chinese commonsense processing
DESCRIPTION
Chinese Commonsense Processing. Presented by Yen-Ling Kuo 2009/3/30. Remember how to build ConceptNet 2…?. The data we collected:. Context Finding Conceptual Analogy Similarity. Dirty Words Filter Swapping List. Link Prediction Add K-Line. Rapport Game Pet Game. subject. subject. - PowerPoint PPT PresentationTRANSCRIPT
1
Chinese Commonsense Processing
Presented by Yen-Ling Kuo
2009/3/30
2
Remember how to build ConceptNet 2…?
• The data we collected:
聖誕節 吃大餐__ 的時候,你會 __
(2, 1, 0)
frequencyfrequency good rankgood rank bad rankbad rank
relationrelation subjectsubjectsubjectsubject
3
Dirty Words Filter
• Idea: Spam filter– Machine learning– Matching/Fuzzy hashing– Black list
• Attribute selection– 近朱者赤,近墨者黑
– Node degree– Ratio of bad rank– # rank– # neighbors in black list– Distance to black list– Confidence of users
Black list White list
Attribute Selection
Classification
Bad subjects
Subjects
Good subjects
4
Link Prediction
• Idea: Social network link prediction
• Application of social network link predication– In Question-Answering Bulletin Board(QABB): 1. Recommend potential answers based on previous communications 2. Predict future hot questions
5
Link Prediction Methods
•Node attribute: not always available• Structural property
Node based topological pattern Node based topological pattern
Path based topological pattern Path based topological pattern
6
Link prediction in Chinese Commonsense?
•Use both node attribute and structural property → Modeled as a supervised learning problem
– Give weights to links according to frequency and good/bad ranks.
– Node attribute
詞類 , relation types
– Structural property
Distance, Weighted common neighbors,Weighted Adamic/Adar,Types of neighbors,Katz
Weighted common neighbors
7
Context Finding
•Determine the context around a concept is useful for building applications.•Context finding is similar to memory search. → Use spreading activation from a source node to get the contextual neighborhood.
Ask 睡覺
枕頭
刷牙浴室
11
0.10.1
0.10.1
0.60.6
0.60.6
0.20.2
0.120.12
※ Different relation
with different weight.
8
Conceptual Analogy
• Employ structure-mapping methods to get a list of structurally analogous concepts given a source concept.
• Structural analogy is not just a measure of semantic distance, ex. “wedding” and “bride”.
貓
貓 有 __
貓 會 __
貓 喜歡 __ 狗
狗 有 __
狗 會 __
狗 喜歡 __
貓 and 狗 are conceptually analogous concepts.
9
Reference
•Hugo Liu, Push Sing. Commonsense Reasoning in and over Natural Language, Lecture Notes in Computer Science, 2004.
•David Liben-Nowell, Jon Kleinberg. The Link Prediction Problem for Social Network, Proceedings of CIKM, 2003.
• Tsuyoshi Murata and Sakiko Moriyasu. Link prediction of social networks based on weighted proximity measures, International Conference on Web Intelligence, 2007