检索结果-内蒙古大学图书馆

Code summarization with structure-induced Transformer

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Wu, Hongqiu Zhao, Hai Zhang, Min Department of Computer Science and Engineering Shanghai Jiao Tong University China Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China Institute of Artificial Intelligence School of Computer Science and Technology Soochow University Suzhou China

Code summarization (CS) is becoming a promising area in recent language understanding, which aims to generate sensible human language automatically for programming language in the format of source code, serving in the most convenience of programmer developing. It is well known that programming languages are highly structured. Thus previous works attempt to apply structure-based traversal (SBT) or non-sequential models like Tree-LSTM and graph neural network (GNN) to learn structural program semantics. However, it is surprising that incorporating SBT into advanced encoder like Transformer instead of LSTM has been shown no performance gain, which lets GNN become the only rest means modeling such necessary structural clue in source code. To release such inconvenience, we propose structure-induced Transformer, which encodes sequential code inputs with multi-view structural clues in terms of a newly-proposed structure-induced self-attention mechanism. Extensive experiments show that our proposed structure-induced Transformer helps achieve new state-of-the-art results on benchmarks. Copyright © 2020, The Authors. All rights reserved.

关键词： Trees (mathematics)

What you say and how you say it: Joint modeling of topics and discourse in microblog conversations

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Zeng, Jichuan Li, Jing He, Yulan Gao, Cuiyun Lyu, Michael R. King, Irwin Department of Computer Science and Engineering Chinese University of Hong Kong HKSAR China Tencent AI Lab Shenzhen China Department of Computer Science University of Warwick United Kingdom

This paper presents an unsupervised framework for jointly modeling topic content and discourse behavior in microblog conversations. Concretely, we propose a neural model to discover word clusters indicating what a conversation concerns (i.e., topics) and those reflecting how participants voice their opinions (i.e., discourse).1Extensive experiments show that our model can yield both coherent topics and meaningful discourse behavior. Further study shows that our topic and discourse representations can benefit the classification of microblog messages, especially when they are jointly trained with the classifier. Copyright © 2019, The Authors. All rights reserved.

关键词：

SPRSound: Open-Source SJTU Paediatric Respiratory Sound Database

学校读者我要写书评

暂无评论

TechRxiv

TechRxiv 2022年

作者： Li, Yongfu Zhang, Qing Zhang, Jing Yuan, Jiajun Huang, Huajie Zhang, Yuhang Zhang, Baoqin Lv, Gaomei Lin, Shuzhu Wang, Na Liu, Xin Tang, Mingyu Wang, Yahua Ma, Hui Liu, Lu Yuan, Shuhua Zhou, Hongyuan Zhao, Jian Yin, Yong Zhao, Liebin Wang, Guoxing Lian, Yong Department of Micro-Nano Electronics and MoE Key Lab of Artificial Intelligence Shanghai Jiao Tong University Shanghai200240 China Pediatric AI Clinical Application and Research Center Shanghai Children’s Medical Center Shanghai Jiao Tong University School of Medicine China Department of Respiratory Medicine Shanghai Children’s Medical Center Shanghai Jiao Tong University School of Medicine China China School of Computer Engineering and Science Shanghai University China Sanya Maternity and Child Care Hospital China Department of Paediatrics Taicang Affiliated Hospital of Soochow University The First People’s Hospital of Taicang China Linyi City People Hospital China Department of Paediatrics Fengcheng Hospital China Department of Pediatrics Fifth Affiliated Hospital of Harbin Medical University China Shanghai Tuoxiao Intelligent Technology Co. Ltd China

It has proved that the auscultation of respiratory sound has advantage in early respiratory diagnosis. Various methods have been raised to perform automatic respiratory sound analysis to reduce subjective diagnosis and physicians’ workload. However, these methods highly rely on the quality of respiratory sound database. In this work, we have developed the first open-access paediatric respiratory sound database, SPRSound. The database consists of 2,683 records and 9,089 respiratory sound events from 292 participants. Accurate label is important to achieve a good prediction for adventitious respiratory sound classification problem. A custom-made sound label annotation software (SoundAnn) has been developed to perform sound editing, sound annotation, and quality assurance evaluation. A team of 11 experienced paediatric physicians is involved in the entire process to establish golden standard reference for the dataset. To verify the robustness and accuracy of the classification model, we have investigated the effects of different feature extraction methods and machine learning classifiers on the classification performance of our dataset. As such, we have achieved a score of 75.22%, 61.57%, 56.71%, and © 2022, CC BY.

