claim and speaker-related metadata, without considering the evidence used by humans in labeling the claims. We extend the LIAR dataset by automatically extracting the justification from the fact-checking article used by humans to label a given claim. We show that modeling the extracted justification in conjunction with the claim (and metadata) provides a significant improvement regardless of the machine learning model used (feature-based or deep learning) both in a binary classification task (true, false) and in a six-way classification task (pants on fire, false, mostly false, half true, mostly true, true).

Common-sense reasoning is becoming increasingly important for the advancement of Natural Language Processing. While word embeddings have been very successful, they cannot explain which aspects of ‘coffee’ and ‘tea’ make them similar, or how they could be related to ‘shop’. In this paper, we propose an explicit word representation that builds upon the Distributional Hypothesis to represent meaning from semantic roles, and allow inference of relations from their meshing, as supported by the affordance-based Indexical Hypothesis. We find that our model improves the state-of-the-art on unsupervised word similarity tasks while allowing for direct inference of new relations from the same vector space.

In this paper we describe our 2nd place FEVER shared-task system that achieved a FEVER score of 62.52% on the provisional test set (without additional human evaluation), and 65.41% on the development set. Our system is a four stage model consisting of document retrieval, sentence retrieval, natural language inference and aggregation. Retrieval is performed leveraging task-specific features, and then a natural language inference model takes each of the retrieved sentences paired with the claimed fact. The resulting predictions are aggregated across retrieved sentences with a Multi-Layer Perceptron, and re-ranked corresponding to the final prediction.

The Fact Extraction and VERification (FEVER) shared task was launched to support the development of systems able to verify claims by extracting supporting or refuting facts from raw text. The shared task organizers provide a large-scale dataset for the consecutive steps involved in claim verification, in particular, document retrieval, fact extraction, and claim classification. In this paper, we present our claim verification pipeline approach, which, according to the preliminary results, scored third in the shared task, out of 23 competing systems. For the document retrieval, we implemented a new entity linking approach. In order to be able to rank candidate facts and classify a claim on the basis of several selected facts, we introduce two extensions to the Enhanced LSTM (ESIM).

We develop a system for the FEVER fact extraction and verification challenge that uses a high precision entailment classifier based on transformer networks pretrained with language modeling, to classify a broad set of potential evidence. The precision of the entailment classifier allows us to enhance recall by considering every statement from several articles to decide upon each claim. We include not only the articles best matching the claim text by TFIDF score, but read additional articles whose titles match named entities and capitalized expressions occurring in the claim text. The entailment module evaluates potential evidence one statement at a time, together with the title of the page the evidence came from (providing a hint about possible pronoun antecedents). In preliminary evaluation, the system achieves .5736 FEVER score, .6108 label accuracy, and .6485 evidence F1 on the FEVER shared task test set.

In this paper we present our system for the FEVER Challenge. The task of this challenge is to verify claims by extracting information from Wikipedia. Our system has two parts. In the first part it performs a search for candidate sentences by treating the claims as query. In the second part it filters out noise from these candidates and uses the remaining ones to decide whether they support or refute or entail not enough information to verify the claim. We show that this system achieves a FEVER score of 0.3927 on the FEVER shared task development data set which is a 25.5% improvement over the baseline score.

This article presents the SIRIUS-LTG system for the Fact Extraction and VERification (FEVER) Shared Task. It consists of three components: 1) Wikipedia Page Retrieval: First we extract the entities in the claim, then we find potential Wikipedia URI candidates for each of the entities using a SPARQL query over DBpedia 2) Sentence selection: We investigate various techniques i.e. Smooth Inverse Frequency (SIF), Word Mover’s Distance (WMD), Soft-Cosine Similarity, Cosine similarity with unigram Term Frequency Inverse Document Frequency (TF-IDF) to rank sentences by their similarity to the claim. 3) Textual Entailment: We compare three models for the task of claim classification. We apply a Decomposable Attention (DA) model (Parikh et al., 2016), a Decomposed Graph Entailment (DGE) model (Khot et al., 2018) and a Gradient-Boosted Decision Trees (TalosTree) model (Sean et al., 2017) for this task. The experiments show that the pipeline with simple Cosine Similarity using TFIDF in sentence selection along with DA model as labelling model achieves the best results on the development set (F1 evidence: 32.17, label accuracy: 59.61 and FEVER score: 0.3778). Furthermore, it obtains 30.19, 48.87 and 36.55 in terms of F1 evidence, label accuracy and FEVER score, respectively, on the test set. Our system ranks 15th among 23 participants in the shared task prior to any human-evaluation of the evidence.

We describe here our system and results on the FEVER shared task. We prepared a pipeline system which composes of a document selection, a sentence retrieval, and a recognizing textual entailment (RTE) components. A simple entity linking approach with text match is used as the document selection component, this component identifies relevant documents for a given claim by using mentioned entities as clues. The sentence retrieval component selects relevant sentences as candidate evidence from the documents based on TF-IDF. Finally, the RTE component selects evidence sentences by ranking the sentences and classifies the claim simultaneously. The experimental results show that our system achieved the FEVER score of 0.4016 and outperformed the official baseline system.

