Multimodal sentiment analysis model for missing modalities under shared semantic conditions

doi:10.11772/j.issn.1001-9081.2025080959

Journal of Computer Applications ›› 2026, Vol. 46 ›› Issue (9): 2761-2768.DOI: 10.11772/j.issn.1001-9081.2025080959

• Artificial intelligence • Previous Articles

Multimodal sentiment analysis model for missing modalities under shared semantic conditions

Shang LIU(), Zhaosen TANG, Hongyue LIU, Linfang DONG, Jin ZHOU

School of Science and Technology，Tianjin University of Finance and Economics，Tianjin 300222，China

Received:2025-08-22 Revised:2025-10-03 Accepted:2025-10-14 Online:2026-09-16 Published:2026-09-10
Contact: Shang LIU
About author:LIU Shang，born in 1977， Ph. D.， professor. Her research interestsinclude pattern recognition， digital image processing.
TANG Zhaosen， born in 2002， M. S. candidate. His researchinterests include pattern recognition， multimodal perception.
LIU Hongyue， born in 2001， M. S. candidate. Her researchinterests include image processing， pose estimation.
DONG Linfang， born in 1972， Ph. D.， associate professor. Herresearch interests include artificial intelligence， natural language processing.
ZHOU Jin， born in 1981， Ph. D.， associate professor. Her researchinterests include modulation recognition and spectrum sensing based ondeep learning.

共享语义条件下面向模态缺失的多模态情感分析模型

刘赏(), 汤兆森, 刘鸿月, 董林芳, 周金

天津财经大学理工学院，天津 300222

通讯作者: 刘赏
作者简介:刘赏（1977—），女，河北辛集人，教授，博士，主要研究方向：模式识别、数字图像处理
汤兆森（2002—），男，河南濮阳人，硕士研究生，主要研究方向：模式识别、多模态感知
刘鸿月（2001—），女，贵州贵阳人，硕士研究生，主要研究方向：图像处理、姿态估计
董林芳（1972—），女，河北张家口人，副教授，博士，主要研究方向：人工智能、自然语言处理
周金（1981—），女，天津人，副教授，博士，主要研究方向：基于深度学习的调制识别及频谱感知。
基金资助:
天津市自然科学基金资助项目(22JCYBJC01550);天津市艺术科学规划项目(C22030)

Abstract

Abstract:

Existing studies on multimodal sentiment analysis in modality missing scenarios often neglect inter-modal correlations when generating missing modalities， leading to semantic inconsistencies between restored and original data. Additionally， missing modality generation based on diffusion models brings large computational overhead. To address these issues， a Multimodal Sentiment Analysis Model for Missing Modalities under Shared Semantic Conditions （MM-SSC） was proposed. First， a shared latent space mapping module was designed to use the shared latent space of Vector Quantized Variational AutoEncoder （VQ-VAE） to capture multimodal distributions effectively， thereby ensuring cross-modal shared semantics. Second， a cross-modal consistency constraint method was proposed to learn mutual information within each modality's latent space， thereby promoting the refinement of semantic information between modalities and enhancing cross-modal consistency. Third， a missing modality reconstruction and alignment module was designed to reconstruct and refine missing modalities while reducing reconstruction computational overhead. Finally， a multimodal fusion and prediction module was introduced to fuse reconstructed and available modalities for consistent sentiment analysis. Experimental results demonstrate that under fixed modality missing conditions， compared with Incomplete Multimodality-Diffused emotion recognition （IMDer） model， the proposed model achieves average improvements of 0.4 and 0.7 percentage points in the F1 score and ACC7 on the CMU-MOSI dataset， respectively； on the CMU-MOSEI dataset， the F1 score and ACC7 are improved by 0.9 and 0.3 percentage points， respectively. Under random modality missing conditions， compared with IMDer model， the proposed model achieves an average improvement of 2.2 percentage points in both F1 score and ACC7 on the CMU-MOSI dataset； on the CMU-MOSEI dataset， the F1 score and ACC7 （accuracy for seven classes） are improved by 0.8 and 0.4 percentage points， respectively. It can be seen that MM-SSC can address multimodal sentiment analysis tasks under modality missing scenarios effectively.