关键词： Pediatrics

Evaluation for Weakly Supervised Object Localization: Protocol, Metrics, and Datasets

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Choe, Junsuk Oh, Seong Joon Chun, Sanghyuk Lee, Seungho Akata, Zeynep Shim, Hyunjung The Department of Computer Science and Engineering Sogang University Seoul Korea Republic of NAVER AI Lab The School of Integrated Technology Yonsei University Korea Republic of University of Tuebingen Max Planck Institute for Informatics Max Planck Institute for Intelligent Systems Germany

Weakly-supervised object localization (WSOL) has gained popularity over the last years for its promise to train localization models with only image-level labels. Since the seminal WSOL work of class activation mapping (CAM), the field has focused on how to expand the attention regions to cover objects more broadly and localize them better. However, these strategies rely on full localization supervision for validating hyperparameters and model selection, which is in principle prohibited under the WSOL setup. In this paper, we argue that WSOL task is ill-posed with only image-level labels, and propose a new evaluation protocol where full supervision is limited to only a small held-out set not overlapping with the test set. We observe that, under our protocol, the five most recent WSOL methods have not made a major improvement over the CAM baseline. Moreover, we report that existing WSOL methods have not reached the few-shot learning baseline, where the full-supervision at validation time is used for model training instead. Based on our findings, we discuss some future directions for WSOL. Source code and dataset are available at https://***/clovaai/wsolevaluation. Copyright © 2020, The Authors. All rights reserved.

关键词： Object recognition

Dialogue graph modeling for conversational machine reading

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Ouyang, Siru Zhang, Zhuosheng Zhao, Hai Department of Computer Science and Engineering Shanghai Jiao Tong University Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China

Conversational Machine Reading (CMR) aims at answering questions in complicated interactive scenarios. Machine needs to answer questions through interactions with users based on given rule document, user scenario and dialogue history, and even initiatively asks questions for clarification if necessary. Namely, the answer to the task needs a machine in the response of either Yes, No, Irrelevant or to raise a follow-up question for further clarification. To effectively capture multiple objects in such a challenging task, graph modeling is supposed to be adopted, though it is surprising that this does not happen until this work proposes a dialogue graph modeling framework by incorporating two complementary graph models, i.e., explicit discourse graph and implicit discourse graph, which respectively capture explicit and implicit interactions hidden in the rule documents. The proposed model is evaluated on the ShARC benchmark and achieves new state-of-the-art by first exceeding the milestone accuracy score of 80%. The source code of our paper is available at https://***/ozyyshr/DGM. Copyright © 2020, The Authors. All rights reserved.

关键词： Graph theory

Reference language based unsupervised neural machine translation

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Li, Zuchao Zhao, Hai Wang, Rui Utiyama, Masao Sumita, Eiichiro Department of Computer Science and Engineering Shanghai Jiao Tong University Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Kyoto Japan

Exploiting common language as an auxiliary for better translation has a long tradition in machine translation, which lets supervised learning based machine translation enjoy the enhancement delivered by the well-used pivot language, in case that the prerequisite of parallel corpus from source language to target language cannot be fully satisfied. The rising of unsupervised neural machine translation (UNMT) seems completely relieving the parallel corpus curse, though still subject to unsatisfactory performance so far due to vague clues available used for its core back-translation training. Further enriching the idea of pivot translation by freeing the use of parallel corpus other than its specified source and target, we propose a new reference language based UNMT framework, in which the reference language only shares parallel corpus with the source, indicating clear enough signal to help the reconstruction training of UNMT through a proposed reference agreement mechanism. Experimental results show that our methods improve the quality of UNMT over that of a strong baseline in terms of only one auxiliary language, demonstrating the usefulness of the proposed reference language based UNMT with a good start. Copyright © 2020, The Authors. All rights reserved.

