Paper Group NANR 126
Semi-Supervised Learning via New Deep Network Inversion. BioAMA: Towards an End to End BioMedical Question Answering System. Telling Stories with Soundtracks: An Empirical Analysis of Music in Film. Social Image Tags as a Source of Word Embeddings: A Task-oriented Evaluation. Towards a Welsh Semantic Annotation System. Bacteria and Biotope Entity R …
Semi-Supervised Learning via New Deep Network Inversion
| Title | Semi-Supervised Learning via New Deep Network Inversion |
| Authors | Balestriero R., Roger V., Glotin H., Baraniuk R. |
| Abstract | We exploit a recently derived inversion scheme for arbitrary deep neural networks to develop a new semi-supervised learning framework that applies to a wide range of systems and problems. The approach reaches current state-of-the-art methods on MNIST and provides reasonable performances on SVHN and CIFAR10. Through the introduced method, residual networks are for the first time applied to semi-supervised tasks. Experiments with one-dimensional signals highlight the generality of the method. Importantly, our approach is simple, efficient, and requires no change in the deep network architecture. |
| Tasks | |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=B1i7ezW0- |
| https://openreview.net/pdf?id=B1i7ezW0- | |
| PWC | https://paperswithcode.com/paper/semi-supervised-learning-via-new-deep-network |
| Repo | |
| Framework | |
BioAMA: Towards an End to End BioMedical Question Answering System
| Title | BioAMA: Towards an End to End BioMedical Question Answering System |
| Authors | Vasu Sharma, Nitish Kulkarni, Srividya Pranavi, Gabriel Bayomi, Eric Nyberg, Teruko Mitamura |
| Abstract | In this paper, we present a novel Biomedical Question Answering system, BioAMA: {``}Biomedical Ask Me Anything{''} on task 5b of the annual BioASQ challenge. In this work, we focus on a wide variety of question types including factoid, list based, summary and yes/no type questions that generate both exact and well-formed {`}ideal{'} answers. For summary-type questions, we combine effective IR-based techniques for retrieval and diversification of relevant snippets for a question to create an end-to-end system which achieves a ROUGE-2 score of 0.72 and a ROUGE-SU4 score of 0.71 on ideal answer questions (7{%} improvement over the previous best model). Additionally, we propose a novel NLI-based framework to answer the yes/no questions. To train the NLI model, we also devise a transfer-learning technique by cross-domain projection of word embeddings. Finally, we present a two-stage approach to address the factoid and list type questions by first generating a candidate set using NER taggers and ranking them using both supervised or unsupervised techniques. | |
| Tasks | Natural Language Inference, Question Answering, Transfer Learning, Word Embeddings |
| Published | 2018-07-01 |
| URL | https://www.aclweb.org/anthology/W18-2312/ |
| https://www.aclweb.org/anthology/W18-2312 | |
| PWC | https://paperswithcode.com/paper/bioama-towards-an-end-to-end-biomedical |
| Repo | |
| Framework | |
Telling Stories with Soundtracks: An Empirical Analysis of Music in Film
| Title | Telling Stories with Soundtracks: An Empirical Analysis of Music in Film |
| Authors | Jon Gillick, David Bamman |
| Abstract | Soundtracks play an important role in carrying the story of a film. In this work, we collect a corpus of movies and television shows matched with subtitles and soundtracks and analyze the relationship between story, song, and audience reception. We look at the content of a film through the lens of its latent topics and at the content of a song through descriptors of its musical attributes. In two experiments, we find first that individual topics are strongly associated with musical attributes, and second, that musical attributes of soundtracks are predictive of film ratings, even after controlling for topic and genre. |
| Tasks | Image Captioning, Question Answering |
| Published | 2018-06-01 |
| URL | https://www.aclweb.org/anthology/W18-1504/ |
| https://www.aclweb.org/anthology/W18-1504 | |
| PWC | https://paperswithcode.com/paper/telling-stories-with-soundtracks-an-empirical |
| Repo | |
| Framework | |
Social Image Tags as a Source of Word Embeddings: A Task-oriented Evaluation
