我已将此表格添加到我的文档中，并且某些单元格中的数据浮动在任何地方，例如第 9 行“优势”和第 14 行“劣势”。
\begin{longtable}{|p{0,6cm}|p{7cm}|p{7cm}|}
\caption{\it{SWOT Analysis}}
\label{tab:data_table}\\ \hline
\textbf{Ref} & \textbf{Strengths} & \textbf{Weaknesses} \\
\hline
\endfirsthead
\multicolumn{3}{l}%
{\tablename\ \thetable\ -- \textit{Continues...}} \\
\hline
\textbf{Ref} & \textbf{Strengths} & \textbf{Weaknesses} \\
\hline
\endhead
\endfoot
\hline
\endlastfoot
\hline
1 & \begin{tabular}[c]{@{}l@{}}Classification of Urdu sentences \\on Document-level, \\Lexicon bases sentiment analysis\end{tabular} & \begin{tabular}[c]{@{}l@{}}No method to tackle implicit negation \\Noun phrases need to be considered\end{tabular} \\
\hline
3 & \begin{tabular}[c]{@{}l@{}}Utilized Long Short-term memory\\ (LSTM) for polarity detection in \\Roman Urdu\end{tabular} & \begin{tabular}[c]{@{}l@{}}No validation of data collection process,\\ no data preprocessing method declared\\ Methods are not transparent \\~\end{tabular} \\
\hline
5 & \begin{tabular}[c]{@{}l@{}}806 Roman Urdu sentences collection, \\features Construction and application\\ on different multilingual classifiers\end{tabular} & \begin{tabular}[c]{@{}l@{}}Limited dataset \\No structure of the dataset\end{tabular} \\
\hline
6 & Lexicon and Rule-based
methods used to construct a RU classification algorithm, ML and phonetic
techniques used & \begin{tabular}[c]{@{}l@{}}Limited categorization of the dataset \\No normalizing of the dataset\end{tabular} \\
\hline
7 & 15,000 roman Urdu
sentences collected & The dataset
contained biographies and not general \\
\hline
8 & \begin{tabular}[c]{@{}l@{}}22,000 sentences of RU were collected, \\Supervised \& unsupervised methods used\end{tabular} & Ambiguous
combination of classifiers \\
\hline
9 & 1200 text documents
of Urdu news were collected, performed a linguistic analysis & \begin{tabular}[c]{@{}l@{}}No character-level features used \\Needs evaluation on state-of-the-art\\ semantic techniques\end{tabular} \\
\hline
11 & Existing
values collated to different techniques & \begin{tabular}[c]{@{}l@{}}No dataset mentioned \\No classification methods mentioned\end{tabular} \\
\hline
12 & A massive
dataset of 5 lac sentiments, Use of lexical classifying techniques & \begin{tabular}[c]{@{}l@{}}Confusing representation of the dataset \\Lack of credible results\end{tabular} \\
\hline
14 & \begin{tabular}[c]{@{}l@{}}1000 reviews collected and Various \\frameworks compared\\ i.e., Hadoop Mapreduce\end{tabular} & Limited
dataset Classifiers were not general and were overfitting on the given
dataset \\
\hline
\end{longtable}
答案1
首先，您可以加载pdflscape包并使用其landscape环境，这实质上会旋转内容并使表格适合页面。
您还可以添加一些额外的空间。作为第一个调整，我会重新定义 \arraystretch为高于1.0以增加行间距。
由于列已定义为p{}，单元格变为常规段落。因此，您可以使用\par和\newline强制换行，并且tabular不再需要 s。此外，\setstretch{<factor>}fromsetspace包还可以更改每个段落中各行之间的间距
通过删除垂直线并添加预定义规则，可以实现不同的布局，在我看来更好booktabs。
希望能帮助到你
第一个截图的代码
\documentclass{article}
\usepackage{setspace}
\usepackage{array}
\usepackage{longtable} % Allows selecting all possible font size
\usepackage{pdflscape}
\newcommand\locstrut{\rule[-9pt]{0pt}{21pt}}
\renewcommand{\arraystretch}{1.5}
\begin{document}
\begin{landscape}
\setstretch{1.17}
\begin{longtable}{|p{0,6cm}|p{7cm}|p{7cm}|}
\caption{\it{SWOT Analysis}}
\label{tab:data_table} \\
\hline
\locstrut \textbf{Ref} & \textbf{Strengths} & \textbf{Weaknesses} \\
\endfirsthead
