COMP9314

Lectures

CSE@UNSW

Week	Topics
1	Introduction Keyword Search in Relational Databases (1) (6-in-1) <updated on 1 Mar> Readings: Databases and IR: Perspectives of a SQL Guy: a talk given by a senior Microsoft researcher, high-level overview, less technical (esp. the 1st half) Proximity Search in Databases: VLDB 99 paper from Stanford. Read it after you want to compare your solution to the homework question (find movies near 'cage' and 'travolta') with theirs.
2	Keyword Search in Relational Databases (2) (6-in-1) Readings: Keyword Searching and Browsing in Databases using BANKS: 1st keyword search on RDBMS system based on the graph model. The algorithm part should be easy once you refresh your knowledge about the Dijkstra algorithm. Bidirectional Expansion For Keyword Search on Graph Databases: Little bit involved in the algorithm part.
3	Keyword Search in Relational Databases (3) (6-in-1) Readings: DISCOVER: Keyword Search in Relational Databases: Pay attention to the system architecture and CN generation algorithm. Efficient IR-Style Keyword Search over Relational Databases: Read the ranking method and the four query processing algorithms. It is the first paper in this area that has reasonable ranking method and optimization for efficient top-k retrieval.
4	Keyword Search in Relational Databases (4) (6-in-1) Readings: SPARK: Top-k Keyword Query in Relational Databases: Check out our latest work on the topic, :)
5	Multimedia Information Retrieval (1) (4-in-1) Readings: Reference books [1] - [4]. Key book: Multimedia database management systems Color Indexing: using Histogram Intersection for object identification and Histogram Backprojection for object location. The QBIC Project Querying Images By Content Using Color, Texture, and Shape: The QBIC Project-Querying Images By Content Using Colour, Texture, and Shape: The original paper for IBM QBIC project. Colour spaces - perceptual, historical and applicational background: An overview of colour spaces used in image processing.
6	Multimedia Information Retrieval (2) (4-in-1) Readings: Reference books [1] - [4]. Key book: Multimedia database management systems Color Indexing: using Histogram Intersection for object identification and Histogram Backprojection for object location. The QBIC Project Querying Images By Content Using Color, Texture, and Shape: The QBIC Project-Querying Images By Content Using Colour, Texture, and Shape: The original paper for IBM QBIC project. Colour spaces - perceptual, historical and applicational background: An overview of colour spaces used in image processing.
7	Multimedia Information Retrieval (3) (4-in-1) Readings: Reference books [1] - [4]. Key book: Multimedia database management systems Color Indexing: using Histogram Intersection for object identification and Histogram Backprojection for object location. The QBIC Project Querying Images By Content Using Color, Texture, and Shape: The QBIC Project-Querying Images By Content Using Colour, Texture, and Shape: The original paper for IBM QBIC project. Colour spaces - perceptual, historical and applicational background: An overview of colour spaces used in image processing.
8	Data Stream Computation(1) (6-in-1) Data Stream Computation(2) (6-in-1) Readings: Models and Issues in Data Stream Systems A survey paper for the data stream systems. Gigascope: A Stream Database for Network Applications Introduction of the Gigascope system. The space complexity of approximating the frequency moments Paper for the frequent moment estimation. Probabilistic Counting Algorithms for Data Base Applications Paper for FM algorithm. An Improved Data Stream Summary: The Count-Min Sketch and its Applications Paper for Count Min algorithm.
9	Data Stream Computation(3 : Heavy Hitter ) (6-in-1) Readings: A simple algorithm for finding frequent elements in streams and bags the frequent elements computation Approximate Frequency Counts over Data Streams the frequent elements computation
10	Data Stream Computation(4 : Order Statistics ) (6-in-1) Readings: Space Efficient Online Computation of Quantile Summaries the GK algorithm paper
11	Data Stream Computation(5 : Skyline Computation) (6-in-1) Readings: Stabbing the Sky: Efficient Skyline Computation over Sliding Windows Skyline Computation over Sliding Windows
12	XML (6-in-1) Readings: Relational Databases for Querying XML Documents: Limitations and Opportunities Querying and Maintaining a Compact XML Storage Other references from W3C etc will be discussed in the lecture
13	XML Part 2 (6-in-1) Readings: Update Synchronization for Mobile XML Data