Xiaodong Shi

Xiaodong Shi

Auto-generated profile, based on publicly available information

Xiaodong Shi

School of Informatics, Xiamen University, Xiamen, China

Research Focus & Metrics

Academic Career Length

21Years (2006-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 July 2019 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
HXHan Xiao
Han Xiao
YCYidong Chen
Yidong Chen
XSXiaodong Shi
Xiaodong Shi
Knowledge representation is one of the critical problems in knowledge engineering and artificial intelligence, while knowledge embedding as a knowledge representation methodology indicates entities and relations in knowledge graph as low-dimensional, continuous vectors. In this way, knowledge graph is compatible with machine learning models. Major knowledge embedding methods employ geometric translation to design score function, which is weak-semantic for natural language processing. To overcome this disadvantage, in this paper, we propose our model based on multi-view clustering framework, which could generate semantic representations of knowledge elements (i.e. entities/relations). With our semantic model, we also present an empowered solution to entity retrieval with entity description. Extensive experiments show that our model achieves substantial improvements against baselines on the task of knowledge graph completion, triple classification, entity classification and entity retrieval.

24 April 2020 in Scientometrics
Verified
Journal Article
Closed access
PDF file
JLJialiang Lin
Jialiang Lin
YYYao Yu
Yao Yu
YZYu Zhou
Yu Zhou
ZZZhiyang Zhou
Zhiyang Zhou
XSXiaodong Shi
Xiaodong Shi

23 September 2023 in Neural Computing and Applications
Verified
Journal Article
Closed access
PDF file
JLJialiang Lin
Jialiang Lin
JSJiaxin Song
Jiaxin Song
ZZZhangping Zhou
Zhangping Zhou
YCYidong Chen
Yidong Chen
XSXiaodong Shi
Xiaodong Shi

11 February 2017 in Proceedings of the AAAI Conference on Artificial Intelligence
Verified
Journal Article
Open access
PDF file
JSJinsong Su
Jinsong Su
ZTZhixing Tan
Zhixing Tan
DXDeyi Xiong
Deyi Xiong
RJRongrong Ji
Rongrong Ji
XSXiaodong Shi
Xiaodong Shi
YLYang Liu
Yang Liu
Neural machine translation (NMT) heavily relies on word-level modelling to learn semantic representations of input sentences.However, for languages without natural word delimiters (e.g., Chinese) where input sentences have to be tokenized first,conventional NMT is confronted with two issues:1) it is difficult to find an optimal tokenization granularity for source sentence modelling, and2) errors in 1-best tokenizations may propagate to the encoder of NMT.To handle these issues, we propose word-lattice based Recurrent Neural Network (RNN) encoders for NMT,which generalize the standard RNN to word lattice topology.The proposed encoders take as input a word lattice that compactly encodes multiple tokenizations, and learn to generate new hidden states from arbitrarily many inputs and hidden states in preceding time steps.As such, the word-lattice based encoders not only alleviate the negative impact of tokenization errors but also are more expressive and flexible to embed input sentences.Experiment results on Chinese-English translation demonstrate the superiorities of the proposed encoders over the conventional encoder.

5 August 2020 in Polish Journal of Environmental Studies
Verified
Journal Article
Open access
HGHuinan Guo
Huinan Guo
XSXiaodong Shi
Xiaodong Shi
LMLijuan Ma
Lijuan Ma
TYTao Yang
Tao Yang
WMWei Min
Wei Min
Irrigation with saline water alters the soil environment, thereby influencing soil microbial processes and crop growth. The objective of this field experiment was to compare the effects of long-term irrigation with saline water or freshwater on cotton growth, soil physicochemical properties, and bacterial community structure. High-throughput sequencing of DNA of the 16S rRNA gene was used to assess the composition of bacterial communities. Saline-water irrigation significantly increased soil salinity and significantly decreased pH, soil organic matter, total nitrogen, and biomass and yield of cotton. Cotton biomass and yield were negatively correlated with soil salinity but were positively correlated with soil organic matter, and total nitrogen. Compared with fresh water irrigation, irrigation with saline water significantly decreased the Shannon diversity index of the bacterial community. Compared with freshwater irrigation, saline-water irrigation significantly increased the relative abundance of two phyla of bacteria (Gemmatimonadetes and Actinobacteria) and decreased that of two (Proteobacteria and Acidobacteria). In addition, the relative abundance of 5 genera was significantly higher (Sphingomonas, Gemmatimonas, Gaiella, Solirubrobacter, and Nocardioides) and that of 7 was significantly lower (RB41, H16, Haliangium, Nitrospira, Lysobacter, Acidobacteria_bacterium, and Bryobacter) in the saline-water treatment. Soil salinity and pH were the major environmental factors that shaped bacterial communities. According to LEfSe analysis, only 3 indicator species were associated with saline-water irrigation, whereas 5 indicator species were associated with freshwater irrigation. Overall, long-term irrigation with saline water adversely affected soil properties and led to decreases in soil bacterial community diversity and cotton growth and yield. These results contribute to a more comprehensive understanding of bacterial communities in saline environments.

3 November 2023 in Neural Networks
Verified
Journal Article
Closed access
HWHui Wu
Hui Wu
YHYuting He
Yuting He
YCYidong Chen
Yidong Chen
YBYu Bai
Yu Bai
XSXiaodong Shi
Xiaodong Shi