How to use long table in latex on overleaf my table is not going on to second page, Its only on one page and truncating

Question

\begin{document}    
\begin{table}[H] 
\caption{Comparative Analysis of most prominent literature work\label{tab1}}
\newcolumntype{x}{>{\centering\arraybackslash\hsize=5\hsize}X}
\newcolumntype{z}{>{\centering\arraybackslash\hsize=10\hsize}X}
\newcolumntype{y}{>{\centering\arraybackslash\hsize=10\hsize}X}

\begin{tabularx}{\textwidth}{xzyy}
\toprule
\textbf{Source} & \textbf{Techniques}   & \textbf{Strength} & \textbf{Weakness}\\
\midrule
Madhu Oruganti 2024 \cite{cite26}    & Curvelet Transform (CLT) and CNN      & Improved     feature similarity, enhanced discriminative power. & Capturing dynamic age-related changes over time.\\  

Yan, Chunman 2024 \cite{cite25}  &  Multi-scale fusion with CNN  & Global and local features & Complex and high computational cost due.\\

Nermeen Nader 2024 \cite{cite24}  & Fusion of Color-Texture and Color Features & Remove Illumination issue. & High computational complexity, extensive preprocessing and normalization steps.\\

Li, Houjie 2024 \cite{cite23} &  Online Re-weighting Relation Network (OR2Net) &  Multi-scale features, Addresses data imbalance & Complex, require significant computational resources and tuning.\\

Peng, Jia Luo 2023 \cite{cite22} & Multi-task Learning with Attention Module and Fairness-aware Contrastive Loss. & Mitigates racial bias, improves accuracy. & High Computation\\

Xu, Min 2023 \cite{cite21}  & Angular Scaling Calibration (ASC) & Implementation, compatibile & Not for all noise. \\

Khammari, M. 2023 \cite{cite20} & MSRCP and GRF & MSRCP and GRF & High computational complexity, preprocssing time. \\

Mahpod, Shahar, 2023 \cite{cite19} &  Multiview hybrid symmetric and asymmetric distance learning network. & Effectively learns. & high computational resources and large amounts of labeled data.\\

Wang, X 2023 \cite{cite18} & Incremental Update Kinship Identification (INK) & Balances privacy and accuracy, adaptable to heterogeneous datasets. & Complex implementation. \\

Pernuš, M 2023 \cite{cite17} & Generative Adversarial Network (GAN) & High resemblance to real parents' children, fine-grained control over synthesis, ability to generate multiple images. & Complexity in training GANs and potential issues with mode collapse.\\

Jiang, Wei 2022 \cite{cite16} & Sample correlation-based genomic relationship matrix (scGRM), UKin & Reduces bias and root mean square error in kinship coefficient estimation. & Complex\\

Liu, Fan 2022 \cite{cite15} & Age-Invariant Adversarial Feature Learning (AIAF) with Identity Feature Weighted module (IFW). & Reduces the impact of age differences. & High implementation complexity and resource requirement.\\

Wu, Xiaoting 2022 \cite{cite14} & Unified Adaptive Adversarial Multimodal Learning (UAAML) & Integrates multiple modalities for better performance. & Complex to implement and train.\\

Almuashi, Mohammed 2022 \cite{cite13} & SCNN-FBOB & Removes weak local blocks to enhance accuracy; utilizes only the most informative regions. & Computationally intensive.\\

Dong, Guan-Nan 2021  \cite{cite10} & Cross-Generation Feature Interaction Learning (CFIL) & Integrates similarity learning and feature extraction. & Complex\\

Yan, Haibin 2021 \cite{cite9} & Deep Relational Network & Local region focus enhances feature extraction. & Increased complexity and computational requirements.\\

Wu, Huishan 2021 \cite{cite11} & Component-based metric learning (CML) and SIFT descriptors. & No need for manual face alignment, improved verification accuracy. & Potentially complex implementation and computationally intensive.\\

Bessaoudi, Mohcene 2021 \cite{cite12} & Tensor Cross-view Quadratic Discriminant Analysis (TCQDA). & Improved accuracy through multi-view representation learning. & Potential complexity in implementation and computational cost.\\

Dahan, Eran 2020 \cite{cite6} & Unified Multi-Task Learning Scheme & Small training sets, prevents overfitting through novel fusion of embeddings, resolves dataset imbalance with adaptive sampling. & Complexity in model design and potential computational overhead.\\

Zhang, Lei 2020 \cite{cite5} & Family ID-based Adversarial Convolutional Network (AdvKin) & Enhanced kinship recognition by incorporating family ID information and adversarial mechanisms. & Complexity in model training due to multiple loss functions and network layers.\\