\multicolumn{3}{l}{\tablename\ \thetable\ -- \textit{Continues...}} \\
\hline
\locstrut \textbf{Ref} & \textbf{Strengths} & \textbf{Weaknesses} \\
\hline \endhead
\hline \endlastfoot
\hline
1 &
Classification of Urdu sentences\par
on Document-level,\par
Lexicon bases sentiment analysis &
No method to tackle implicit negation\par
Noun phrases need to be considered\\
\hline
3 &
Utilized Long Short-term memory\par
(LSTM) for polarity detection in\par
Roman Urdu &
No validation of data collection process,\par
no data preprocessing method declared\par
Methods are not transparent \\
\hline
5 &
806 Roman Urdu sentences collection,\par
features Construction and application\par
on different multilingual classifiers &
Limited dataset\par
No structure of the dataset \\
\hline
6 &
Lexicon and Rule-based methods used to construct a RU classification algorithm, ML and phonetic techniques used &
Limited categorization of the dataset\par
No normalizing of the dataset \\
\hline
7 &
15,000 roman Urdu sentences collected &
The dataset contained biographies and not general \\
\hline
8 &
22,000 sentences of RU were collected,\par
Supervised \& unsupervised methods used &
Ambiguous combination of classifiers \\
\hline
9 &
1200 text documents of Urdu news were collected, performed a linguistic analysis &
No character-level features used\par
Needs evaluation on state-of-the-art\par
semantic techniques \\
\hline
11 &
Existing values collated to different techniques &
No dataset mentioned\par
No classification methods mentioned \\
\hline
12 &
A massive dataset of 5 lac sentiments, Use of lexical classifying techniques &
Confusing representation of the dataset\par
Lack of credible results \\
\hline
14 &
1000 reviews collected and Various\par
frameworks compared\par
i.e., Hadoop Mapreduce &
Limited dataset Classifiers were not general and were overfitting on the given dataset \\
\end{longtable}
\end{landscape}
\end{document}
第二张截图的代码
\documentclass{article}
\usepackage{setspace}
\usepackage{array}
\usepackage{longtable} % Allows selecting all possible font size
\usepackage{booktabs}
\usepackage{pdflscape}
\usepackage[table]{xcolor}
\newcommand\locstrut{\rule[-9pt]{0pt}{21pt}}
\renewcommand{\arraystretch}{1.65}
\begin{document}
\begin{landscape}
\setstretch{1.05}
\begin{longtable}{p{0,6cm}p{7cm}p{7cm}}
\caption{\it{SWOT Analysis}} \label{tab:data_table}\\
\toprule
\locstrut \textbf{Ref} & \textbf{Strengths} & \textbf{Weaknesses} \\
\midrule
\endfirsthead
\multicolumn{3}{l}{\tablename\ \thetable\ -- \textit{Continues...}} \\
\toprule
\locstrut \textbf{Ref} & \textbf{Strengths} & \textbf{Weaknesses} \\
\midrule
\endhead
\midrule \endfoot
\bottomrule \endlastfoot
1 &
Classification of Urdu sentences\newline
on Document-level,\newline
Lexicon bases sentiment analysis &
No method to tackle implicit negation\newline
Noun phrases need to be considered\\
%\hline
3 &
Utilized Long Short-term memory\newline
(LSTM) for polarity detection in\newline
Roman Urdu &
No validation of data collection process,\newline
no data preprocessing method declared\newline
Methods are not transparent \\
%\hline
5 &
806 Roman Urdu sentences collection,\newline
features Construction and application\newline
on different multilingual classifiers &
Limited dataset\newline
No structure of the dataset \\
%\hline
6 &
