一周新论文 | 2020年第11周 | 自然语言处理相关

《一周新论文》系列之2020年第11周：自然语言处理相关

本周重点关注：

• Google: [9], [12], [22], [36]
• Microsoft: [19], [64]
• Facebook: [11], [39]
• 其他：[15], [17], [22], [23], [27], [41], [47], [49], [59], [73], [86]

2020年3月13日

[1]. KGvec2go – Knowledge Graph Embeddings as a Service
链接 | https://arxiv.org/abs/2003.05809
作者 | Jan Portisch, Michael Hladik, Heiko Paulheim
备注 | to be published in the Proceedings of the International Conference on Language Resources and Evaluation (LREC) 2020

[2]. It Means More if It Sounds Good: Yet Another Hypotheses Concerning the Evolution of Polysemous Words
链接 | https://arxiv.org/abs/2003.05758
作者 | Ivan P. Yamshchikov, Cyrille Merleau Nono Saha, Igor Samenko, Jürgen Jost

[3]. Sentiment Analysis with Contextual Embeddings and Self-Attention
链接 | https://arxiv.org/abs/2003.05574
作者 | Katarzyna Biesialska, Magdalena Biesialska, Henryk Rybinski
备注 | Accepted at the 25th International Symposium on Methodologies for Intelligent Systems (ISMIS 2020)

[4]. Learning word-referent mappings and concepts from raw inputs
链接 | https://arxiv.org/abs/2003.05573
作者 | Wai Keen Vong, Brenden M. Lake
单位 | New York University

[5]. Semantic Holism and Word Representations in Artificial Neural Networks
链接 | https://arxiv.org/abs/2003.05522
作者 | Tomáš Musil

[6]. Investigating Entity Knowledge in BERT with Simple Neural End-To-End Entity Linking
链接 | https://arxiv.org/abs/2003.05473
作者 | Samuel Broscheit
备注 | Published at CoNLL 2019

[7]. A Precisely Xtreme-Multi Channel Hybrid Approach For Roman Urdu Sentiment Analysis
链接 | https://arxiv.org/abs/2003.05443
作者 | Faiza Memood, Muhammad Usman Ghani, Muhammad Ali Ibrahim, Rehab Shehzadi, Muhammad Nabeel Asim

2020年3月12日

[8]. Brazilian Lyrics-Based Music Genre Classification Using a BLSTM Network
链接 | https://arxiv.org/abs/2003.05377
作者 | Raul de Araújo Lima, Rômulo César Costa de Sousa, Simone Diniz Junqueira Barbosa, Hélio Cortês Vieira Lopes

[9]. Capturing document context inside sentence-level neural machine translation models with self-training
链接 | https://arxiv.org/abs/2003.05259
作者 | Elman Mansimov, Gábor Melis, Lei Yu
单位 | New York University; DeepMind;

[10]. Vector symbolic architectures for context-free grammars
链接 | https://arxiv.org/abs/2003.05171
作者 | Peter beim Graben, Markus Huber, Werner Meyer, Ronald Römer, Constanze Tschöpe, Matthias Wolff

[11]. A Benchmark for Systematic Generalization in Grounded Language Understanding
链接 | https://arxiv.org/abs/2003.05161
作者 | Laura Ruis, Jacob Andreas, Marco Baroni, Diane Bouchacourt, Brenden M. Lake
单位 | University of Amsterdam; Massachusetts Institute of Technology; ICREA; Facebook AI Research; New York University

[12]. TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages
链接 | https://arxiv.org/abs/2003.05002
作者 | Jonathan H. Clark, Eunsol Choi, Michael Collins, Dan Garrette, Tom Kwiatkowski, Vitaly Nikolaev, Jennimaria Palomaki
单位 | Google Research

[13]. Toward Interpretability of Dual-Encoder Models for Dialogue Response Suggestions
链接 | https://arxiv.org/abs/2003.04998
作者 | Yitong Li, Dianqi Li, Sushant Prakash, Peng Wang
单位 | Duke University; University of Washington; Google Research

[14]. GASP! Generating Abstracts of Scientific Papers from Abstracts of Cited Papers
链接 | https://arxiv.org/abs/2003.04996
作者 | Fabio Massimo Zanzotto, Viviana Bono, Paola Vocca, Andrea Santilli, Danilo Croce, Giorgio Gambosi, Roberto Basili

[15]. Masking Orchestration: Multi-task Pretraining for Multi-role Dialogue Representation Learning
链接 | https://arxiv.org/abs/2003.04994
作者 | Tianyi Wang, Yating Zhang, Xiaozhong Liu, Changlong Sun, Qiong Zhang
单位 | Alibaba Group; Indiana University Bloomington
备注 | 8 pages, 4 figures, AAAI2020

[16]. Learning to mirror speaking styles incrementally
链接 | https://arxiv.org/abs/2003.04993
作者 | Siyi Liu, Ziang Leng, Derry Wijaya
单位 | Boston University

[17]. Multi-task Learning with Multi-head Attention for Multi-choice Reading Comprehension
链接 | https://arxiv.org/abs/2003.04992
作者 | Hui Wan
单位 | IBM Research

[18]. Unsupervised and Interpretable Do