Dornaika, Fadi 2020 \cite{cite4} & Kinship-oriented Discriminant Data Projection with Multi-Level Fusion & Efficient feature selection. & Computationally intensive. Complexity in merging strategies.\\

Charpentier, Marie 2020 \cite{cite7} & Deep learning for face recognition. & Effective in identifying kinship signals. & Not suitable for all environmental factors influencing resemblance.\\

Brumpton, Ben 2020 \cite{cite8} & Within-family Mendelian randomization. & Reduces biases from familial effects. & Requires large family-based datasets for accurate analysis\\

Moujahid, Abdelmalik 2019 \cite{cite3} & Pyramid Multi-level (PML) Representation with Covariance Descriptor & Explicitly encodes scales and face parts, integrates various texture and color features. & Computationally complex due to covariance calculations across multiple blocks and scales.\\


\bottomrule
\end{tabularx}
\end{table}
\unskip
\end{document}

Answer 1

Some suggestions, in no particular order.

• Whatever else you do, don't use a table environment, since the contents of table environments cannot be broken across pages. Likewise, you can't use a tabularx environment since its contents can't be broken across pages either.

• I suggest you employ an xltabular environment instead, which combines the capabilities of the longtable (span two or more pages) and tabularx (take a target width, e.g., \textwidth) environments. Observe that an xltabular environment is not a "float" (in the LaTeX-specific sense of the term).

• I also wouldn't center-set the cell contents; suppressing full justification (while still allowing hyphenation) would seem to be preferable.

• Do please make an effort to avoid bad line breaks in non-English language words -- such as the given names "Xiaoting" and "Huishan".

The following screenshot shows the first page of the 3-page table that results from implementing these suggestions. (The result also depends on auxiliary assumptions -- such as which document class is in use, or how wide and tall the page is -- that I had to make in order to make the document minimally compilable.)

\documentclass{article} % of some other suitable document class
\usepackage[T1]{fontenc}
\usepackage[english]{babel}
\usepackage{xltabular,ragged2e,booktabs}
\newcolumntype{L}[1]{>{\RaggedRight\hsize=#1\hsize}X}

\hyphenation{xiao-ting hui-shan} % must avoid "xiaot-ing" and "huis-han"

\begin{document}


\begingroup % localize scope of next instruction
\setlength\extrarowheight{2pt} 

\begin{xltabular}{\textwidth}{@{} L{0.7} *{3}{L{1.1}} @{}}
%% Note: 0.7 + 3*1.1 = 4 = # of X-type columns
\caption{Comparative Analysis of most prominent literature work}\label{tab1}\\


%%% headers and footers

\toprule
Source & Techniques & Strength & Weakness \\
\midrule
\endfirsthead

\multicolumn{4}{@{}l}{Table \thetable, continued} \\[1ex]
\toprule
Source & Techniques & Strength & Weakness \\
\midrule
\endhead


\midrule
\multicolumn{4}{r@{}}{\small(continued on next page)}\\
\endfoot

\bottomrule
\endlastfoot


%%% body of table

Madhu Oruganti 2024 \cite{cite26} & Curvelet Transform (CLT) and CNN & Improved feature similarity, enhanced discriminative power. & Capturing dynamic age-related changes over time.\\

Yan, Chunman 2024 \cite{cite25} & Multi-scale fusion with CNN & Global and local features & Complex and high computational cost due.\\

Nermeen Nader 2024 \cite{cite24} & Fusion of Color-Texture and Color Features & Remove Illumination issue. & High computational complexity, extensive preprocessing and normalization steps.\\

Li, Houjie 2024 \cite{cite23} & Online Re-weighting Relation Network (OR2Net) & Multi-scale features, Addresses data imbalance & Complex, require significant computational resources and tuning.\\

Peng, Jia Luo 2023 \cite{cite22} & Multi-task Learning with Attention Module and Fairness-aware Contrastive Loss. & Mitigates racial bias, improves accuracy. & High Computation\\

Xu, Min 2023 \cite{cite21} & Angular Scaling Calibration (ASC) & Implementation, compatibile & Not for all noise. \\

Khammari, M. 2023 \cite{cite20} & MSRCP and GRF & MSRCP and GRF & High computational complexity, preprocssing time. \\

Mahpod, Shahar, 2023 \cite{cite19} & Multiview hybrid symmetric and asymmetric distance learning network. & Effectively learns. & high computational resources and large amounts of labeled data.\\

Wang, X 2023 \cite{cite18} & Incremental Update Kinship Identification (INK) & Balances privacy and accuracy, adaptable to heterogeneous datasets. & Complex implementation. \\

