Imperial College London Engagement Goal

Recent advances in Deep Content Learning helps support data-driven research at the Data Science Institute at Imperial College London

Use information gathered from academic abstracts to create a reference of individual topics in a Wikipedia type format, in a quick and easy to use format. The main focus is a general framework called Deep Content Learning using two different example datasets. In the traffic prediction project, a new large-scale traffic dataset was released with auxiliary information including search queries from the Baidu Map app. Proposed were hybrid models to achieve state-of-the-art prediction accuracy. The second project involved zero-shot text classification which integrated semantic knowledge and used a two-phase architecture to tackle the challenging zero-shot learning in textual data. In parallel with this research is also their work on TensorFlow. HPCC Systems is perfectly placed to handle the data processing side and delivery of the data. The idea is to use a Thor cluster for the data cleaning, normalization and linking and a ROXIE cluster for the fast delivery of the data to the user. TensorFlow fits snugly in the middle of this workflow, providing the modeling based mainly on recurrent neural networks.

Key Contact

Jingqing Zhang

PhD Candidate

Yike Guo

Professor, Computing Science

Articles, Blogs, Papers & Presentations

Article

How deep learning could give doctors a helping hand

Blog

Update – Integrating Prior Knowledge with Learning in Biomedical Natural Language Processing

Blog

Integrating Prior Knowledge with Learning in Biomedical Natural Language Processing

Tech Talk

Integrating Prior Knowledge with Learning in Biomedical Natural Language Processing

Blog

Enhanced Interpretability in Question/Answering Systems

Tech Talk

Deep Sequence Learning in Traffic Prediction and Text Classification

Paper

Deep Sequence Learning with Auxiliary Information for Traffic Prediction

Presentation

Deep Content Learning in Traffic Prediction and Text Classification

Paper

Dest-ResNet: a Deep Spatiotemporal Residual Network for Hotspot Traffic Speed Prediction

Paper

Integrating Semantic Knowledge to Tackle Zero-shot Text Classification

Paper

Learning Text to Image Synthesis with Textual Data Augmentation

Paper

TensorDB: Database Infrastructure for Continuous MachineLearning

Paper

The Deep Poincaré Map: A Novel Approach forLeft Ventricle Segmentation

Article

How Researchers are Using NLP and Machine Learning to Ease your Information Overload

Presentation Slides

Towards Trustable AI for Complex Systems: a Life Science and AIOps Perspective