| Title | Social Image Tags as a Source of Word Embeddings: A Task-oriented Evaluation |
| Authors | Mika Hasegawa, Tetsunori Kobayashi, Yoshihiko Hayashi |
| Abstract | |
| Tasks | Semantic Textual Similarity, Word Embeddings |
| Published | 2018-05-01 |
| URL | https://www.aclweb.org/anthology/L18-1156/ |
| https://www.aclweb.org/anthology/L18-1156 | |
| PWC | https://paperswithcode.com/paper/social-image-tags-as-a-source-of-word |
| Repo | |
| Framework | |
Towards a Welsh Semantic Annotation System
| Title | Towards a Welsh Semantic Annotation System |
| Authors | Scott Piao, Paul Rayson, Dawn Knight, Gareth Watkins |
| Abstract | |
| Tasks | |
| Published | 2018-05-01 |
| URL | https://www.aclweb.org/anthology/L18-1158/ |
| https://www.aclweb.org/anthology/L18-1158 | |
| PWC | https://paperswithcode.com/paper/towards-a-welsh-semantic-annotation-system |
| Repo | |
| Framework | |
Bacteria and Biotope Entity Recognition Using A Dictionary-Enhanced Neural Network Model
| Title | Bacteria and Biotope Entity Recognition Using A Dictionary-Enhanced Neural Network Model |
| Authors | Qiuyue Wang, Xiaofeng Meng |
| Abstract | Automatic recognition of biomedical entities in text is the crucial initial step in biomedical text mining. In this pa-per, we investigate employing modern neural network models for recognizing biomedical entities. To compensate for the small amount of training data in biomedical domain, we propose to integrate dictionaries into the neural model. Our experiments on BB3 data sets demonstrate that state-of-the-art neural network model is promising in recognizing biomedical entities even with very little training data. When integrated with dictionaries, its performance could be greatly improved, achieving the competitive performance compared with the best dictionary-based system on the entities with specific terminology, and much higher performance on the entities with more general terminology. |
| Tasks | Feature Engineering |
| Published | 2018-07-01 |
| URL | https://www.aclweb.org/anthology/W18-2317/ |
| https://www.aclweb.org/anthology/W18-2317 | |
| PWC | https://paperswithcode.com/paper/bacteria-and-biotope-entity-recognition-using |
| Repo | |
| Framework | |
Code Synthesis with Priority Queue Training
| Title | Code Synthesis with Priority Queue Training |
| Authors | Daniel A. Abolafia, Quoc V. Le, Mohammad Norouzi |
| Abstract | We consider the task of program synthesis in the presence of a reward function over the output of programs, where the goal is to find programs with maximal rewards. We introduce a novel iterative optimization scheme, where we train an RNN on a dataset of K best programs from a priority queue of the generated programs so far. Then, we synthesize new programs and add them to the priority queue by sampling from the RNN. We benchmark our algorithm called priority queue training (PQT) against genetic algorithm and reinforcement learning baselines on a simple but expressive Turing complete programming language called BF. Our experimental results show that our deceptively simple PQT algorithm significantly outperforms the baselines. By adding a program length penalty to the reward function, we are able to synthesize short, human readable programs. |
| Tasks | Program Synthesis |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=r1AoGNlC- |
| https://openreview.net/pdf?id=r1AoGNlC- | |
| PWC | https://paperswithcode.com/paper/code-synthesis-with-priority-queue-training |
| Repo | |
| Framework | |
The Case for Systematically Derived Spatial Language Usage
| Title | The Case for Systematically Derived Spatial Language Usage |
| Authors | Bonnie Dorr, Clare Voss |
| Abstract | This position paper argues that, while prior work in spatial language understanding for tasks such as robot navigation focuses on mapping natural language into deep conceptual or non-linguistic representations, it is possible to systematically derive regular patterns of spatial language usage from existing lexical-semantic resources. Furthermore, even with access to such resources, effective solutions to many application areas such as robot navigation and narrative generation also require additional knowledge at the syntax-semantics interface to cover the wide range of spatial expressions observed and available to natural language speakers. We ground our insights in, and present our extensions to, an existing lexico-semantic resource, covering 500 semantic classes of verbs, of which 219 fall within a spatial subset. We demonstrate that these extensions enable systematic derivation of regular patterns of spatial language without requiring manual annotation. |