Pernuš, M 2023 \cite{cite17} & Generative Adversarial Network (GAN) & High resemblance to real parents' children, fine-grained control over synthesis, ability to generate multiple images. & Complexity in training GANs and potential issues with mode collapse.\\

Jiang, Wei 2022 \cite{cite16} & Sample correlation-based genomic relationship matrix (scGRM), UKin & Reduces bias and root mean square error in kinship coefficient estimation. & Complex\\

Liu, Fan 2022 \cite{cite15} & Age-Invariant Adversarial Feature Learning (AIAF) with Identity Feature Weighted module (IFW). & Reduces the impact of age differences. & High implementation complexity and resource requirement.\\

Wu, Xiaoting 2022 \cite{cite14} & Unified Adaptive Adversarial Multimodal Learning (UAAML) & Integrates multiple modalities for better performance. & Complex to implement and train.\\

Almuashi, Mohammed 2022 \cite{cite13} & SCNN-FBOB & Removes weak local blocks to enhance accuracy; utilizes only the most informative regions. & Computationally intensive.\\

Dong, Guan-Nan 2021 \cite{cite10} & Cross-Generation Feature Interaction Learning (CFIL) & Integrates similarity learning and feature extraction. & Complex\\

Yan, Haibin 2021 \cite{cite9} & Deep Relational Network & Local region focus enhances feature extraction. & Increased complexity and computational requirements.\\

Wu, Huishan 2021 \cite{cite11} & Component-based metric learning (CML) and SIFT descriptors. & No need for manual face alignment, improved verification accuracy. & Potentially complex implementation and computationally intensive.\\

Bessaoudi, Mohcene 2021 \cite{cite12} & Tensor Cross-view Quadratic Discriminant Analysis (TCQDA). & Improved accuracy through multi-view representation learning. & Potential complexity in implementation and computational cost.\\

Dahan, Eran 2020 \cite{cite6} & Unified Multi-Task Learning Scheme & Small training sets, prevents overfitting through novel fusion of embeddings, resolves dataset imbalance with adaptive sampling. & Complexity in model design and potential computational overhead.\\

Zhang, Lei 2020 \cite{cite5} & Family ID-based Adversarial Convolutional Network (AdvKin) & Enhanced kinship recognition by incorporating family ID information and adversarial mechanisms. & Complexity in model training due to multiple loss functions and network layers.\\

Dornaika, Fadi 2020 \cite{cite4} & Kinship-oriented Discriminant Data Projection with Multi-Level Fusion & Efficient feature selection. & Computationally intensive. Complexity in merging strategies.\\

Charpentier, Marie 2020 \cite{cite7} & Deep learning for face recognition. & Effective in identifying kinship signals. & Not suitable for all environmental factors influencing resemblance.\\

Brumpton, Ben 2020 \cite{cite8} & Within-family Mendelian randomization. & Reduces biases from familial effects. & Requires large family-based datasets for accurate analysis\\

Moujahid, Abdelmalik 2019 \cite{cite3} & Pyramid Multi-level (PML) Representation with Covariance Descriptor & Explicitly encodes scales and face parts, integrates various texture and color features. & Computationally complex due to covariance calculations across multiple blocks and scales.\\

\end{xltabular}
\endgroup

\end{document} 

Answer 2

An alternative way, based on use of the tabularray package, \small font size for better filling texts in cells, in tables and slightly reduced int3erlines spaces:

\documentclass{article} 
\usepackage{geometry}

\usepackage[T1]{fontenc}
\usepackage[english]{babel}
\usepackage{ragged2e}

\usepackage{tabularray}
\UseTblrLibrary{booktabs}

\begin{document}

\begin{longtblr}[
caption = {Comparative Analysis of most prominent literature work},
  label = {tab1},
                ]{colsep  = 3pt,
                  colspec = {@{} X[0.7,l] *{3}{X[1.1, cmd=\RaggedRight]} @{}},
                  cells  = {font=\small\linespread{0.92}\selectfont},
                  rowhead = 1}
    \toprule
Source & Techniques & Strength & Weakness \\
    \midrule
Madhu Oruganti 2024 \cite{cite26} 
    & Curvelet Transform (CLT) and CNN 
        & Improved feature similarity, enhanced discriminative power. 
            & Capturing dynamic age-related changes over time.\\

Yan, Chunman 2024 \cite{cite25} 
    & Multi-scale fusion with CNN 
        & Global and local features 
            & Complex and high computational cost due.\\

Nermeen Nader 2024 \cite{cite24} 
    & Fusion of Color-Texture and Color Features 
        & Remove Illumination issue. 
            & High computational complexity, extensive preprocessing and normalization steps.\\

