Federated high order tensor fusion for privacy preserving multimodal social media analysis

Li Wan, Bin Zhang · 2026

Read Federated high order tensor fusion for privacy preserving multimodal social media analysis by Li Wan, Bin Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0344980.txt.

Description

Li Wan and Bin Zhang tackle a privacy problem in social media analysis. Posts mix text, images, and sound, and reading them together gives the best sense of public mood, but pooling everyone's data in one place exposes private content. Their 2026 paper proposes a federated system in which each user's data stays on their own device while models still learn from the whole network.

Mathematically heavy tensor methods capture how the different content types relate to each other, improving how well the system reads mixed posts. Tests on a medical abstracts collection and a standard video sentiment benchmark beat earlier approaches. The authors present the design as a route to opinion monitoring that respects user privacy.

Catalog record

CREATORLi Wan, Bin Zhang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE60126
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0344980

Browse the full library · About the Tower of Babel