Key words: multimodal sentiment analysis, missing modality, deep learning, Vector Quantized Variational AutoEncoder (VQ-VAE), diffusion model

摘要：

现有研究在进行模态缺失场景下的多模态情感分析时，在缺失模态生成过程中常忽略模态间的相关性，导致恢复数据与原始数据的语义不一致，同时基于扩散模型的缺失模态生成会带来较大的计算开销。针对以上问题，提出一种共享语义条件下面向模态缺失的多模态情感分析模型（MM-SSC）。首先，设计共享潜空间映射模块，利用矢量量化变分自编码器（VQ-VAE）的共享潜空间有效捕捉多模态分布，确保模态间共享语义；其次，给出跨模态一致性约束方法，以学习每种模态潜空间内的互信息，促进模态间语义信息的完善，增强跨模态一致性；再次，设计缺失模态重建与对齐模块，以重建与细化缺失模态，同时降低重建计算开销；最后，引入多模态融合与预测模块，融合重建模态与可用模态，用于一致性的情感分析。实验结果显示，在固定模态缺失的情况下，相较于不完全多模态扩散情感识别（IMDer）模型，所提模型在CMU-MOSI数据集上的F1得分和七分类准确率（ACC7）指标分别平均提升了0.4与0.7个百分点，在CMU-MOSEI数据集上的F1得分和ACC7分别平均提升了0.9与0.3个百分点；在随机模态缺失的情况下，相比IMDer模型，所提模型在CMU-MOSI数据集上的F1得分和ACC7都平均提升了2.2个百分点，在CMU-MOSEI数据集上的F1得分和ACC7分别平均提升了0.8与0.4个百分点。MM-SSC能有效应对模态缺失场景下的多模态情感分析任务。

关键词: 多模态情感分析, 模态缺失, 深度学习, 矢量量化变分自编码器, 扩散模型

CLC Number:

TP391.4

Shang LIU, Zhaosen TANG, Hongyue LIU, Linfang DONG, Jin ZHOU. Multimodal sentiment analysis model for missing modalities under shared semantic conditions[J]. Journal of Computer Applications, 2026, 46(9): 2761-2768.

刘赏, 汤兆森, 刘鸿月, 董林芳, 周金. 共享语义条件下面向模态缺失的多模态情感分析模型[J]. 《计算机应用》唯一官方网站, 2026, 46(9): 2761-2768.

Figures/Tables 9

Fig. 1 Framework of MM-SSC

Tab. 1 Experimental results on CMU-MOSI dataset with fixed modality missing

模态组合	DCCAE		MCTN		MMIN		GCNet		IMDer		MM-SSC
模态组合	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7
平均	64.8	24.4	69.9	31.3	71.4	31.6	72.8	32.1	75.3	35.5	75.7	36.2
${l}$	76.5	28.3	79.2	41.0	83.8	41.6	83.6	42.3	84.7	44.8	85.7	46.9
${v}$	51.1	17.1	54.4	16.5	54.0	15.5	55.7	16.9	60.8	22.2	61.1	22.5
${a}$	42.1	16.9	54.5	16.5	51.5	15.5	54.5	16.6	62.2	22.0	62.3	22.5
${l, v}$	76.8	30.0	81.2	42.1	83.9	42.0	84.2	43.4	85.4	45.3	85.8	47.7
${l, a}$	77.0	30.2	81.0	43.2	84.0	42.3	84.4	43.4	85.3	45.0	85.7	47.1
${v, a}$	52.5	17.4	57.4	16.8	58.5	19.5	61.9	17.2	63.4	23.8	63.3	18.5
${l, v, a}$	77.4	31.2	81.5	43.4	84.4	44.8	85.1	44.9	85.6	45.3	86.1	48.1

