查看“Sunwei”的源代码


{| border="0" cellpadding="1" cellspacing="1"
|-
| 
<span style="font-size:larger;">M.Sc. Student @ IIP&nbsp;Group<br/> Department of Computer Science and Technology<br/> Nanjing University</span>

<span style="font-size:larger;">Email: weisun_@outlook.com</span>

|-
| &nbsp;
| &nbsp;
|}

<span style="font-size:larger;"><span style="color:#3498db">'''Supervisor'''</span></span>

----

*<span style="font-size:larger;">Professor Jun-Yuan Xie</span> 

<span style="font-size:larger;"><span style="color:#3498db">'''Biography'''</span></span>

----

*<span style="font-size:larger;">I&nbsp;received my B.Sc. degree in&nbsp;of Soochow University in June 2017. In the same year, I was admitted to study for a Master degree in Nanjing University&nbsp;without entrance examination.&nbsp; Currently I am a second&nbsp;year M.Sc. student of Department of Computer Science and Technology&nbsp;in Nanjing University&nbsp;and a member of IIP&nbsp;Group, led by professor Jun-Yuan Xie and&nbsp; Chong-Jun Wang.</span> 

<span style="font-size:larger;"><span style="color:#3498db">'''Research Interest'''</span></span>

<span style="font-size:larger;">&nbsp;'''Machine Learning &&nbsp;Multi-label Learning'''.&nbsp;</span>

*<span style="font-size:larger;">On '''Multi-label Text Classification (MLTC),''' text features can be regarded as '''detailed description of documents''' and label sets can be '''a summarization of documents'''. '''Hybrid Topics''' from text features and label sets by LDA (a method of '''topic model''') can effectively mine global label correlations and deeper&nbsp;features. Meanwhile, a pair including topics and labels can mitigate the imbalanced problem of labels.</span> 
*<span style="font-size:larger;">Deep learning For multi-label text classification. We utilize '''dilated convolution''' to obtain the '''semantic understanding''' of the text and design a hybrid '''attention mechansim''' for '''different labels''' (Specifically, each label should attend to most relevant textual contents).&nbsp; Firstly,&nbsp;we initialize trainable label embeddings. Then After obationing word-level information based on Bi-LSTM, we&nbsp;get semantic understanding of texts&nbsp;based on word-level information by dilated convolution. Finally,&nbsp;we design a hybrid attention for different labels based on label embeddings.&nbsp; Besides, we add '''label cooccurrence matrix into loss function '''to guide the whole network to learn and achieve good results.&nbsp;&nbsp;</span> 
*<span style="font-size:larger;">'''GCN (Graph Convolution Network) '''can be used to exploit more complex label correlations on&nbsp;Multi-label Learning.</span> 

<span style="font-size:larger;"><font color="#3498db">'''Publications'''</font></span>

----

*<span style="font-size:larger;">Ran X., Pan Y., Sun W. and&nbsp;Wang C.. Learn to Select via Hierarchical Gate Mechanism for Aspect-Based Sentiment Analysis.&nbsp;In&nbsp;''Proceedings of IJCAL 2019.''</span> 
*<span style="font-size:larger;">Sun W., Wang C..Multi-label Classification: Select Distinct Semantic Understanding for Different Labels[C]. In&nbsp;''Proceedings of&nbsp;ApWeb-WAIM 2019.''</span> 
*<span style="font-size:larger;">Sun W., Ran X.. Luo X., and Wang C..An Efficient Framework by Topic Model for Multi-label Text Classification. In&nbsp;''Proceedings of IJCNN 2019.''</span> 
*<span style="font-size:larger;">Xu Y., Ran X.. Sun W., Luo X.&nbsp;and Wang C..Gated Neural Network with Regularized Loss for Multi-label Text Classification. In&nbsp;''Proceedings of IJCNN 2019.''</span> 
*<span style="font-size:larger;">Ran X., Pan Y., Sun W. and&nbsp;Wang C.. Modeling More Globally: A Hierarchical Attention Network via Multi-Task Learning for Aspect-Based Sentiment Analysis. In&nbsp;''Proceedings of DASFAA 2019'', Chiang Mai, Thailand, Apr. 22-25,&nbsp;2019: 505-509.</span> 

<span style="font-size:larger;"><span style="color:#3498db">'''Resources'''</span></span>

----

*<span style="font-size:larger;">[http://manikvarma.org/downloads/XC/XMLRepository.html Extreme Classiﬁcation Repository]: for large-scale multi-label datasets and off-the-shelf eXtreme Multi-Label Learning (XML) solvers.</span> 
*<span style="font-size:larger;">[http://mulan.sourceforge.net/datasets-mlc.html Mulan Multi-Label Learning Datasets]: regular/traditional multi-label learning datasets.</span> 
*<span style="font-size:larger;">[https://github.com/XSilverBullet/Multi-label-Paper Related Work]: This page categorizes a list of works of my interest, mainly in Multi-Label Learning.</span> 

<span style="font-size:larger;"><span style="color:#3498db">'''Rewards or Honors'''</span></span>

*<span style="font-size:larger;">Second-Class Academic Scholarship, 2018-2019</span> 
*<span style="font-size:larger;">First-Class Academic Scholarship, 2017-2018</span> 
*<span style="font-size:larger;">Outstanding Graduate Student, 2017.06</span> 
*<span style="font-size:larger;">CCF Excellent University Student, 2016.10</span> 
*<span style="font-size:larger;">National Scholarship, 2015.11</span>