关键词： Neural machine translation

Topic-aware multi-turn dialogue modeling

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Xu, Yi Zhao, Hai Zhang, Zhuosheng Department of Computer Science and Engineering Shanghai Jiao Tong University China Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University China

In the retrieval-based multi-turn dialogue modeling, it remains a challenge to select the most appropriate response according to extracting salient features in context utterances. As a conversation goes on, topic shift at discourse-level naturally happens through the continuous multi-turn dialogue context. However, all known retrieval-based systems are satisfied with exploiting local topic words for context utterance representation but fail to capture such essential global topic-aware clues at discourse-level. Instead of taking topic-agnostic ngram utterance as processing unit for matching purpose in existing systems, this paper presents a novel topic-aware solution for multi-turn dialogue modeling, which segments and extracts topic-aware utterances in an unsupervised way, so that the resulted model is capable of capturing salient topic shift at discourse-level in need and thus effectively track topic flow during multi-turn conversation. Our topic-aware modeling is implemented by a newly proposed unsupervised topic-aware segmentation algorithm and Topic-Aware Dual-attention Matching (TADAM) Network, which matches each topic segment with the response in a dual cross-attention way. Experimental results on three public datasets show TADAM can outperform the state-of-the-art method by a large margin, especially by 3.4% on E-commerce dataset that has an obvious topic shift. Copyright © 2020, The Authors. All rights reserved.

关键词： Large dataset

OpenVoting: Recoverability from Failures in Dual Voting

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Agrawal, Prashant Tomer, Kabir Nakarmi, Abhinav Jhanwar, Mahabir Prasad Sharma, Subodh Banerjee, Subhashis Department of Computer Science and Engineering Iit Delhi New Delhi India Department of Computer Science University of Illinois Urbana-Champaign United States Department of Computer Science Centre for Digitalisation Ai and Society Ashoka University Sonipat India

In this paper we address the problem of recovery from failures without re-running entire elections when elections fail to verify. We consider the setting of dual voting protocols, where the cryptographic guarantees of end-to-end verifiable voting (E2E-V) are combined with the simplicity of audit using voter-verified paper records (VVPR). We first consider the design requirements of such a system and then suggest a protocol called OpenVoting, which identifies a verifiable subset of error-free votes consistent with the VVPRs, and the polling booths corresponding to the votes that fail to verify with possible reasons for the failures. To an ordinary voter OpenVoting looks just like an old fashioned paper based voting system, with minimal additional cognitive overload. Copyright © 2019, The Authors. All rights reserved.

关键词： Electronic voting

Towards Open-World Recommendation: An Inductive Model-based Collaborative Filtering Approach

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Wu, Qitian Zhang, Hengrui Gao, Xiaofeng Yan, Junchi Zha, Hongyuan Department of Computer Science and Engineering Shanghai Jiao Tong University China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University China School of Data Science Shenzhen Institute of Artificial Intelligence and Robotics for Society The Chinese University of Hong Kong Shenzhen China

Recommendation models can effectively estimate underlying user interests and predict one's future behaviors by factorizing an observed user-item rating matrix into products of two sets of latent factors. However, the user-specific embedding factors can only be learned in a transductive way, making it difficult to handle new users on-the-fly. In this paper, we propose an inductive collaborative filtering framework that contains two representation models. The first model follows conventional matrix factorization which factorizes a group of key users' rating matrix to obtain meta latents. The second model resorts to attention-based structure learning that estimates hidden relations from query to key users and learns to leverage meta latents to inductively compute embeddings for query users via neural message passing. Our model enables inductive representation learning for users and meanwhile guarantees equivalent representation capacity as matrix factorization. Experiments demonstrate that our model achieves promising results for recommendation on few-shot users with limited training ratings and new unseen users which are commonly encountered in open-world recommender systems. The codes are available at https://***/qitianwu/IDCF. © 2020, CC BY.

关键词： Collaborative filtering