Tab. 1 Experimental results on CMU-MOSI dataset with fixed modality missing

模态组合	DCCAE		MCTN		MMIN		GCNet		IMDer		MM-SSC
模态组合	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7
平均	64.8	24.4	69.9	31.3	71.4	31.6	72.8	32.1	75.3	35.5	75.7	36.2
${l}$	76.5	28.3	79.2	41.0	83.8	41.6	83.6	42.3	84.7	44.8	85.7	46.9
${v}$	51.1	17.1	54.4	16.5	54.0	15.5	55.7	16.9	60.8	22.2	61.1	22.5
${a}$	42.1	16.9	54.5	16.5	51.5	15.5	54.5	16.6	62.2	22.0	62.3	22.5
${l, v}$	76.8	30.0	81.2	42.1	83.9	42.0	84.2	43.4	85.4	45.3	85.8	47.7
${l, a}$	77.0	30.2	81.0	43.2	84.0	42.3	84.4	43.4	85.3	45.0	85.7	47.1
${v, a}$	52.5	17.4	57.4	16.8	58.5	19.5	61.9	17.2	63.4	23.8	63.3	18.5
${l, v, a}$	77.4	31.2	81.5	43.4	84.4	44.8	85.1	44.9	85.6	45.3	86.1	48.1

Tab. 2 Experimental results on CMU-MOSEI dataset with fixed modality missing

模态组合	DCCAE		MCTN		MMIN		GCNet		IMDer		MM-SSC
模态组合	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7
平均	70.2	44.6	72.5	46.8	73.5	47.1	73.7	47.1	75.3	48.5	76.2	48.8
${l}$	79.5	47.0	82.8	50.2	82.4	51.4	83.2	51.2	84.5	52.5	85.4	53.5
${v}$	57.2	40.1	57.1	41.6	60.0	40.7	61.6	41.7	63.6	42.6	64.2	42.1
${a}$	53.8	40.9	54.5	41.4	59.5	40.4	60.3	41.1	60.6	41.7	62.2	41.7
${l, v}$	80.4	47.1	83.2	50.4	83.4	51.2	84.4	51.1	85.0	53.1	85.8	53.8
${l, a}$	80.0	47.4	83.3	50.7	83.3	52.0	84.4	51.3	85.1	53.1	85.5	53.7
${v, a}$	59.2	41.6	62.7	42.1	61.9	41.8	57.2	42.0	63.5	42.8	64.6	42.7
${l, v, a}$	81.2	48.2	84.2	51.2	84.2	52.4	85.1	51.5	85.1	53.4	86.0	53.8

Tab. 2 Experimental results on CMU-MOSEI dataset with fixed modality missing

模态组合	DCCAE		MCTN		MMIN		GCNet		IMDer		MM-SSC
模态组合	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7	F1	ACC7
平均	70.2	44.6	72.5	46.8	73.5	47.1	73.7	47.1	75.3	48.5	76.2	48.8
${l}$	79.5	47.0	82.8	50.2	82.4	51.4	83.2	51.2	84.5	52.5	85.4	53.5
${v}$	57.2	40.1	57.1	41.6	60.0	40.7	61.6	41.7	63.6	42.6	64.2	42.1
${a}$	53.8	40.9	54.5	41.4	59.5	40.4	60.3	41.1	60.6	41.7	62.2	41.7
${l, v}$	80.4	47.1	83.2	50.4	83.4	51.2	84.4	51.1	85.0	53.1	85.8	53.8
${l, a}$	80.0	47.4	83.3	50.7	83.3	52.0	84.4	51.3	85.1	53.1	85.5	53.7
${v, a}$	59.2	41.6	62.7	42.1	61.9	41.8	57.2	42.0	63.5	42.8	64.6	42.7
${l, v, a}$	81.2	48.2	84.2	51.2	84.2	52.4	85.1	51.5	85.1	53.4	86.0	53.8

