Logistics Big Data Analytics

Program Code: 
Description of Work: 


The research will carry out data analytics, finding correlations and pattern recognitions through intelligent computational technologies, various statistical analytic methods and software that turn the large complex data set into meaningful information and knowledge. In here, we study and conduct research into the following topics: Complex data structures including trees, graphs, text and spatial-temporal data; Mixed information types including image, video, web and text; Data that is distributed across multiple data sources; Joint mining of structured, semi-structured and unstructured information; Rare events whose significance could be masked by more frequently occurring events; Data, text and content mining; Web mining and document management; Knowledge discovery, representation and knowledge mining; Integrating a knowledge base from a data mining system and applying this knowledge during the data mining; Integrating a wide range of data mining techniques and methods and deriving incremental new knowledge from large data sets and prior knowledge. The types of knowledge extracted from information mining activities include: Embedded structures and relationships leading to associations between these embedded structures and embedded trees rather than just between simple variables; and also Knowledge that conforms to a certain model structure to enhance the model. The sources of information addressed in such mining activities include: Information gathered by a corporation or enterprise about its clients; Information provided by customers or viewers as a result of their own choice such as product reviews and trustworthiness information. Such information is of considerable importance in trust, reputation and risk assessment systems; and Information on social networking sites, to permit opinion mining.

Description of Work:

This project involves design and development of algorithms, information analysis techniques, good mathematical knowledge as well as numerical modelling.


Prof Elizabeth Chang (e.chang@adfa.edu.au)