Web Search and Text Analysis

This subject has the following teaching availabilities in 2014:

200 hours

One of the following:

None

None

433-460 Human Language Technology
433-467 Text and Document Management
433-660 Human Language Technology
433-667 Text and Document Management
433-476 Text and Document Management

For the purposes of considering request for Reasonable Adjustments under the Disability Standards for Education (Cwth 2005), and Student Support and Engagement Policy, academic requirements for this subject are articulated in the Subject Overview, Learning Outcomes, Assessment and Generic Skills sections of this entry.

It is University policy to take all reasonable steps to minimise the impact of disability upon academic study, and reasonable adjustments will be made to enhance a student's participation in the University's programs. Students who feel their disability may impact on meeting the requirements of this subject are encouraged to discuss this matter with a Faculty Student Adviser and Student Equity and Disability Support: http://services.unimelb.edu.au/disability

AIMS

The aims for this subject is for students to develop an understanding of the main algorithms used in natural language processing and text retrieval, for use in a diverse range of applications including search engines, cross-language information retrieval, machine translation, text mining, question answering, summarisation, and grammar correction. Topics to be covered include text normalisation, sentence boundary detection, part-of-speech tagging, n-gram language modelling, and text classification. The programming language used is Python.

INDICATIVE CONTENT

Topics covered will include:

• Document classification, including gender detection, topic detection and language identification
• Weighted finite state transducers, hidden Markov models
• N-gram language modelling, including statistical estimation
• Sentence segmentation and alignment, the IBM models, expectation maximisation
• Search algorithms including beam search, A* search
• Term indexing, vector space model, term weighting

NTENDED LEARNING OUTCOMES

On completion of this subject the student is expected to:

1. Articulate issues relevant to the efficient implementation of language processing systems and text retrieval systems
2. Apply natural language processing and information retrieval methodologies to textual data
3. Develop and evaluate computational models of language, based on results from the research literature

• Project assignments will be done during the semester and are expected to take approximately 60 hours in total (40%). There are two projects, due around week 6 and week 12.
• A research-oriented workshop presentation (10%)
• One 2-hour end-of-semester examination (50%)

Hurdle requirement: To pass the subject, students must obtain at least:

• 50% overall
• 25/50 in the continuous assessment
• 25/50 in the end-of-semester written examination

Intended Learning Outcomes (ILOs) 1 and 2 are addressed in the lectures, workshops, and exam; ILOs 3 and 4 are addressed in the project work and oral presentation.

This subject is not available as a breadth subject.

On completing this subject, students should have the following skills:

• Formulate and implement algorithmic solutions to computational problems, with reference to the research literature
• Apply a systems approach to complex problems, and design for operational efficiency
• Design, implement and test programs for small and medium size problems in the Python programming language.

LEARNING AND TEACHING METHODS

The subject comprises a weekly 2 hour lecture followed by a 1 hour laboratory exercise. Weekly readings are assigned from the research literature, and weekly laboratory exercises are assigned. Additionally, a significant amount of project work is assigned.

INDICATIVE KEY LEARNING RESOURCES

At the beginning of the semester, the coordinator will post a list of readings from the research literature and research monographs which will form the basis of the intellectual content of the subject. An indicative monograph is Statistical Machine Translation, by Philipp Koehn (2010).

CAREERS / INDUSTRY LINKS

A growing sector of the IT industry is concerned with leveraging the information that is locked up in semi-structured text data on the web. Large scale analysis and exploitation of this information depends on graduates with a solid grounding in natural language processing and text retrieval algorithms, and experience with implementing systems that are informed by the research literature.