Tab. 3 Experimental results on CMU-MOSI dataset with random modality missing

缺失率	DCCAE		MCTN		MMIN		GCNet		IMDer		MM-SSC
缺失率	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%
平均	66.3	25.2	70.1	34.8	73.1	35.8	75.1	36.7	79.3	40.5	81.5	42.7
0.0	77.4	31.2	81.5	43.4	84.4	44.8	85.1	44.9	85.6	45.3	86.1	48.1
0.1	74.7	28.1	78.5	39.8	81.8	41.2	82.3	42.1	84.8	44.8	85.7	47.5
0.2	71.9	27.6	75.7	38.5	79.1	38.9	79.5	40.0	83.4	44.3	84.1	46.4
0.3	66.7	25.8	71.2	35.5	76.2	36.9	77.2	38.2	81.0	42.5	83.0	46.1
0.4	62.8	24.2	67.6	32.9	71.6	34.9	74.4	36.6	78.5	39.7	80.1	41.0
0.5	61.3	23.0	64.8	31.2	66.5	32.2	69.8	39.9	75.9	37.9	79.5	39.8
0.6	58.5	20.9	62.5	29.7	64.0	29.1	66.7	29.8	74.0	35.8	77.6	38.0
0.7	57.4	20.6	59.0	27.5	61.0	28.4	65.4	28.1	71.2	33.4	75.5	34.6

Tab. 4 Experimental results on CMU-MOSEI dataset with random modality missing

缺失率	DCCAE		MCTN		MMIN		GCNet		IMDer		MM-SSC
缺失率	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%	F1/%	ACC7/%
平均	71.2	45.0	76.1	46.9	75.4	48.2	76.8	48.3	79.8	50.6	80.6	51.0
0.0	81.2	48.2	84.2	51.2	84.2	52.4	85.1	51.5	85.1	53.4	86.0	53.8
0.1	78.3	46.9	81.6	49.8	81.3	50.6	82.1	51.2	84.6	53.1	85.6	53.5
0.2	75.4	46.3	78.7	48.6	78.8	49.6	79.9	50.2	82.4	52.0	83.5	52.4
0.3	72.2	45.6	76.2	47.4	75.5	48.1	76.8	49.2	80.7	51.3	81.1	51.8
0.4	70.0	44.0	74.1	45.6	72.6	47.5	74.9	48.0	78.1	50.0	80.4	50.4
0.5	66.4	43.3	72.6	45.1	70.7	46.7	73.2	46.7	77.4	49.2	78.1	50.0
0.6	63.2	42.9	71.1	43.8	70.3	45.6	72.1	45.1	75.5	48.5	77.0	47.5
0.7	62.6	42.5	70.5	43.6	69.5	44.8	70.4	44.5	74.6	47.6	73.0	48.3

Tab. 5 Training overhead comparison of models on CMU-MOSI dataset

模型	参数量/10⁶	GPU浮点数运算量/GFLOPs
IMDer	89.79	230.7
MM-SSC	88.46	137.9

Tab. 6 Training time comparison of different models on CMU-MOSI dataset

模型

不同缺失率下的训练时间/s

平均训练

时间/s

Tab. 7 Ablation experimental results on CMU-MOSI and CMU-MOSEI datasets

CMC	SMR	CMU-MOSI		CMU-MOSEI
CMC	SMR	F1/%	MAE	F1/%	MAE
		79.8	0.862	78.7	0.625
√		81.2	0.828	79.5	0.605
	√	80.9	0.806	80.0	0.620
√	√	82.5	0.799	80.4	0.596

