Intelligent Automation Phase 2 Forward look Automation Centreeducation
- Slides: 7
Intelligent Automation Phase 2 Forward look Automation. Centre@education. gov. uk
Phase 2 Aims Add a layer of intelligence to our automation Next 18 -24 months Undertake research across government and private sector Raise awareness Train staff Deliver proof-of-concepts Deliver working solutions Automation. Centre@education. gov. uk
Intelligent Automation Question for the group – what is it? We think: Hallmarks of human interactions Ability to converse (ask and respond) in natural language Ability to identify images Ability to learn Automation. Centre@education. gov. uk
Human interactions What we need Tools to help our customers interact with us well Efficiency Understanding needs Reducing internal burden Deep understanding and application of the right methods A joined up approach Good communication What we (probably) don’t need Bots that pass as human (Turing Test) Automation. Centre@education. gov. uk
Natural language processing What is NLP? A set of methods to analyse and manipulate text and speech Based on mathematical concepts Some techniques relatively easy to pick up: Term frequency, inverse-document frequency Others conceptually trickier: Latent Direchlet allocation, latent semantic analysis, supervised learning Being comfortable in R or Python is a standard prerequisite Incredibly clever things are possible, most often seen in companies with big R&D budgets or in academia Automation. Centre@education. gov. uk
Next steps Automation Centre Further development of topic modelling and search bot apps Training in machine learning techniques for deep learning More in-depth research into COTS products Proof-of-concepts in new methods Automation. Centre@education. gov. uk
Questions Automation. Centre@education. gov. uk