This paper presents the ColumbiaNLP submission for the FEVER Workshop Shared Task. Our system is an end-to-end pipeline that extracts factual evidence from Wikipedia and infers a decision about the truthfulness of the claim based on the extracted evidence. Our pipeline achieves significant improvement over the baseline for all the components (Document Retrieval, Sentence Selection and Textual Entailment) both on the development set and the test set. Our team finished 6th out of 24 teams on the leader-board based on the preliminary results with a FEVER score of 49.06 on the blind test set compared to 27.45 of the baseline system.

In this paper, we describe DeFactoNLP, the system we designed for the FEVER 2018 Shared Task. The aim of this task was to conceive a system that can not only automatically assess the veracity of a claim but also retrieve evidence supporting this assessment from Wikipedia. In our approach, the Wikipedia documents whose Term Frequency-Inverse Document Frequency (TFIDF) vectors are most similar to the vector of the claim and those documents whose names are similar to those of the named entities (NEs) mentioned in the claim are identified as the documents which might contain evidence. The sentences in these documents are then supplied to a textual entailment recognition module. This module calculates the probability of each sentence supporting the claim, contradicting the claim or not providing any relevant information to assess the veracity of the claim. Various features computed using these probabilities are finally used by a Random Forest classifier to determine the overall truthfulness of the claim. The sentences which support this classification are returned as evidence. Our approach achieved a 0.4277 evidence F1-score, a 0.5136 label accuracy and a 0.3833 FEVER score.

With huge amount of information generated every day on the web, fact checking is an important and challenging task which can help people identify the authenticity of most claims as well as providing evidences selected from knowledge source like Wikipedia. Here we decompose this problem into two parts: an entity linking task (retrieving relative Wikipedia pages) and recognizing textual entailment between the claim and selected pages. In this paper, we present an end-to-end multi-task learning with bi-direction attention (EMBA) model to classify the claim as “supports”, “refutes” or “not enough info” with respect to the pages retrieved and detect sentences as evidence at the same time. We conduct experiments on the FEVER (Fact Extraction and VERification) paper test dataset and shared task test dataset, a new public dataset for verification against textual sources. Experimental results show that our method achieves comparable performance compared with the baseline system.

In this system description of our pipeline to participate at the Fever Shared Task, we describe our sentence-based approach. Throughout all steps of our pipeline, we regarded single sentences as our processing unit. In our IR-Component, we searched in the set of all possible Wikipedia introduction sentences without limiting sentences to a fixed number of relevant documents. In the entailment module, we judged every sentence separately and combined the result of the classifier for the top 5 sentences with the help of an ensemble classifier to make a judgment whether the truth of a statement can be derived from the given claim.

Many tasks such as question answering and reading comprehension rely on information extracted from unreliable sources. These systems would thus benefit from knowing whether a statement from an unreliable source is correct. We present experiments on the FEVER (Fact Extraction and VERification) task, a shared task that involves selecting sentences from Wikipedia and predicting whether a claim is supported by those sentences, refuted, or there is not enough information. Fact checking is a task that benefits from not only asserting or disputing the veracity of a claim but also finding evidence for that position. As these tasks are dependent on each other, an ideal model would consider the veracity of the claim when finding evidence and also find only the evidence that is relevant. We thus jointly model sentence extraction and verification on the FEVER shared task. Among all participants, we ranked 5th on the blind test set (prior to any additional human evaluation of the evidence).

This paper describes our system submission to the 2018 Fact Extraction and VERification (FEVER) shared task. The system uses a heuristics-based approach for evidence extraction and a modified version of the inference model by Parikh et al. (2016) for classification. Our process is broken down into three modules: potentially relevant documents are gathered based on key phrases in the claim, then any possible evidence sentences inside those documents are extracted, and finally our classifier discards any evidence deemed irrelevant and uses the remaining to classify the claim’s veracity. Our system beats the shared task baseline by 12% and is successful at finding correct evidence (evidence retrieval F1 of 62.5% on the development set).

We describe our system used in the 2018 FEVER shared task. The system employed a frame-based information retrieval approach to select Wikipedia sentences providing evidence and used a two-layer multilayer perceptron to classify a claim as correct or not. Our submission achieved a score of 0.3966 on the Evidence F1 metric with accuracy of 44.79%, and FEVER score of 0.2628 F1 points.

Many approaches to automatically recognizing entailment relations have employed classifiers over hand engineered lexicalized features, or deep learning models that implicitly capture lexicalization through word embeddings. This reliance on lexicalization may complicate the adaptation of these tools between domains. For example, such a system trained in the news domain may learn that a sentence like “Palestinians recognize Texas as part of Mexico” tends to be unsupported, but this fact (and its corresponding lexicalized cues) have no value in, say, a scientific domain. To mitigate this dependence on lexicalized information, in this paper we propose a model that reads two sentences, from any given domain, to determine entailment without using lexicalized features. Instead our model relies on features that are either unlexicalized or are domain independent such as proportion of negated verbs, antonyms, or noun overlap. In its current implementation, this model does not perform well on the FEVER dataset, due to two reasons. First, for the information retrieval portion of the task we used the baseline system provided, since this was not the aim of our project. Second, this is work in progress and we still are in the process of identifying more features and gradually increasing the accuracy of our model. In the end, we hope to build a generic end-to-end classifier, which can be used in a domain outside the one in which it was trained, with no or minimal re-training.