Li, Houjie 2024 \cite{cite23} 
    & Online Re-weighting Relation Network (OR2Net) 
        & Multi-scale features, Addresses data imbalance 
            & Complex, require significant computational resources and tuning.\\

Peng, Jia Luo 2023 \cite{cite22} 
    & Multi-task Learning with Attention Module and Fairness-aware Contrastive Loss. 
        & Mitigates racial bias, improves accuracy. 
            & High Computation  \\

Xu, Min 2023 \cite{cite21} 
    & Angular Scaling Calibration (ASC) 
        & Implementation, compatibile 
            & Not for all noise. \\

Khammari, M. 2023 \cite{cite20} & MSRCP and GRF & MSRCP and GRF & High computational complexity, preprocssing time. \\

Mahpod, Shahar, 2023 \cite{cite19} 
    & Multiview hybrid symmetric and asymmetric distance learning network. 
        & Effectively learns. 
            & high computational resources and large amounts of labeled data.\\

Wang, X 2023 \cite{cite18} 
    & Incremental Update Kinship Identification (INK) 
        & Balances privacy and accuracy, adaptable to heterogeneous datasets. 
            & Complex implementation. \\

Pernuš, M 2023 \cite{cite17} 
    & Generative Adversarial Network (GAN)  
        & High resemblance to real parents' children, fine-grained control over synthesis, ability to generate multiple images. 
            & Complexity in training GANs and potential issues with mode collapse.\\

Jiang, Wei 2022 \cite{cite16} 
    & Sample correlation-based genomic relationship matrix (scGRM), UKin & Reduces bias and root mean square error in kinship coefficient estimation. 
        & Complex\\

Liu, Fan 2022 \cite{cite15} 
    & Age-Invariant Adversarial Feature Learning (AIAF) with Identity Feature Weighted module (IFW). 
        & Reduces the impact of age differences. 
            & High implementation complexity and resource requirement.\\

Wu, Xiaoting 2022 \cite{cite14} 
    & Unified Adaptive Adversarial Multimodal Learning (UAAML) 
        & Integrates multiple modalities for better performance. 
            & Complex to implement and train.\\

Almuashi, Mohammed 2022 \cite{cite13} 
    & SCNN-FBOB 
        & Removes weak local blocks to enhance accuracy; utilizes only the most informative regions. 
            & Computationally intensive.\\

Dong, Guan-Nan 2021 \cite{cite10} 
    & Cross-Generation Feature Interaction Learning (CFIL) 
        & Integrates similarity learning and feature extraction. 
            & Complex\\

Yan, Haibin 2021 \cite{cite9} 
    & Deep Relational Network 
        & Local region focus enhances feature extraction. 
            & Increased complexity and computational requirements.\\

Wu, Huishan 2021 \cite{cite11} 
    & Component-based metric learning (CML) and SIFT descriptors. 
        & No need for manual face alignment, improved verification accuracy. 
            & Potentially complex implementation and computationally intensive.\\

Bessaoudi, Mohcene 2021 \cite{cite12} 
    & Tensor Cross-view Quadratic Discriminant Analysis (TCQDA). & Improved accuracy through multi-view representation learning. 
        & Potential complexity in implementation and computational cost.\\

Dahan, Eran 2020 \cite{cite6} 
    & Unified Multi-Task Learning Scheme 
        & Small training sets, prevents overfitting through novel fusion of embeddings, resolves dataset imbalance with adaptive sampling. 
            & Complexity in model design and potential computational overhead.\\

Zhang, Lei 2020 \cite{cite5} 
    & Family ID-based Adversarial Convolutional Network (AdvKin) 
        & Enhanced kinship recognition by incorporating family ID information and adversarial mechanisms. 
            & Complexity in model training due to multiple loss functions and network layers.\\

Dornaika, Fadi 2020 \cite{cite4} 
    & Kinship-oriented Discriminant Data Projection with Multi-Level Fusion 
        & Efficient feature selection. 
            & Computationally intensive. Complexity in merging strategies.\\

Charpentier, Marie 2020 \cite{cite7} 
    & Deep learning for face recognition. 
        & Effective in identifying kinship signals. 
            & Not suitable for all environmental factors influencing resemblance.\\

Brumpton, Ben 2020 \cite{cite8} 
    & Within-family Mendelian randomization. 
        & Reduces biases from familial effects. 
            & Requires large family-based datasets for accurate analysis\\

Moujahid, Abdelmalik 2019 \cite{cite3} 
    & Pyramid Multi-level (PML) Representation with Covariance Descriptor 
        & Explicitly encodes scales and face parts, integrates various texture and color features. 
            & Computationally complex due to covariance calculations across multiple blocks and scales.\\
    \bottomrule
\end{longtblr}

\end{document}