Large language models have recently risen in popularity due to their ability to perform many natural language tasks without requiring any fine-tuning. In this work, we focus on two novel ideas: (1) generating definitions from examples and using them for zero-shot classification, and (2) investigating how an LLM makes use of the definitions. We thoroughly analyze the performance of GPT-3 model for fine-grained multi-label conspiracy theory classification of tweets using zero-shot labeling. In doing so, we asses how to improve the labeling by providing minimal but meaningful context in the form of the definitions of the labels. We compare descriptive noun phrases, human-crafted definitions, introduce a new method to help the model generate definitions from examples, and propose a method to evaluate GPT-3’s understanding of the definitions. We demonstrate that improving definitions of class labels has a direct consequence on the downstream classification results.

In the last few years, the natural language processing community has witnessed advances in neural representations of free texts with transformer-based language models (LMs). Given the importance of knowledge available in tabular data, recent research efforts extend LMs by developing neural representations for structured data. In this article, we present a survey that analyzes these efforts. We first abstract the different systems according to a traditional machine learning pipeline in terms of training data, input representation, model training, and supported downstream tasks. For each aspect, we characterize and compare the proposed solutions. Finally, we discuss future work directions.

One reason for the positive impact of Pre-trained Language Models (PLMs) in NLP tasks is their ability to encode semantic types, such as ‘European City’ or ‘Woman’. While previous work has analyzed such information in the context of interpretability, it is not clear how to use types to steer the PLM output. For example, in a cloze statement, it is desirable to steer the model to generate a token that satisfies a user-specified type, e.g., predict a date rather than a location. In this work, we introduce Type Embeddings (TEs), an input embedding that promotes desired types in a PLM. Our proposal is to define a type by a small set of word examples. We empirically study the ability of TEs both in representing types and in steering masking predictions without changes to the prompt text in BERT. Finally, using the LAMA datasets, we show how TEs highly improve the precision in extracting facts from PLMs.

Computational fact-checking aims at supporting the verification process of textual claims by exploiting trustworthy sources. However, there are large classes of complex claims that cannot be automatically verified, for instance those related to temporal reasoning. To this aim, in this work, we focus on the verification of economic claims against time series sources. Starting from given textual claims in natural language, we propose a neural machine translation approach to produce respective queries expressed in a recently proposed temporal fragment of the Datalog language. The adopted deep neural approach shows promising preliminary results for the translation of 10 categories of claims extracted from real use cases.

Computational fact-checking has gained a lot of traction in the machine learning and natural language processing communities. A plethora of solutions have been developed, but methods which leverage both structured and unstructured information to detect misinformation are of particular relevance. In this paper, we tackle the FEVEROUS (Fact Extraction and VERification Over Unstructured and Structured information) challenge which consists of an open source baseline system together with a benchmark dataset containing 87,026 verified claims. We extend this baseline model by improving the evidence retrieval module yielding the best evidence F1 score among the competitors in the challenge leaderboard while obtaining an overall FEVEROUS score of 0.20 (5th best ranked system).

We present a generic method to compute thefactual accuracy of a generated data summarywith minimal user effort. We look at the prob-lem as a fact-checking task to verify the nu-merical claims in the text. The verification al-gorithm assumes that the data used to generatethe text is available. In this paper, we describehow the proposed solution has been used toidentify incorrect claims about basketball tex-tual summaries in the context of the AccuracyShared Task at INLG 2021.

While pre-trained language models (PLMs) are the go-to solution to tackle many natural language processing problems, they are still very limited in their ability to capture and to use common-sense knowledge. In fact, even if information is available in the form of approximate (soft) logical rules, it is not clear how to transfer it to a PLM in order to improve its performance for deductive reasoning tasks. Here, we aim to bridge this gap by teaching PLMs how to reason with soft Horn rules. We introduce a classification task where, given facts and soft rules, the PLM should return a prediction with a probability for a given hypothesis. We release the first dataset for this task, and we propose a revised loss function that enables the PLM to learn how to predict precise probabilities for the task. Our evaluation results show that the resulting fine-tuned models achieve very high performance, even on logical rules that were unseen at training. Moreover, we demonstrate that logical notions expressed by the rules are transferred to the fine-tuned model, yielding state-of-the-art results on external datasets.