Lexicon and Rule-based methods used to construct a RU classification algorithm, ML and phonetic techniques used &
Limited categorization of the dataset\newline
No normalizing of the dataset \\
%\hline
7 &
15,000 roman Urdu sentences collected &
The dataset contained biographies and not general \\
%\hline
8 &
22,000 sentences of RU were collected,\newline
Supervised \& unsupervised methods used &
Ambiguous combination of classifiers \\
%\hline
9 &
1200 text documents of Urdu news were collected, performed a linguistic analysis &
No character-level features used\newline
Needs evaluation on state-of-the-art\newline
semantic techniques \\
%\hline
11 &
Existing values collated to different techniques &
No dataset mentioned\newline
No classification methods mentioned \\
%\hline
12 &
A massive dataset of 5 lac sentiments, Use of lexical classifying techniques &
Confusing representation of the dataset\newline
Lack of credible results \\
%\hline
14 &
1000 reviews collected and Various\newline
frameworks compared\newline
i.e., Hadoop Mapreduce &
Limited dataset Classifiers were not general and were overfitting on the given dataset \\
\end{longtable}
\end{landscape}
\end{document}
答案2
- 对于您的表格，遗漏了一些重要的信息，因为使用了
documentclass页面布局，并且如果您真的需要长表格（从代码片段来看，这并不明显。 - 还不清楚为什么你将表格嵌套在单元格中
- 我猜，下面的表格设计应该接近你想要的：
[![在此处输入图片描述][1]][1]
对于此表，我使用包longtblr中定义的，tabularray因为在您的文档中您使用了longtable。但是，如上所示，此代码片段可以放在一页上。
\documentclass{article}
\usepackage{geometry}
\usepackage[english]{babel}
\hyphenation{do-cu-me-nt}
\usepackage{microtype}
\usepackage{tabularray}
\begin{document}
\begin{longtblr}[
caption = {SWOT Analysis},
label = {tab:data_table}
]{hlines, vlines,
colspec = {c *{2}{X[j]}},
colsep = 3pt,
row{1} = {font=\small\bfseries},
rowhead = 1
}
Ref.& Strengths
& Weaknesses \\
1 & Classification of Urdu sentences on Document- level,
Lexicon bases sentiment analysis
& No method to tackle implicit negation
Noun phrases need to be considered \\
3 & Utilized Long Short-term memory
(LSTM) for polarity detection in
Roman Urdu
& No validation of data collection process,
no data preprocessing method declared
Methods are not transparent \\
5 & 806 Roman Urdu sentences collection,
features Construction and application
on different multilingual classifiers
& Limited dataset
No structure of the dataset \\
6 & Lexicon and Rule-based methods used to construct a RU classification algorithm,
ML and phonetic techniques used
& Limited categorization of the dataset
No normalizing of the dataset \\
7 & 15,000 roman Urdu sentences collected
& The dataset contained biographies and not general \\
8 & 22,000 sentences of RU were collected,
Supervised \& unsupervised methods used
& Ambiguous combination of classifiers \\
9 & 1200 text documents of Urdu news were collected, performed a linguistic analysis
& No character-level features used
Needs evaluation on state-of-the-art
semantic techniques \\
11 & Existing values collated to different techniques
& No dataset mentioned
No classification methods mentioned \\
12 & A massive dataset of 5 lac sentiments, Use of lexical classifying techniques
& Confusing representation of the dataset\par
Lack of credible results \\
14 & 1000 reviews collected and Various
frameworks compared i.e., Hadoop Mapreduce
& Limited dataset Classifiers were not general and
were overfitting on the given dataset \\
\end{longtblr}
\end{document}
[1]: https://i.stack.imgur.com/Slihe.png