搜狗搜狗成立于2003年,是中国搜索行业的挑战者,AI领域的创新者。目前搜狗月活跃用户数仅次于BAT,是中国用户规模第四大互联网公司。2004年8月,搜狗推出搜狗搜索,现已成为中国第二大搜索引擎。2006年6月,推出搜狗输入法,重新定义了中文输入,截止2019年9月,搜狗输入法日活用户数4.5亿,是国内第一大中文输入法。
IEEE Transactions on Audio, Speech, and Language Processing, pp.1-1, (2021)
While the attention matrix only reveals the relevance between source and target words, the phrase alignment generated by our model is more useful for achieving lexically and structurally constrained decoding
Cited by2BibtexViews178DOI
0
0
Niu Yilin, Jiao Fangkai, Zhou Mantong, Yao Ting,Xu Jingfang,Huang Minlie
ACL, pp.3916-3927, (2020)
We present a Self-Training method, which supervises the evidence extractor with auto-generated evidence labels in an iterative process
Cited by7BibtexViews213
0
0
national conference on artificial intelligence, (2020)
We present ReCO, a large scale humancurated Chinese reading comprehension dataset focusing on opinion questions
Cited by2BibtexViews156
0
0
IJCAI 2020, (2020): 3694-3701
We have presented a voting mechanism for system combination in machine translation
Cited by1BibtexViews455DOI
0
0
national conference on artificial intelligence, (2020)
We introduce a novel hidden answer pivot module to the neural question generation which explicitly modeling the hidden answer information
Cited by1BibtexViews208
0
0
Yi-Ning Liu,Yan-Ping Wang, Xiao-Fen Wang, Zhe Xia,Jingfang Xu
Computer Networks, (2019): 340-348
A practical Internet of Things sensing collection protocol is proposed without a trusted authority, which preserves the raw data and unlinks the data with its contributor
Cited by35BibtexViews148DOI
0
0
Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Informatio..., pp.85-94, (2019)
To tackle the large-scale open-domain question answering problem, we proposed a document gated reader, which combines the convolutional neural network based document selection and LSTM based answer generation in a single model
Cited by3BibtexViews204DOI
0
0
EMNLP/IJCNLP (1), pp.6121-6131, (2019)
By making the source sentence and machine translation attend to each other, representations learned in such an interactive way help to identify whether a target word should be copied or be re-generated
Cited by1BibtexViews280DOI
0
0
ICTIR, pp.93-100, (2019)
To deal with story comprehension application Story Cloze Test, unlike most previous works which utilize the small development set for training
Cited by1BibtexViews266DOI
0
0
Jianhua Yuan, Yanyan Zhao,Bing Qin,Jingfang Xu
national conference on artificial intelligence, (2019)
We develop the Recurrent Conditional Attention model for answer stance detection in community QA
Cited by1BibtexViews157
0
0
Yan Zhang, Kien Tsong Chau,JingFang Xu, CuiTing Liu
Proceedings of the 3rd International Conference on Cryptography, Security and Privacy, pp.99-102, (2019)
That means 67% of user always or usually view the “Editor Recommendation” portion which is in the middle. This module has been used on 66rpg website for seven years and most experienced visitors are familiar with this module
Cited by0BibtexViews98DOI
0
0
Peilu Wang, Hao Jiang,Jingfang Xu,Qi Zhang
Data Intell., no. 4 (2019): 333-349
We introduce three applications of a knowledge graph in Sogou Inc.: entity detection and linking, knowledge-based question answering and knowledge-based dialog system which have been used to the Web search products to make knowledge acquisition more efficient
Cited by0BibtexViews164DOI
0
0
ACL (1), pp.1213-1223, (2019)
We find that language-based sharing manner is unstable because it achieves the highest BLEU score on Multi-NMT of similar languages, but the worst quality on dissimilar languages
Cited by0BibtexViews156DOI
0
0
IJCAI, pp.4623-4629, (2018)
We present a commonsense knowledge aware conversational model to demonstrate how commonsense knowledge can facilitate language understanding and generation in open-domain conversational systems
Cited by174BibtexViews920
1
0
EMNLP, (2018): 533-542
The past several years have witnessed the rapid development of neural machine translation, which investigates the use of neural networks to model the translation process
Cited by98BibtexViews336DOI
0
0
IJCAI, pp.4279-4285, (2018)
We address the issue of generating responses that are coherent to a pre-specified personality or profile
Cited by44BibtexViews163DOI
0
0
SC, pp.56:1-56:11, (2018)
Before describing ShenTu’s innovations, we introduce the Sunway TaihuLight system, which is optimized for traditional scientific computing applications and ever ranked
Cited by18BibtexViews179
0
0
EMNLP, pp.2955-2960, (2018)
We have proposed three effective strategies to improve the universal one-to-many multilingual translation, including special label initialization, language-dependent positional embedding and a new parameter-sharing mechanism
Cited by13BibtexViews317
0
0
Yi-Ning Liu,Yan-Ping Wang, Xiao-Fen Wang, Zhe Xia,Jingfang Xu
ISPEC, pp.506-518, (2018)
Assuming a data center, who recruits a group of participants to perform a time series sensing task, the data from the individual user is privacy-sensitive, and users do not trust any third-party to obtain the relationship between their identity and their sensing data
Cited by0BibtexViews68
0
0
CIKM, (2017): 257-266
We propose a new deep learning architecture, namely DeepRank
Cited by127BibtexViews152DOI
0
0
小科