Answering natural language questions over tables is usually seen as a semantic parsing task. To alleviate the collection cost of full logical forms, one popular approach focuses on weak supervision consisting of denotations instead of logical forms. However, training semantic parsers from weak supervision poses difficulties, and in addition, the generated logical forms are only used as an intermediate step prior to retrieving the denotation. In this paper, we present TAPAS, an approach to question answering over tables without generating logical forms. TAPAS trains from weak supervision, and predicts the denotation by selecting table cells and optionally applying a corresponding aggregation operator to such selection. TAPAS extends BERT's architecture to encode tables as input, initializes from an effective joint pre-training of text segments and tables crawled from Wikipedia, and is trained end-to-end. We experiment with three different semantic parsing datasets, and find that TAPAS outperforms or rivals semantic parsing models by improving state-of-the-art accuracy on SQA from 55.1 to 67.2 and performing on par with the state-of-the-art on WIKISQL and WIKITQ, but with a simpler model architecture. We additionally find that transfer learning, which is trivial in our setting, from WIK-ISQL to WIKITQ, yields 48.7 accuracy, 4.2 points above the state-of-the-art.

Structured information about entities is critical for many semantic parsing tasks. We present an approach that uses a Graph Neural Network (GNN) architecture to incorporate information about relevant entities and their relations during parsing. Combined with a decoder copy mechanism, this approach provides a conceptually simple mechanism to generate logical forms with entities. We demonstrate that this approach is competitive with the stateof-the-art across several tasks without pretraining, and outperforms existing approaches when combined with BERT pre-training.

We present a novel approach to answering sequential questions based on structured objects such as knowledge bases or tables without using a logical form as an intermediate representation. We encode tables as graphs using a graph neural network model based on the Transformer architecture. The answers are then selected from the encoded graph using a pointer network. This model is appropriate for processing conversations around structured data, where the attention mechanism that selects the answers to a question can also be used to resolve conversational references. We demonstrate the validity of this approach with competitive results on the Sequential Question Answering (SQA) task (Iyyer et al., 2017).

The need to bridge between the unstructured data on the Document Web and the structured data on the Web of Data has led to the development of a considerable number of annotation tools. However, these tools are currently still hard to compare since the published evaluation results are calculated on diverse datasets and evaluated based on different measures. We present GERBIL, an evaluation framework for semantic entity annotation. The rationale behind our framework is to provide developers, end users and researchers with easy-to-use interfaces that allow for the agile, fine-grained and uniform evaluation of annotation tools on multiple datasets. By these means, we aim to ensure that both tool developers and end users can derive meaningful insights pertaining to the extension, integration and use of annotation applications. In particular, GERBIL provides comparable results to tool developers so as to allow them to easily discover the strengths and weaknesses of their implementations with respect to the state of the art. With the permanent experiment URIs provided by our framework, we ensure the reproducibility and archiving of evaluation results. Moreover, the framework generates data in machineprocessable format, allowing for the efficient querying and post-processing of evaluation results. Finally, the tool diagnostics provided by GERBIL allows deriving insights pertaining to the areas in which tools should be further refined, thus allowing developers to create an informed agenda for extensions and end users to detect the right tools for their purposes. GERBIL aims to become a focal point for the state of the art, driving the research agenda of the community by presenting comparable objective evaluation results.

In this paper we propose a novel entity annotator for texts which hinges on Tagme's algorithmic technology, currently the best one available [6,4]. The novelty is twofold: from the one hand, we have engineered the software in order to be modular and more efficient; from the other hand, we have improved the annotation pipeline by re-designing all of its three main modules: spotting, disambiguation and pruning. In particular, the re-design has involved the detailed inspection of the performance of these modules by developing new algorithms which have been in turn tested over all publicly available datasets (i.e. AIDA, IITB, MSN, AQUAINT, and the one of the ERD Challenge).This extensive experimentation allowed us to derive the best combination which achieved on the ERD development dataset an F1 score of 74.8%, which turned to be 67.2% F1 for the test dataset. This final result was due to an impressive precision equal to 87.6%, but very low recall 54.5%. With respect to classic Tagme on the development dataset the improvement ranged from 1% to 9% on the D2W benchmark, depending on the disambiguation algorithm being used.As a side result, the final software can be interpreted as a flexible library of several parsing/disambiguation and pruning modules that can be used to build up new and more sophisticated entity annotators. We plan to release our library to the public as an open-source project.