knowledge_based topic model - 基于知识的主题模型概述
2017-08-14 00:00
465 查看
http://blog.csdn.net/pipisorry/article/details/44040701
LDA (Bleiet al., 2003)
术语
Mustlink states that two words should belong to the same topic
Cannot-link states that two words should not belong to the same topic.
DF-LDA[Andrzejewski, David, Zhu, Xiaojin, and Craven, Mark. Incorporating domain knowledge into topic modeling via Dirichlet Forest priors. In ICML, pp. 2009.]
is perhaps the earliest KBTM, which can incorporate two forms of prior knowledge from the user:must-links and cannot-links.
DF-LDA [1]: A knowledge-based topic model that can use both must-links and cannot-links, but itassumes all the knowledge is correct.
DF-LDA [1] cannot distinguish multiple senses because its de nition of must-link is transitive.That is, if terms w1 and w2 form a must-link, and terms w2 and w3 form a must-link, it implies a must-link between w1 and w3, i.e., w1, w2, and w3 should be in the same topic. We call it thetransitivity problem.
http://blog.csdn.net/pipisorry
GK-LDA (Chen et al., 2013a)
A knowledge-based topic model that uses the ratio of word probabilities under each topic to reduce the effect of wrong knowledge. However, it can only use the must-link type of knowledge.
MC-LDA (Chen et al., EMNLP 2013)
is a recent knowledge-based model for aspect extraction.
A knowledge-based topic model that also use both the must-link and the cannot-link knowledge. It assumes that all knowledge is correct as well.
MC-LDA [10] assumes each must-link represents a distinct sense, and thus assigns each term only one relevant must-link and ignores the rest. This misses a lot of good must-links.
Note:although both DF-LDA and GK-LDA can take prior knowledge from the user, they cannot mine any prior knowledge.
Note:
Both of DF-LDA && MC-LDA assume the knowledge to be correct and manually provided, both of them assume that the user-provided must-links and cannot-links are correct and there is no conflict among them.
AKL (Automated Knowledge LDA),(Chen et al., 2014: Aspect Extraction with Automated Prior Knowledge Learning):
A knowledge-based topic model that applies clustering to learn the knowledge and utilizes the knowledge in the form of knowledge clusters.
whose inference can exploit the automatically learned prior knowledge and handle the issues of incorrect knowledge to produce superior aspects.
LTM [7]:
A lifelong learning topic model that learns only the must-link type of knowledge automatically. It outperformed [8].
from:http://blog.csdn.net/pipisorry/article/details/44040701
LDA (Bleiet al., 2003)
术语
Mustlink states that two words should belong to the same topic
Cannot-link states that two words should not belong to the same topic.
DF-LDA[Andrzejewski, David, Zhu, Xiaojin, and Craven, Mark. Incorporating domain knowledge into topic modeling via Dirichlet Forest priors. In ICML, pp. 2009.]
is perhaps the earliest KBTM, which can incorporate two forms of prior knowledge from the user:must-links and cannot-links.
DF-LDA [1]: A knowledge-based topic model that can use both must-links and cannot-links, but itassumes all the knowledge is correct.
DF-LDA [1] cannot distinguish multiple senses because its de nition of must-link is transitive.That is, if terms w1 and w2 form a must-link, and terms w2 and w3 form a must-link, it implies a must-link between w1 and w3, i.e., w1, w2, and w3 should be in the same topic. We call it thetransitivity problem.
http://blog.csdn.net/pipisorry
GK-LDA (Chen et al., 2013a)
A knowledge-based topic model that uses the ratio of word probabilities under each topic to reduce the effect of wrong knowledge. However, it can only use the must-link type of knowledge.
MC-LDA (Chen et al., EMNLP 2013)
is a recent knowledge-based model for aspect extraction.
A knowledge-based topic model that also use both the must-link and the cannot-link knowledge. It assumes that all knowledge is correct as well.
MC-LDA [10] assumes each must-link represents a distinct sense, and thus assigns each term only one relevant must-link and ignores the rest. This misses a lot of good must-links.
Note:although both DF-LDA and GK-LDA can take prior knowledge from the user, they cannot mine any prior knowledge.
Note:
Both of DF-LDA && MC-LDA assume the knowledge to be correct and manually provided, both of them assume that the user-provided must-links and cannot-links are correct and there is no conflict among them.
AKL (Automated Knowledge LDA),(Chen et al., 2014: Aspect Extraction with Automated Prior Knowledge Learning):
A knowledge-based topic model that applies clustering to learn the knowledge and utilizes the knowledge in the form of knowledge clusters.
whose inference can exploit the automatically learned prior knowledge and handle the issues of incorrect knowledge to produce superior aspects.
LTM [7]:
A lifelong learning topic model that learns only the must-link type of knowledge automatically. It outperformed [8].
from:http://blog.csdn.net/pipisorry/article/details/44040701
相关文章推荐
- knowledge_based topic model - 基于知识的主题模型概述
- TopicModel主题模型 - EM算法
- knowledge_based topic model - AMC
- 从主题模型(Topic Model)到隐语义模型(Latent Factor Model)
- [学习笔记]学习主题模型(Topic Model)和PLSA( probabilistic latent semantic analysis)
- 主题模型(topic model)的小survey
- 基于知识的图文档建模(Knowledge-based Graph Document Modeling)
- 基于记忆(Memory-Based)与基于模型(Model-Based)的辨析
- Notes:An Ontology-based Information Retrieval Model--基于本体的信息检索模型
- TopicModel主题模型 - LDA主题模型的评估
- 主题模型(topic model)的提出及发展历史
- knowledge_based topic model - AMC
- 主题模型(topic model)的提出及发展历史
- LDA主题模型用于BUG修复人推荐《DRETOM: developer recommendation based on topic models for bug resolution》
- TopicModel主题模型 - LDA的编程实现
- [学习笔记]学习主题模型(Topic Model)和PLSA( probabilistic latent semantic analysis)
- #Paper Reading# Online Knowledge-Based Model for Big Data Topic Extraction
- [学习笔记]学习主题模型(Topic Model)和PLSA( probabilistic latent semantic analysis)
- [学习笔记]主题模型(Topic Model)和PLSA( probabilistic latent semantic analysis)
- 主题模型(Topic Model)相关学习资料与文献