Transcript Personalized e-news monitoring agent system

國立雲林科技大學
National Yunlin University of Science and Technology
N.Y.U.S.T.
I. M.
Personalized e-news monitoring agent system for
tracking user-interested Chinese news events
Presenter : Yu-hui Huang
Authors :Chih-Ming Chen · Chao-Yu Liu
AppInt 2009
1
Intelligent Database Systems Lab
Outline

Motivation

Objective

Methodology

Experiments

Conclusion

Comments
N.Y.U.S.T.
I. M.
2
Intelligent Database Systems Lab
Motivation
N.Y.U.S.T.
I. M.

reading or monitoring everyday news stories from the Internet is a
difficult and time-consuming job for modern humans.

But many irrelevant news events are retrieved, resulting in a high
recall rate and low precision rate.
3
Intelligent Database Systems Lab
Objective
N.Y.U.S.T.
I. M.

To improve the performance of Google news alert

Develop a personalized e-news monitoring agent system to tracking
user-interested news based on topic-based scheme.

And develop extension word segmentation system that is ECScanner.
4
Intelligent Database Systems Lab
Methodology

N.Y.U.S.T.
I. M.
System architecture
5
Intelligent Database Systems Lab
Methodology
N.Y.U.S.T.
I. M.

Word segmentation system-ECScanner

Step1：determine new work

Step2： judge new work or not by self-defining threshold
6
Intelligent Database Systems Lab
Methodology

N.Y.U.S.T.
I. M.
The news words management interface for linguistic experts
7
Intelligent Database Systems Lab
Methodology
N.Y.U.S.T.
I. M.

The news event monitoring agent by two-phase scheme for tracking

First phase：cosine measure

Measure similarity between the user-interested and news category

Measure similarity between the user-interested and other news

Ex：user-interested A (書雅，帥哥)=(1,1)
B (書雅，帥哥)=(1,1)
but if A (帥哥，書雅)=(1,1)
x
B (書雅，帥哥)=(1,1)
8
Intelligent Database Systems Lab
Methodology

Hamming distance

Modified cosine measure
N.Y.U.S.T.
I. M.
second phase-filter out misclassified & recommend highest Avg_Msim

9
Intelligent Database Systems Lab
Experiments
N.Y.U.S.T.
I. M.
10
Intelligent Database Systems Lab
Experiments
N.Y.U.S.T.
I. M.
11
Intelligent Database Systems Lab
Experiments
N.Y.U.S.T.
I. M.
12
Intelligent Database Systems Lab
Conclusion
N.Y.U.S.T.
I. M.

Strict threshold value will reduce the number of possible candidate
new words. On the contrary, a loose threshold value will lead to over
large number of candidate new words.

However the ECScanner performance is superior to CKIP
13
Intelligent Database Systems Lab
Comments

Advantage


This paper’s writing step by step is very clear .
Drawback


N.Y.U.S.T.
I. M.
…
Application

News tracking , blog mining , customer behavior analysis…
14
Intelligent Database Systems Lab