@inproceedings{5b925cc4880c4e1daf18c7a565ddbde9,
title = "Multimodal data fusion in text-image heterogeneous graph for social media recommendation",
abstract = "Every day, millions of texts, images, audios, videos, and other information with different modalities are posted on social media. These multimodal data provide abundant resources for information recommendation. In this paper, a new method based on multimodal data fusion is proposed for more effective recommendation on social media. Firstly, a heterogeneous graph on texts and images is created effectively to represent the relationship of multimodal data. Then the relationship of multimodal data is fused based on graph clustering to improve the quality of social media recommendation. Finally, the multimodal social media information recommendation is performed as a process of walk on the proposed heterogeneous graph. The experiment on texts and images of microblogs shows social media recommendation using multimodal data fusion is better than that on single modality.",
keywords = "data fusion, graph clustering, heterogeneous graph, multimodal data, recommendation",
author = "Yu Xiong and Daling Wang and Yifei Zhang and Shi Feng and Guoren Wang",
year = "2014",
doi = "10.1007/978-3-319-08010-9\_12",
language = "English",
isbn = "9783319080093",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "96--99",
booktitle = "Web-Age Information Management - 15th International Conference, WAIM 2014, Proceedings",
address = "Germany",
note = "15th International Conference on Web-Age Information Management, WAIM 2014 ; Conference date: 16-06-2014 Through 18-06-2014",
}