TY  - CONF
ID  - SisLab3633
UR  - https://eprints.uet.vnu.edu.vn/eprints/id/eprint/3633/
A1  - Pham, Thi Ngan
A1  - Ha, Quang Thuy
A1  - Nguyen, Minh Chau
A1  - Nguyen, Tri Thanh
Y1  - 2019///
N2  - Lifelong machine learning has recently become a hot topic attracting the re-searchers all over the world by its effectiveness in dealing with current problem by exploiting the past knowledge. The combination of topic modeling on pre-vious domain knowledge (such as topic modeling with Automatically generat-ed Must-links and Cannot-links, which exploits must-link and cannot-link of two terms), and lifelong topic modeling (which employs the modeling of previous tasks) is widely used to produce better topics. This paper proposes a close domain metric based on probability to choose reliable (prior) knowledge learnt from the past to generate more coherent topics on the current domain. This knowledge is, then,  used to enrich features for multi-label classifier. Several experiments performed on review dataset of hotel show that the proposed approach leads to an improvement in performance over the baseline.
TI  - A probability-based close domain metric in lifelong learning for multi-label classification
AV  - none
T2  - The 6th International Conference on Computer Science, Applied Mathematics and Applications (ICCSAMA 2019)
ER  -