Please use this identifier to cite or link to this item: https://scholarbank.nus.edu.sg/handle/10635/38930
DC FieldValue
dc.titleSemi-supervised text classification using partitioned EM
dc.contributor.authorCong, G.
dc.contributor.authorLee, W.S.
dc.contributor.authorWu, H.
dc.contributor.authorLiu, B.
dc.date.accessioned2013-07-04T07:30:07Z
dc.date.available2013-07-04T07:30:07Z
dc.date.issued2004
dc.identifier.citationCong, G.,Lee, W.S.,Wu, H.,Liu, B. (2004). Semi-supervised text classification using partitioned EM. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2973 : 482-493. ScholarBank@NUS Repository.
dc.identifier.issn03029743
dc.identifier.urihttp://scholarbank.nus.edu.sg/handle/10635/38930
dc.description.abstractText classification using a small labeled set and a large unlabeled data is seen as a promising technique to reduce the labor-intensive and time consuming effort of labeling training data in order to build accurate classifiers since unlabeled data is easy to get from the Web. In [16] it has been demonstrated that an unlabeled set improves classification accuracy significantly with only a small labeled training set. However, the Bayesian method used in [16] assumes that text documents are generated from a mixture model and there is a one-to-one correspondence between the mixture components and the classes. This may not be the case in many applications. In many real-life applications, a class may cover documents from many different topics, which violates the one-to-one correspondence assumption. In such cases, the resulting classifiers can be quite poor. In this paper, we propose a clustering based partitioning technique to solve the problem. This method first partitions the training documents in a hierarchical fashion using hard clustering. After running the expectation maximization (EM) algorithm in each partition, it prunes the tree using the labeled data. The remaining tree nodes or partitions are likely to satisfy the one-to-one correspondence condition. Extensive experiments demonstrate that this method is able to achieve a dramatic gain in classification performance. © Springer-Verlag 2004.
dc.sourceScopus
dc.subjectLabeled and unlabeled data
dc.subjectSemi-supervised learning
dc.subjectText classification
dc.subjectText mining
dc.typeArticle
dc.contributor.departmentCOMPUTER SCIENCE
dc.description.sourcetitleLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
dc.description.volume2973
dc.description.page482-493
dc.identifier.isiutNOT_IN_WOS
Appears in Collections:Staff Publications

Show simple item record
Files in This Item:
There are no files associated with this item.

Page view(s)

101
checked on Aug 19, 2019

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.