Panda Alert Research Proposal

The Panda Alert System aims to improve the state of the art in early disease outbreak
detection by incorporating linguistic features into the language models which were
previously not known/seen to indicate any risk to human health. The project will
develop a fully functioning, real-time alerting and mapping system.

This kind of research project assumes significant knowledge/use of software
engineering methodologies combined with novel research into unsupervised or
semi-supervised NLP models. We propose a layered, modular architecture
detachable from the core NLP engine, which should demonstrate a universal
detection capability easily transferable to new domains.

The most likely approach would be a semi-supervised, bootstrapping model, which
learns from a small amount of training data to generalise over the unknown domain.
Harnessing the latest NLP machine learning methods such as neural networks (deep
learning), the model analyses news reports to flag any risks to public health. All tools
and data sets from this research will be open sourced and available for download
from public repositories such as GitHub or Google Code.

It is vital for the future reusability of the core NLP engine to be domain-agnostic so
that it can be extended and adapted via a clear programming interface to new event
detection tasks. The real-time analysis system will have to handle hundreds of
thousands of news articles and social media items per day in multiple languages.
The research part of the project will take the state of the art in NLP and AI and
extend it to establish novel methods of NLP general discourse pattern detection in
order to analyse statements for known and unknown features potentially indicative
of the fact-bearing target knowledge. The new technique should be capable of
handling the identification and reporting of other target knowledge/facts from casual
online user/news activity with minimum or no training.

In year one, I shall be mainly focusing on toponym resolution and geoparsing. These
techniques comprise NLP techniques for identification of place names in text and
their subsequent resolution to geographical coordinates. This task is easy for humans
to perform, however for a machine to tell which London (UK, Canada, etc.) was
mentioned in a text is a non-trivial job. I aim to improve on the existing baselines by
researching a novel method of geoparsing. We aim to publish the findings next year
at a respected conference.