Learning cell embeddings for understanding table layouts

摘要

There is a large amount of data on the web in tabular form, such as Excel sheets, CSV files, and web tables. Often, tabular data is meant for human consumption, using data layouts that are difficult for machines to interpret automatically. Previous work uses the stylistic features of tabular cells (such as font size, border type, and background color) to classify tabular cells by their role in the data layout of the document (top attribute, data, metadata, etc.). In this paper, we propose a deep neural network model which can embed semantic and contextual information about tabular cells in a low-dimensional cell embedding space. We pre-train this cell embedding model on a large corpus of tabular documents from various domains. We then propose a classification technique based on recurrent neural networks (RNNs) to use our pre-trained cell embeddings, combining them with stylistic features introduced in previous work, in order to improve the performance of cell type classification in complex documents. We evaluate the performance of our system on three datasets containing documents with various data layouts, in two settings: in-domain and cross-domain training. Our evaluation result shows that our proposed cell vector representations in combination with our RNN-based classification technique significantly improve cell type classification performance.