Dyer |
JOHN |
·
Enjoyed most of collaborative and
content based filtering. Complimentary to each other, using both attributes
and what users rated them. ·
Liked LSI, dim
reduction, principal component
analysis, etc ·
Intrigued by how patterns can emerge in the
data and use them to draw conclusions. |
COMMISSO |
STEPHEN |
·
LSI was favorite,
use of linear algebra, dimensionality reduction. How it affects information
retrieval, synonymy, etc. ·
Automatically
clustering. ·
Collaborative Filtering,
how it actually works for amazon. |
|
BRYCE |
·
Surprised how a
few matrix calculations can capture some of the semantic meaning of a text
document. Reminds him of a friend that said that CS was based on math.
Definitely true now. o Class shows cutting edge ways that you can solve problems
with math. ·
The way google used link structure. Seems
that you should always get your hands on all data you can. ·
Fact that
clustering can be done automatically. |
CIESIELSKI |
|
·
A lot of the
ideas are simple, but when web is brought in, then they can get much more complicated
for scale issues. o
Even in project
it showed up. ·
Also interesting
to note how many “fudge factors” are there in Google. |
LIPP |
JOHN |
·
Originally didn't
like LSI at all. Was one of those who slammed LSI in the first survey. But then he got matlab and sby using it, it started
making sense in and got to like it. o Also LSI, used eigenvectors for image retrieval –a la
Eigen Faces (by Kanav Kahol in the multi-media class) |
HIRST |
|
·
Enjoyed reading
academic papers. It's missing when not reading those papers in other classes.
·
Figured out how
bayes classification works for spam email software. ·
Preferred IR
topics to II topics |
SELLERS |
MICHAEL |
------AWOL-------------- |
|
CHRISTIAN |
·
Interesting to
actually see some XML and use it. Used to be just a buzz word. Same with
Xquery. |
MCFADDEN |
MICHAEL |
·
Got to use math
he learned. Got some meaning. Not just formulas and just crunching in
numbers. Actual applications for the math. Also collaborative filtering. |
FATNANI |
NIKHIL |
·
Regular
communication between instructor and students. Website and mails. ·
Projects helpful
to understand. |
BALONEK |
BENJAMIN |
·
Professor was
enthusiastic. In contrast to some other classes. Maybe just only 494 classes
are like this. ·
Liked the emails sent on
the class list . Made him want to come to class. ·
Had no idea how Google worked. Now
understands. |
KALE |
SHREYAS |
·
How you can take
basic text, categorize into matrices, and compute similarities, connection
b/w docs, all this just mathematically. ·
Liked google paper, google-lecture. ·
Noticed how
google kept on adding things, especially things that we said should be there.
·
Also yahoo paper
(on using LSI clustering). ·
Structured data, xml because didn't know xml
at all. |
MARINICK |
JOHN |
Absent |
|
|
|
|
|
|
FAN |
JIANCHUN |
·
Using simple
mathematical models of text turns out to be working pretty well. ·
Cutting edge topics of the class. Especially
last classes on combining DB and IR to help people find info more easily. |
SINGHI |
SURENDRA |
·
Analysis in
projects. Liked how ideas are simple but powerful in practice. |
VADREVU |
SRINIVAS |
·
Liked power
method for eigenvectors. ·
Discussion on Sabre
database. Used to think expedia etc were already doing information
integration. ·
Tyranny of
majority. ·
DB & IR link. |
ZHAO |
JICHENG |
·
Collaborative
and, content based filtering. Especially on how to combine them with clustering. ·
Also clustering results of a search engine. ·
Topic specific
pagerank. |
BHIMAVARAPU |
|
·
LSI was most
appealing. ·
Page Rank, structure
of the web. ·
Content boosted
collaborative filtering. ·
Database refresher
by Ullas. ·
Thought first
half was more interesting. |
RYAN |
JOSHUA |
·
LSI and Page
Rank. ·
Handling
multi-dimensional vectors(?) ·
Liked the
projects. Best in-class projects of ASU. |
|
J |
·
Xqueries, and
see how it works. o
Even the fact
that it existed. ·
Projects were
great help in understanding what was done in class. |
ASWATH |
DIPTI |
·
Clustering
lectures. ·
DB and IR
techniques. ·
Support of
imprecise queries. |
JANAKIRAMAN |
SHIVASHANKARI |
·
The fact that we
focused on how new ideas could be applied. o
Rather than just on what has
been done. ·
Like the slides a
lot. ·
Projects
required a lot of help and were a lot of fun. |