PhD Scholarships in Computer Science and IT

The School of Computer Science and IT is offering six PhD scholarships to high achieving students for various areas of research. The scholarships are spread across four projects listed in the Further infomation section below.

Value and duration

Scholarships are valued at up to $24,500 per annum for three years (no extension possible). International candidates may be eligible for a tuition waiver.

The project 1 scholarship may also include an additional $5000 for conference travel and a $3000 top-up scholarship per annum depending on the applicant's track record.

Eligibility

To be eligible for this scholarship you must:

  • have an honours/master degree in the area of Computer Science or related discipline
  • meet RMIT’s PhD entry requirements
  • have programming skills in C/C++, Java, Python or Ruby.

Prior research experience is highly desirable.

Applicants should have some prior exposure and interest in one or more of the following areas: information retrieval, databases, algorithms/data structures, natural language processing, and/or data mining.

How to apply

Applications are closed.

Open date

Applications are closed.

Close date

Applications are closed.

Further information

The six PhD scholarships available for PhD research are in the following projects:

Project 1: efficient and effective algorithms for top-k document retrieval.
Supervisor: Shane.
PhD positions available: two.
Summary: the goal of this project is to develop new indexing and query processing algorithms for efficient and effective rank-aware text retrieval. Efficient algorithm design for big data is increasingly important as energy costs continue to soar and can now exceed hardware costs. In this project, two important problems in scalable web search are explored: real-time indexing and long query processing. Possible topic specialisations include suffix-based or inverted indexing, data compression, distributed or parallel text processing, and natural language processing for information retrieval.

Project 2: improving web search using structured and unstructured geospatial information.
Supervisors: Timos Sellis and Shane Culpepper.
PhD positions available: two.
Summary: this project will investigate new approaches to ranked retrieval for location-aware search. Close to 20% of all web search queries contain location information; this number is expected to continue growing as users become increasingly dependent on mobile devices for all of their daily internet activities. We intend to combine state-of-the-art research from two domains: spatial keyword search in databases and ad-hoc search in information retrieval to improve the overall quality of search results.

Project 3: effective summaries for search results.
Supervisors: Mark Sanderson and Falk Scholer.
PhD positions available: one.
Summary: search engines return a search result page, which lists short summaries of each retrieved document. However, recent work has shown that users often fail to click on potentially useful documents due to poor summary quality. This project will take a new approach to the understanding, design, and construction of such summaries. To enhance search result summaries, this project will model how users determine document relevance when inspecting a summary; it will exploit a previously untapped source of information to dramatically improve summary quality; and it will create a new approach to retrieving relevant documents.

Project 4: sub-collection retrieval: understanding and improving search engines.    
Supervisors: Mark Sanderson and Falk Scholer.
PhD positions available: one.
Summary: modern search engines need to find useful answers from vast collections of diverse documents. Currently, a single ranking function is used to identify candidate answers. However, our recent pilot work has shown that using different ranking approaches for different parts of a document collection has the potential to significantly boost search performance. This project will analyse different definitions of sub collections, and study which features of ranking functions lead to different performances on distinct types of documents. This new knowledge will lead to a deeper understanding of search systems, and be used to create new ranking approaches, substantially improving current search techniques, and benefiting all users of such tools.

Contact

For general information, contact Associate Professor Xiaodong Li.

For further information about the specific projects listed above, email the relevant supervisor(s).