scholarly journals HC-Search: A Learning Framework for Search-based Structured Prediction

2014 ◽  
Vol 50 ◽  
pp. 369-407 ◽  
Author(s):  
J.R. Doppa ◽  
A. Fern ◽  
P. Tadepalli
Keyword(s):  
Semantic Segmentation ◽  
Structured Prediction ◽  
Training Procedure ◽  
High Quality ◽  
Structured Input ◽  
Learning Framework ◽  
Part Of Speech Tagging ◽  
Part Of Speech ◽  
Segmentation Of Images ◽  
Structured Outputs

Structured prediction is the problem of learning a function that maps structured inputs to structured outputs. Prototypical examples of structured prediction include part-of-speech tagging and semantic segmentation of images. Inspired by the recent successes of search-based structured prediction, we introduce a new framework for structured prediction called HC-Search. Given a structured input, the framework uses a search procedure guided by a learned heuristic H to uncover high quality candidate outputs and then employs a separate learned cost function C to select a final prediction among those outputs. The overall loss of this prediction architecture decomposes into the loss due to H not leading to high quality outputs, and the loss due to C not selecting the best among the generated outputs. Guided by this decomposition, we minimize the overall loss in a greedy stage-wise manner by first training H to quickly uncover high quality outputs via imitation learning, and then training C to correctly rank the outputs generated via H according to their true losses. Importantly, this training procedure is sensitive to the particular loss function of interest and the time-bound allowed for predictions. Experiments on several benchmark domains show that our approach significantly outperforms several state-of-the-art methods.

scholarly journals Learning and Inference for Structured Prediction: A Unifying Perspective

2019 ◽  
Author(s):  
Aryan Deshwal ◽  
Janardhan Rao Doppa ◽  
Dan Roth
Keyword(s):  
Semantic Segmentation ◽  
Input Image ◽  
Structured Prediction ◽  
Classification Problems ◽  
Structured Input ◽  
Part Of Speech ◽  
Input Sentence ◽  
Weakly Supervised ◽  
Segmentation Of Images ◽  
Prediction Problems

In a structured prediction problem, one needs to learn a predictor that, given a structured input, produces a structured object, such as a sequence, tree, or clustering output. Prototypical structured prediction tasks include part-of-speech tagging (predicting POS tag sequence for an input sentence) and semantic segmentation of images (predicting semantic labels for pixels of an input image). Unlike simple classification problems, here there is a need to assign values to multiple output variables accounting for the dependencies between them. Consequently, the prediction step itself (aka ``inference" or ``decoding") is computationally-expensive, and so is the learning process, that typically requires making predictions as part of it. The key learning and inference challenge is due to the exponential size of the structured output space and depend on its complexity. In this paper, we present a unifying perspective of the different frameworks that address structured prediction problems and compare them in terms of their strengths and weaknesses. We also discuss important research directions including integration of deep learning advances into structured prediction, and learning from weakly supervised signals and active querying to overcome the challenges of building structured predictors from small amount of labeled data.

scholarly journals Randomized Greedy Search for Structured Prediction: Amortized Inference and Learning

2019 ◽  
Author(s):  
Chao Ma ◽  
F A Rezaur Rahman Chowdhury ◽  
Aryan Deshwal ◽  
Md Rakibul Islam ◽  
Janardhan Rao Doppa ◽  
...  
Keyword(s):  
Structured Prediction ◽  
Greedy Search ◽  
Computationally Efficient ◽  
Structured Input ◽  
Part Of Speech Tagging ◽  
Part Of Speech ◽  
Training Approach ◽  
Complex Models ◽  
Simple Class ◽  
Structured Output

In a structured prediction problem, we need to learn a predictor that can produce a structured output given a structured input (e.g., part-of-speech tagging). The key learning and inference challenge is due to the exponential size of the structured output space. This paper makes four contributions towards the goal of a computationally-efficient inference and training approach for structured prediction that allows to employ complex models and to optimize for non-decomposable loss functions. First, we define a simple class of randomized greedy search (RGS) based inference procedures that leverage classification algorithms for simple outputs. Second, we develop a RGS specific learning approach for amortized inference that can quickly produce high-quality outputs for a given set of structured inputs. Third, we plug our amortized RGS inference solver inside the inner loop of parameter-learning algorithms (e.g., structured SVM) to improve the speed of training. Fourth, we perform extensive experiments on diverse structured prediction tasks. Results show that our proposed approach is competitive or better than many state-of-the-art approaches in spite of its simplicity.

