(Enter summary)
Abstract: Language models are used in various fields that deal with sequences of words. We present in this paper automatic methods to cluster sets of documents into topic trees (that go from general to specific). Conditional entropy is used to evaluate the clustering methods. Then we use the resulting topic trees to estimate topic-sensitive language models, that can better capture words dependencies. When computing language model probabilities, dynamic topic modeling is used. (Update)
Similar documents based on text: More All
0.4: Dynamic Nonlocal Language Modeling via Hierarchical.. - Florian, Yarowsky (1999)
(Correct)
0.3: Dynamic Nonlocal Language Modeling via - Hierarchical Topic-Based Adaptation
(Correct)
0.1: Deriving Phrase-based Language Models - Heeman, Damnati (1997)
(Correct)
BibTeX entry: (Update)
R. Florian, "Exploiting Nonlocal and Syntactic Word Relationships in Language Models," Ph.D. Qualifying Project Report, CS Dept, Johns Hopkins University, Aug., 1998. http://citeseer.ist.psu.edu/florian98exploiting.html More
@misc{ florian98exploiting,
author = "R. Florian",
title = "Exploiting Nonlocal and Syntactic Word Relationships in Language Models",
text = "Ph.D. Qualifying Project Report, CS Dept, Johns Hopkins University,
Aug., 1998.",
year = "1998",
url = "citeseer.ist.psu.edu/florian98exploiting.html" }
Citations (may not include all citations):
805
Algorithms for Clustering Data (context) - Jain, Dubes - 1988
274
Estimation of probabilities from sparse data for the languag.. (context) - Katz
202
Statistical Methods for Speech Recognition (context) - Jelinek - 1997
2
Hierarchical topic-sensitive language models for automatic s..
- Mangu - 1997
1
The population of species and the estimation of population p.. (context) - Good
1
Using stopy topics for language model adaptation (context) - Seymour, Rosenfeld
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC