A Sequence-to-Sequence Model for Semantic Role Labeling

Daza, Angel; Frank, Anette

doi:10.18653/v1/w18-3027

Cited by 19 publications

(25 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These two additions already show improvements compared to the reported results inDaza and Frank (2018).…”

mentioning

confidence: 50%

“…We define the SRL task as a sequence transduction problem: given an input sequence of tokens X = x 1 , ..., x i , the system is tasked to generate a sequence Y = y 1 , ..., y j consisting of words interleaved with SRL annotations. Defining the task in this fashion allows X and Y to be of different lengths and therefore target sequences may also (Daza and Frank, 2018). We generalize this architecture to multilingual and cross-lingual SRL.…”

Section: One Model To Treat Them Allmentioning

confidence: 99%

“…Encoder. In our prior work (Daza and Frank, 2018) we used a 2-layer BiLSTM as encoder. In this paper, we adopt the Deep BiLSTM Encoder from He et al (2017) which has been shown to work well for SRL models.…”

Section: Encoder-decoder Architecturementioning

confidence: 99%

“…In this work, we propose a general Encoder-Decoder (Enc-Dec) architecture for SRL (see Figure 1). We extend our previous Enc-Dec approach for SRL (Daza and Frank, 2018) to a cross-lingual model that translates sentences from a source language to a (lower-resource) target language, and during decoding jointly labels it with SRL annotations. 1 Our contributions are as follows:…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Translate and Label! An Encoder-Decoder Approach for Cross-lingual Semantic Role Labeling

Daza

Frank

2019

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conferen

Self Cite

View full text Add to dashboard Cite

We propose a Cross-lingual Encoder-Decoder model that simultaneously translates and generates sentences with Semantic Role Labeling annotations in a resource-poor target language. Unlike annotation projection techniques, our model does not need parallel data during inference time. Our approach can be applied in monolingual, multilingual and cross-lingual settings and is able to produce dependencybased and span-based SRL annotations. We benchmark the labeling performance of our model in different monolingual and multilingual settings using well-known SRL datasets. We then train our model in a cross-lingual setting to generate new SRL labeled data. Finally, we measure the effectiveness of our method by using the generated data to augment the training basis for resource-poor languages and perform manual evaluation to show that it produces high-quality sentences and assigns accurate semantic role annotations. Our proposed architecture offers a flexible method for leveraging SRL data in multiple languages.

show abstract

“…These two additions already show improvements compared to the reported results inDaza and Frank (2018).…”

mentioning

confidence: 50%

Section: One Model To Treat Them Allmentioning

confidence: 99%

Section: Encoder-decoder Architecturementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Translate and Label! An Encoder-Decoder Approach for Cross-lingual Semantic Role Labeling

Daza

Frank

2019

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conferen

Self Cite

View full text Add to dashboard Cite

show abstract

“…We found no attempt to adapt trend filtering to an online setup where data becomes available as time passes. On the other hand, sequence-to-sequence methods, see [34], have been successfully applied in a Natural Language Processing (NLP) context, for translation [2], text generation using Generative Adversarial Networks (GAN) like [26], speech recognition in [13], semantic labelling in [7]...Yet, these methods are supervised methods where the labels are given and unambiguous. Also, classification of time series is well established, see [20] for a recent survey.…”

Section: Background and Related Workmentioning

confidence: 99%

An Empirical Study of Neural Networks for Trend Detection in Time Series

Miot¹,

Drigout²

2020

SN COMPUT. SCI.

View full text Add to dashboard Cite

Even if trend is probably one of the most intuitive notions in time series dynamics, this notion is usually ambiguous and model dependent. We first cast the trend detection problem into a sequence-to-sequence classification problem. Then, we simulate various dynamics with labelled trends. Using those simulated time-series we build a baseline trend estimator showing good performance on various dynamics. Comparing this baseline estimator with various other trend estimators, we find that some recurrent neural networks structures compare favourably against other estimators including convolutional neural networks. Those sequence-to-sequence trend classifiers could be used as efficient basic blocks to build more complex time series estimators.

show abstract

Tamil Paraphrase Detection Using Encoder-Decoder Neural Networks

Kumar¹,

Thenmozhi²,

Subramanian³

2020

IFIP Advances in Information and Communication Technology

View full text Add to dashboard Cite

Detecting paraphrases in Indian languages require critical anlysis on the lexical, syntactic and semantic features. Since the structure of Indian languages differ from the other languages like English, the usage of lexico-syntactic features vary between the Indian languages and plays a critical role in determining the performance of the system. Instead of using various lexico-syntactic similarity features, we aim to apply a complete end-to-end system using deep learning networks with no lexicosyntactic features. In this paper we exploited the encoder-decoder model of deep neural network to analyze the paraphrase sentences in Tamil language and to classify. In this encoder-decoder model, LSTM, GRU units and gNMT are used as layers along with attention mechanism. Using this end-to-end model, there is an increase in f1-measure by 0.5% for the subtask-1 when compared to the state-of-the-art systems. The system was trained and evaluated on DPIL@FIRE2016 Shared Task dataset. To our knowledge, ours is the first deep learning model which validates the training instances of both the subtask-1 and subtask-2 dataset of DPIL shared task.

show abstract

A Sequence-to-Sequence Model for Semantic Role Labeling

Cited by 19 publications

References 24 publications

Translate and Label! An Encoder-Decoder Approach for Cross-lingual Semantic Role Labeling

Translate and Label! An Encoder-Decoder Approach for Cross-lingual Semantic Role Labeling

An Empirical Study of Neural Networks for Trend Detection in Time Series

Tamil Paraphrase Detection Using Encoder-Decoder Neural Networks

Contact Info

Product

Resources

About