Fig. 2 Visualization results on CMU-MOSI dataset

References 23

[1]	夏鑫雨，朱磊，聂秀山，等. 典型概念驱动的模态缺失深度跨模态检索［J］. 计算机辅助设计与图形学学报， 2025， 37（3）： 519-532.
	Xia Xinyu， Zhu Lei， Nie Xiushan， et al. Typical concept-driven modality-missing deep cross-modal retrieval［J］. Journal of Computer-Aided Design and Computer Graphics， 2025， 37（3）： 519-532.
[2]	Ho J， Jain A， Abbeel P. Denoising diffusion probabilistic models［C］// NeurIPS 2020. Red Hook： Curran Associates Inc.， 2020： 6840-6851.
[3]	Van den Oord A， Vinyals O， Kavukcuoglu K. Neural discrete representation learning ［C］// NeurIPS 2017. Red Hook： Curran Associates Inc.， 2017： 6309-6318.
[4]	Zhao J， Li R， Jin Q. Missing modality imagination network for emotion recognition with uncertain missing modalities ［C］// ACL-IJCNLP 2021 （Volume 1： Long Papers）. Stroudsburg： ACL， 2021： 2608-2618.
[5]	Zuo H， Liu R， Zhao J， et al. Exploiting modality-invariant feature for robust multimodal emotion recognition with missing modalities［C］// ICASSP 2023. Piscataway： IEEE， 2023： 1-5.
[6]	Zeng J， Liu T， Zhou J. Tag-assisted multimodal sentiment analysis under uncertain missing modalities ［C］// SIGIR 2022. New York： ACM， 2022： 1545-1554.
[7]	Yuan Z， Li W， Xu H， et al. Transformer-based feature reconstruction network for robust multimodal sentiment analysis［C］// MM 2021. New York： ACM， 2021： 4400-4407.
[8]	Lian Z， Chen L， Sun L， et al. GCNet： graph completion network for incomplete multimodal learning in conversation ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence， 2023， 45（7）： 8419-8432.
[9]	Wang Y， Li Y， Cui Z. Incomplete multimodality-diffused emotion recognition ［C］// NeurIPS 2023. Red Hook： Curran Associates Inc.， 2023： 17117-17128.
[10]	Song Y， Sohl-Dickstein J， Kingma D P， et al. Score-based generative modeling through stochastic differential equations ［PP/OL］. V2. arXiv （2021-02-10）［2025-05-16］..
[11]	罗渊贻，吴锐，刘家锋，等. 面向情感语义不一致的多模态情感分析方法［J］. 计算机研究与发展， 2025， 62（2）： 374-382.
	Luo Yuanyi， Wu Rui， Liu Jiafeng， et al. Multimodal sentiment analysis method for sentimental semantic inconsistency ［J］. Journal of Computer Research and Development， 2025， 62（2）： 374-382.
[12]	Radhakrishnan A， Friedman S F， Khurshid S， et al. Cross-modal autoencoder framework learns holistic representations of cardiovascular state ［J］. Nature Communications， 2023， 14： No.2436.
[13]	Sadok S， Leglaive S， Girin L， et al. A multimodal dynamical variational autoencoder for audiovisual speech representation learning ［J］. Neural Networks， 2024， 172： No.106120.
[14]	Ashuach T， Gabitto M I， Koodli R V， et al. MultiVI： deep generative model for the integration of multimodal data ［J］. Nature Methods， 2023， 20（8）： 1222-1231.
[15]	Wu Y， Song T， Wu Z， et al. CodeBrain： imputing any brain MRI via instance-specific scalar-quantized codes ［PP/OL］. V2. arXiv （2025-03-09）［2025-05-23］..
[16]	Redekop E， Pleasure M， Wang Z， et al. Codebook VQ-VAE approach for prostate cancer diagnosis using multiparametric MRI［C］// CVPRW 2024. Piscataway： IEEE， 2024： 2365-2372.
[17]	Devlin J， Chang M W， Lee K， et al. BERT： pre-training of deep bidirectional Transformers for language understanding ［C］// NAACL-HLT 2019， Volume 1 （Long and Short Papers）. Stroudsburg： ACL， 2019： 4171-4186.
[18]	Stöckli S， Schulte-Mecklenbeck M， Borer S， et al. Facial expression analysis with AFFDEX and FACET： a validation study［J］. Behavior Research Methods， 2018， 50（4）： 1446-1460.
[19]	Degottex G， Kane J， Drugman T， et al. COVAREP： a collaborative voice analysis repository for speech technologies ［C］// ICASSP 2014. Piscataway： IEEE， 2014： 960-964.
[20]	Zadeh A， Zellers R， Pincus E， et al. Multimodal sentiment intensity analysis in videos： facial gestures and verbal messages［J］. IEEE Intelligent Systems， 2016， 31（6）： 82-88.
[21]	Bagher Zadeh A， Liang P P， Poria S， et al. Multimodal language analysis in the wild： CMU-MOSEI dataset and interpretable dynamic fusion graph ［C］// ACL （Volume 1： Long Papers）. Stroudsburg： ACL， 2018： 2236-2246.
[22]	Wang W， Arora R， Livescu K， et al. On deep multi-view representation learning ［J］. Proceedings of Machine Learning Research， 2015， 37： 1083-1092.
[23]	Pham H， Liang P P， Manzini T， et al. Found in translation： learning robust joint representations by cyclic translations between modalities ［J］. Proceedings of the AAAI Conference on Artificial Intelligence， 2019， 33（1）： 6892-6899.

