Advanced Topics in Information Retrieval
Spring 2008
This is largely a placeholder at the moment. For the nonce, here are
some general outline thoughts for the class:
- Weekly meeting, Fridays, 2:30-4:30, room 140.
- Read papers on topics in IR
- Present for at least one session (depending on enrollment)
- Auditors must read papers but will not be required to present
- SPIRE schedule number 82296.
- 3 credits
Friday, February 8. Passages/XML
James presenting
We consider the topic of passage retrieval and XML element retrieval
for this class. The general idea is that there are situations in
which a portion of a document is more useful than the entire thing.
In passage retrieval, a key question is finding the portions. In XML
retrieval, the XML markup elements may provide the right portions.
- (Required) Kaszkiel and Zobel, "Passage retrieval revisited", SIGIR 1997 (9
pages, at ACM).
This paper evaluates a range of passage retrieval approaches. It
builds on a classic SIGIR 1994 paper
by Callan.
- (At least the poster) Jiang and Zhai, "Extraction of coherent relevant passages using
hidden
Markov models". ACM Transactions on Information Systems,
24(3):295-319. (25 pages, at ACM).
This
paper talks about a way to extract just the right length of passage.
A sketch of the work is available in a poster presented at SIGIR:
-
Jiang and Zhai, "Accurately extracting coherent relevant passages
using hidden Markov models", (2 page poster summary, at ACM)
- (Skim at least) Fuhr et al, "Overview of the INEX 2007 Ad Hoc Track".
Pre-proceedings
of INEX 2007. (22 pages, pdf).
Provides an overview of retrieval using XML elements, much of which is
summary evaluation results. You may find
this description of the query format useful:
- Trotman and Sigurbjornsson, "Narrowed Extended XPath I (NEXI)".
INEX
2004 workshop proceedings.
Defines the query language used for structured queries in INEX. Worth
skimming to figure out the queries. (11 pages, pdf)
- Ogilvie and Callan, "Hierarchical language models for XML
component
retrieval". In Proceedings of the INEX 2005 workshop. (15 pages, pdf).
This paper sketches how element retrieval in XML can be done in a
language modeling framework.
- (Required) Kams and Koolen, "On the relation between relevant passages and
XML
document structure." Proceedings of the SIGIR 2007 workshop on focused
retrieval. (5 pages, pdf,
pdf slides,
full workshop proceedings). This paper links
XML element retrieval and passage retrieval: is there a relationship
at all?
-
Itakura and Clarke, "From passages into elements in XML
retrieval". Proceedings of the SIGIR 2007 workshop on focused
retrieval. (6 pages, pdf, proceedings).
How should those relevant XML elements be found?
Friday, February 15. Learning to rank
Ben presenting