Factor semantics and IR using Hebbian learning
Rafael A. Calvo

Basic Information

• Project Title:Factor semantics and IR using Hebbian learning
• Name: Rafael A. Calvo
• Presentation Date: 12/1??

---





Contents

• Abstract
• Proposal and Timelines
• System Description
• Report

Abstract

Principal Component Analysis (PCA) is a useful technique for reducing the dimensionality of datasets, a paramount need in high dimensional term spaces. We study neural networks with Hebbian-like learning that approximately produce the principal components of a document database. The explained variance of the solutions shows how much information the reduced space retains. The retrieval performance on the reduced space is compared with other methods: Smart, LSI and LLSF.

Proposal


Produce for nnn, ntc, ltc:
recall/precsion curves
factors/expvariance

Measure time for each process: smart2hebb, hebb, search
look at time improvements options.
compare with LSI

look at factor semantics.
adaptive learning
scalability
Schedulle

Task	to be done by	status
Set up the web page	October 20th	done
Run Smart, hebb and lsi for nnn,ntc,ltc Ploting results / timing	November 15	done
Analyze Factor semantics (if any).	December 1st	done
Documentation. Write report.	December 10	done

System Description

Draft report