我使用 longtable，因为我想在文本后嵌入，并让它继续到下一页。另外，我使用 multirow。不幸的是，其中一个单元格的文本奇怪地挂在脚注区域。我的代码：
\documentclass[11pt,titlepage,oneside,openany]{book}
\begin{document}
\begin{longtable}{|l|l|l|l|}
\hline
\begin{tabular}[c]{@{}l@{}}Benchmark\\ Data Set\end{tabular} & \begin{tabular}[c]{@{}l@{}}Accuracy \\ / Error\end{tabular} & Model & Paper \\ \hline
\multirow{ 3}{*}{SST-5} & 56.2 & \begin{tabular}[c]{@{}l@{}}BCN+Suffix \\ BiLSTM-Tied+CoVe\end{tabular} & \begin{tabular}[c]{@{}l@{}}Improved Sentence Modeling \\ using Suffix Bidirectional LSTM\end{tabular} \\ \cline{2-4}
& 55.5 & BERTLARGE & \begin{tabular}[c]{@{}l@{}}Fine-grained Sentiment \\ Classification using BERT\end{tabular} \\ \cline{2-4}
& 54.7 & BCN+ELMo & \begin{tabular}[c]{@{}l@{}}Deep contextualized \\ word representations\end{tabular} \\ \hline
\multirow{3}{*}{\begin{tabular}[c]{@{}l@{}}Yelp\\ Review\\ Full\end{tabular}} & \begin{tabular}[c]{@{}l@{}}Error: \\ 27.80\end{tabular} & XLNet & \begin{tabular}[c]{@{}l@{}}XLNet: Generalized Autoregressive\\ Pretraining for Language Understanding\end{tabular} \\ \cline{2-4}
& \begin{tabular}[c]{@{}l@{}}Error:\\ 28.62\end{tabular} & BERT\_large+ITPT & \begin{tabular}[c]{@{}l@{}}How to Fine-Tune BERT \\ for Text Classification?\end{tabular} \\ \cline{2-4}
& \begin{tabular}[c]{@{}l@{}}Error:\\ 29.32\end{tabular} & BERT\_large & \begin{tabular}[c]{@{}l@{}}Unsupervised Data Augmentation\\ for Consistency Training\end{tabular} \\ \hline
\thead\multirow{3}{*}{\begin{tabular}[c]{@{}l@{}}Amazon \\ Review\\ Full\end{tabular}} & 65.83 & BERT\_large & \begin{tabular}[c]{@{}l@{}}Unsupervised Data Augmentation\\ for Consistency Training\end{tabular} \\ \cline{2-4}
& 65.19 & Deep CNN & \begin{tabular}[c]{@{}l@{}}Deep Pyramid Convolutional Neural \\ Networks for Text Categorization\end{tabular} \\ \cline{2-4}
& 64.43 & Deep RNN & \begin{tabular}[c]{@{}l@{}}Disconnected Recurrent Neural \\ Networks for Text Categorization\end{tabular} \\ \hline
\multirow{3}{*}{\begin{tabular}[c]{@{}l@{}}SemEval \\ 2017 4C\end{tabular}} & \begin{tabular}[c]{@{}l@{}}Error:\\ 0.481\end{tabular} & \begin{tabular}[c]{@{}l@{}}LSTMs+CNNs\\ ensemble with \\ multiple conv. ops\end{tabular} & \begin{tabular}[c]{@{}l@{}}BB twtr at SemEval-2017 Task 4: \\ Twitter Sentiment Analysis \\ with CNNs and LSTMs\end{tabular} \\ \cline{2-4}
& \begin{tabular}[c]{@{}l@{}}Error:\\ 0.555\end{tabular} & \begin{tabular}[c]{@{}l@{}}Deep Bi-LSTM\\ +attention\end{tabular} & \begin{tabular}[c]{@{}l@{}}DataStories at SemEval-2017 Task 4: \\ Deep LSTM with Attention for Message-\\ level and Topic-based Sentiment Analysis\end{tabular} \\ \cline{2-4}
& \begin{tabular}[c]{@{}l@{}}Error:\\ 0.599\end{tabular} & RNTN & \begin{tabular}[c]{@{}l@{}}Amobee at SemEval-2017 Task 4: \\ Deep Learning System for Sentiment \\ Detection on Twitter\end{tabular} \\ \hline
\caption[State of the Art Benchmark Results]{State of the Art Benchmark Results}
\label{fig:related_work}
\end{longtable}
\end{document}
先感谢您