Aggregating Crowd Wisdom with Side Information via a Clustering-based Label-aware Autoencoder

Aggregating Crowd Wisdom with Side Information via a Clustering-based Label-aware Autoencoder

Li'ang Yin, Yunfei Liu, Weinan Zhang, Yong Yu

Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence
Main track. Pages 1542-1548. https://doi.org/10.24963/ijcai.2020/214

Aggregating crowd wisdom infers true labels for objects, from multiple noisy labels provided by various sources. Besides labels from sources, side information such as object features is also introduced to achieve higher inference accuracy. Usually, the learning-from-crowds framework is adopted. However, the framework considers each object in isolation and does not make full use of object features to overcome label noise. In this paper, we propose a clustering-based label-aware autoencoder (CLA) to alleviate label noise. CLA utilizes clusters to gather objects with similar features and exploits clustering to infer true labels, by constructing a novel deep generative process to simultaneously generate object features and source labels from clusters. For model inference, CLA extends the framework of variational autoencoders and utilizes maximizing a posteriori (MAP) estimation, which prevents the model from overfitting and trivial solutions. Experiments on real-world tasks demonstrate the significant improvement of CLA compared with the state-of-the-art aggregation algorithms.
Keywords:
Humans and AI: Human Computation and Crowdsourcing
Machine Learning: Deep Generative Models
Machine Learning: Clustering
Machine Learning: Unsupervised Learning