| Tasks | Robot Navigation |
| Published | 2018-06-01 |
| URL | https://www.aclweb.org/anthology/W18-1408/ |
| https://www.aclweb.org/anthology/W18-1408 | |
| PWC | https://paperswithcode.com/paper/the-case-for-systematically-derived-spatial |
| Repo | |
| Framework | |
Proceedings of the First Workshop on Storytelling
| Title | Proceedings of the First Workshop on Storytelling |
| Authors | |
| Abstract | |
| Tasks | |
| Published | 2018-06-01 |
| URL | https://www.aclweb.org/anthology/W18-1500/ |
| https://www.aclweb.org/anthology/W18-1500 | |
| PWC | https://paperswithcode.com/paper/proceedings-of-the-first-workshop-on-7 |
| Repo | |
| Framework | |
Arabic Data Science Toolkit: An API for Arabic Language Feature Extraction
| Title | Arabic Data Science Toolkit: An API for Arabic Language Feature Extraction |
| Authors | Paul Rodrigues, Valerie Novak, C. Anton Rytting, Julie Yelle, Jennifer Boutz |
| Abstract | |
| Tasks | |
| Published | 2018-05-01 |
| URL | https://www.aclweb.org/anthology/L18-1198/ |
| https://www.aclweb.org/anthology/L18-1198 | |
| PWC | https://paperswithcode.com/paper/arabic-data-science-toolkit-an-api-for-arabic |
| Repo | |
| Framework | |
Supervised autoencoders: Improving generalization performance with unsupervised regularizers
| Title | Supervised autoencoders: Improving generalization performance with unsupervised regularizers |
| Authors | Lei Le, Andrew Patterson, Martha White |
| Abstract | Generalization performance is a central goal in machine learning, particularly when learning representations with large neural networks. A common strategy to improve generalization has been through the use of regularizers, typically as a norm constraining the parameters. Regularizing hidden layers in a neural network architecture, however, is not straightforward. There have been a few effective layer-wise suggestions, but without theoretical guarantees for improved performance. In this work, we theoretically and empirically analyze one such model, called a supervised auto-encoder: a neural network that predicts both inputs (reconstruction error) and targets jointly. We provide a novel generalization result for linear auto-encoders, proving uniform stability based on the inclusion of the reconstruction error—particularly as an improvement on simplistic regularization such as norms or even on more advanced regularizations such as the use of auxiliary tasks. Empirically, we then demonstrate that, across an array of architectures with a different number of hidden units and activation functions, the supervised auto-encoder compared to the corresponding standard neural network never harms performance and can significantly improve generalization. |
| Tasks | |
| Published | 2018-12-01 |
| URL | http://papers.nips.cc/paper/7296-supervised-autoencoders-improving-generalization-performance-with-unsupervised-regularizers |
| http://papers.nips.cc/paper/7296-supervised-autoencoders-improving-generalization-performance-with-unsupervised-regularizers.pdf | |
| PWC | https://paperswithcode.com/paper/supervised-autoencoders-improving |
| Repo | |
| Framework | |
Multi-Sentence Compression with Word Vertex-Labeled Graphs and Integer Linear Programming
| Title | Multi-Sentence Compression with Word Vertex-Labeled Graphs and Integer Linear Programming |
| Authors | Elvys Linhares Pontes, St{'e}phane Huet, Thiago Gouveia da Silva, Andr{'e}a carneiro Linhares, Juan-Manuel Torres-Moreno |
| Abstract | Multi-Sentence Compression (MSC) aims to generate a short sentence with key information from a cluster of closely related sentences. MSC enables summarization and question-answering systems to generate outputs combining fully formed sentences from one or several documents. This paper describes a new Integer Linear Programming method for MSC using a vertex-labeled graph to select different keywords, and novel 3-gram scores to generate more informative sentences while maintaining their grammaticality. Our system is of good quality and outperforms the state-of-the-art for evaluations led on news dataset. We led both automatic and manual evaluations to determine the informativeness and the grammaticality of compressions for each dataset. Additional tests, which take advantage of the fact that the length of compressions can be modulated, still improve ROUGE scores with shorter output sentences. |