Multimodal sentiment analysis model for missing modalities under shared semantic conditions

共享语义条件下面向模态缺失的多模态情感分析模型

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 9

References 23

Related Articles 15

Recommended Articles

Metrics

[1]	Ming LIU, Dongqi SHEN, Ziyang MENG. Point cloud registration network with dual-branch multi-level feature fusion [J]. Journal of Computer Applications, 2026, 46(8): 2584-2593.
[2]	Fengchun LIU, Xinying SHAO, Chunying ZHANG, Liya WANG, Jing REN. FCMdepth： monocular depth estimation framework with multi-scale feature optimization [J]. Journal of Computer Applications, 2026, 46(8): 2603-2611.
[3]	Wei LIU, Weigang LI, Zhiqiang TIAN. Representation learning method of hierarchical rotation-invariant geometric structure for point cloud classification and segmentation [J]. Journal of Computer Applications, 2026, 46(8): 2594-2602.
[4]	Xinliang LIU, Yushi XU, Dubai LI, Yanzhao REN. Survey of knowledge graph-based question answering methods [J]. Journal of Computer Applications, 2026, 46(8): 2394-2410.
[5]	Xiaojin GUO, Xuyang SUI, Kenan ZHOU. Channel estimation of STAR-RIS-assisted hybrid-field communication system based on deep learning [J]. Journal of Computer Applications, 2026, 46(8): 2548-2554.
[6]	Jing LIU, Shaoze ZHAO, Xingang LIU, Haozhe NIU, Haipeng JI. Cross-condition microstructure data generation method for titanium alloys based on improved CGAN [J]. Journal of Computer Applications, 2026, 46(7): 2373-2382.
[7]	Miaogen LING, Rui JING, Wei FANG. Survey of research on applications of explainable deep learning in tropical cyclone forecasting [J]. Journal of Computer Applications, 2026, 46(7): 2318-2326.
[8]	Xiuli DU, Xing GAO, Xiaoyu ZHANG, Chengsheng PAN, Qijie ZOU. Video snapshot compressive imaging reconstruction method based on dense spatio-temporal deformable attention [J]. Journal of Computer Applications, 2026, 46(7): 2288-2296.
[9]	Xiangyi WU, Hailiang YE, Feilong CAO. Point cloud completion method based on smooth-sharpen graph convolution [J]. Journal of Computer Applications, 2026, 46(7): 2267-2276.
[10]	Xiaohong QIU, Zixiang YU, Yujie HU. Sample augmentation method for fault diagnosis with high-frequency guided diffusion model [J]. Journal of Computer Applications, 2026, 46(7): 2119-2129.
[11]	Yi DU, Mingjin XU, Jiayi KONG, Liyao WANG, Chen ZHAO. Low-rank adaptive parameter-efficient fine-tuning algorithm based on YOLOv11 [J]. Journal of Computer Applications, 2026, 46(6): 1738-1745.
[12]	Zhenkai XIONG, Mengjun XU, Yinyin SUN, Xin WANG. Maritime ship detection algorithm under complex weather environments based on enhanced YOLOv8 [J]. Journal of Computer Applications, 2026, 46(6): 1998-2006.
[13]	Yunping HE, Leichun WANG, Ruirui SONG, Xiangfeng LU, Jinxiang WEI, Xiaomeng LIU. Dual-channel multimodal sentiment analysis model based on contrast invariance and reinforcement specificity [J]. Journal of Computer Applications, 2026, 46(6): 1767-1775.
[14]	Liwan YAO, Hailong LIU, Zhangfan ZENG. Frequency-domain driven and diffusion-based fusion for sonar image enhancement algorithm [J]. Journal of Computer Applications, 2026, 46(6): 1947-1955.
[15]	Lili HE, Meng CAO, Lei ZHANG, Hongjun PAN, Yi LIU, Chengxin SUN. Sign language generation model based on Kolmogorov-Arnold network and diffusion Transformer [J]. Journal of Computer Applications, 2026, 46(6): 1801-1810.