Term Weighting and Ranking Algorithms

Table of Contents

Review

xDocuments in 3D Space

Vector Space Model

xDocuments in Vector Space

xVector Space Documents and Queries

Similarity Measures

xText Clustering

xAgglomerative Clustering

xAgglomerative Clustering

xAgglomerative Clustering

xAutomatic Class Assignment

xPPT Slide

Today

xFinding Out About

Ranking Algorithms

xStructure of an IR System

PPT Slide

Vector Representation (revisited; see Salton article in Science)

Assigning Weights to Terms

Assigning Weights to Terms

xBinary Weights

xRaw Term Weights

Assigning Weights

tf x idf

Inverse Document Frequency

tf x idf normalization

xVector space similarity (use the weights to compare the documents)

xVector Space Similarity Measure combine tf x idf into a similarity measure

xTo Think About

xxComputing Similarity Scores

xComputing a similarity score

xOther Major Ranking Schemes

xOther Major Ranking Schemes

xProbabilistic Models

xProbabilistic Models: Some Notation

xProbabilistic Models

xProbabilistic Models

xLogistic Regression

xProbabilistic Models: Logistic Regression

xLogistic Regression

xProbabilistic Models: Logistic Regression attributes

xProbabilistic Models: Logistic Regression

xSimplified Logistic Regression

xProbabilistic Models

xVector and Probabilistic Models

Email: ray@sherlock.berkeley.edu

Home Page: http://sims.berkeley.edu/~ray

Download presentation source
 

There is much more in the original slides, 
many of the connections do NOT WORK
because I did not downlaoad pages we dont want.
Someof those we will use later.

Dr. R