| Tasks | Question Answering, Sentence Compression, Text Summarization |
| Published | 2018-06-01 |
| URL | https://www.aclweb.org/anthology/W18-1704/ |
| https://www.aclweb.org/anthology/W18-1704 | |
| PWC | https://paperswithcode.com/paper/multi-sentence-compression-with-word-vertex |
| Repo | |
| Framework | |
Learning Document Embeddings With CNNs
| Title | Learning Document Embeddings With CNNs |
| Authors | Shunan Zhao, Chundi Lui, Maksims Volkovs |
| Abstract | This paper proposes a new model for document embedding. Existing approaches either require complex inference or use recurrent neural networks that are difficult to parallelize. We take a different route and use recent advances in language modeling to develop a convolutional neural network embedding model. This allows us to train deeper architectures that are fully parallelizable. Stacking layers together increases the receptive filed allowing each successive layer to model increasingly longer range semantic dependences within the document. Empirically we demonstrate superior results on two publicly available benchmarks. Full code will be released with the final version of this paper. |
| Tasks | Document Embedding, Language Modelling, Network Embedding |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=ryHM_fbA- |
| https://openreview.net/pdf?id=ryHM_fbA- | |
| PWC | https://paperswithcode.com/paper/learning-document-embeddings-with-cnns |
| Repo | |
| Framework | |
Feature Map Variational Auto-Encoders
| Title | Feature Map Variational Auto-Encoders |
| Authors | Lars Maaløe, Ole Winther |
| Abstract | There have been multiple attempts with variational auto-encoders (VAE) to learn powerful global representations of complex data using a combination of latent stochastic variables and an autoregressive model over the dimensions of the data. However, for the most challenging natural image tasks the purely autoregressive model with stochastic variables still outperform the combined stochastic autoregressive models. In this paper, we present simple additions to the VAE framework that generalize to natural images by embedding spatial information in the stochastic layers. We significantly improve the state-of-the-art results on MNIST, OMNIGLOT, CIFAR10 and ImageNet when the feature map parameterization of the stochastic variables are combined with the autoregressive PixelCNN approach. Interestingly, we also observe close to state-of-the-art results without the autoregressive part. This opens the possibility for high quality image generation with only one forward-pass. |
| Tasks | Image Generation, Omniglot |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=Hy_o3x-0b |
| https://openreview.net/pdf?id=Hy_o3x-0b | |
| PWC | https://paperswithcode.com/paper/feature-map-variational-auto-encoders |
| Repo | |
| Framework | |
Reading Comprehension with Graph-based Temporal-Casual Reasoning
| Title | Reading Comprehension with Graph-based Temporal-Casual Reasoning |
| Authors | Yawei Sun, Gong Cheng, Yuzhong Qu |
| Abstract | Complex questions in reading comprehension tasks require integrating information from multiple sentences. In this work, to answer such questions involving temporal and causal relations, we generate event graphs from text based on dependencies, and rank answers by aligning event graphs. In particular, the alignments are constrained by graph-based reasoning to ensure temporal and causal agreement. Our focused approach self-adaptively complements existing solutions; it is automatically triggered only when applicable. Experiments on RACE and MCTest show that state-of-the-art methods are notably improved by using our approach as an add-on. |
| Tasks | Dependency Parsing, Reading Comprehension |
| Published | 2018-08-01 |
| URL | https://www.aclweb.org/anthology/C18-1069/ |
| https://www.aclweb.org/anthology/C18-1069 | |
| PWC | https://paperswithcode.com/paper/reading-comprehension-with-graph-based |
| Repo | |
| Framework | |