scholarly journals Learning from Disagreement: A Survey

2021 ◽  
Vol 72 ◽  
pp. 1385-1470
Author(s):  
Alexandra N. Uma ◽  
Tommaso Fornaciari ◽  
Dirk Hovy ◽  
Silviu Paun ◽  
Barbara Plank ◽  
...  
Keyword(s):  
Language Processing ◽  
Gold Standard ◽  
Training Methods ◽  
High Quality ◽  
Training Models ◽  
Part Of Speech Tagging ◽  
Part Of Speech ◽  
Growing Body ◽  
Research Questions ◽  
Speech Tagging

Many tasks in Natural Language Processing (NLP) and Computer Vision (CV) offer evidence that humans disagree, from objective tasks such as part-of-speech tagging to more subjective tasks such as classifying an image or deciding whether a proposition follows from certain premises. While most learning in artificial intelligence (AI) still relies on the assumption that a single (gold) interpretation exists for each item, a growing body of research aims to develop learning methods that do not rely on this assumption. In this survey, we review the evidence for disagreements on NLP and CV tasks, focusing on tasks for which substantial datasets containing this information have been created. We discuss the most popular approaches to training models from datasets containing multiple judgments potentially in disagreement. We systematically compare these different approaches by training them with each of the available datasets, considering several ways to evaluate the resulting models. Finally, we discuss the results in depth, focusing on four key research questions, and assess how the type of evaluation and the characteristics of a dataset determine the answers to these questions. Our results suggest, first of all, that even if we abandon the assumption of a gold standard, it is still essential to reach a consensus on how to evaluate models. This is because the relative performance of the various training methods is critically affected by the chosen form of evaluation. Secondly, we observed a strong dataset effect. With substantial datasets, providing many judgments by high-quality coders for each item, training directly with soft labels achieved better results than training from aggregated or even gold labels. This result holds for both hard and soft evaluation. But when the above conditions do not hold, leveraging both gold and soft labels generally achieved the best results in the hard evaluation. All datasets and models employed in this paper are freely available as supplementary materials.

scholarly journals Learning syntactic tagging of Macedonian language

2018 ◽  
Vol 15 (3) ◽  
pp. 799-820
Author(s):  
Martin Bonchanoski ◽  
Katerina Zdravkova
Keyword(s):  
Dynamic Features ◽  
Slavic Languages ◽  
Learning Framework ◽  
Part Of Speech Tagging ◽  
Part Of Speech ◽  
Guided Learning ◽  
Initial Stage ◽  
Dependency Network ◽  
Pos Tagger ◽  
Speech Tagging

This paper presents the creation of machine learning based systems for Part-of-speech tagging of Macedonian language. Four well-known PoS tagger systems implemented for English and Slavic languages: TnT, cyclic dependency network, guided learning framework for bidirectional sequence classification, and dynamic features induction were trained. Orwell?s novel ?1984? was manually tagged from the authors and it was used split into training and test set. After the training of the models, a comparison between the models was made. At the end, a POS tagger with an accuracy that reaches 97.5% was achieved, making it very appropriate for the future grammatical tagging of the National corpus of Macedonian language, which is currently in its initial stage. The Part-of-speech tagger that was create is published online and free to use.

Lexical Rule and Lexicon Effect for Part of Speech Tagging Bahasa Madura

2018 ◽  
Vol 18 (1) ◽  
pp. 65-72
Author(s):  
Nindian Puspa Dewi ◽  
Ubaidi Ubaidi
Keyword(s):  
Text Processing ◽  
Part Of Speech Tagging ◽  
Pos Tagging ◽  
Part Of Speech ◽  
Speech Tagging ◽  
Bahasa Indonesia

POS Tagging adalah dasar untuk pengembangan Text Processing suatu bahasa. Dalam penelitian ini kita meneliti pengaruh penggunaan lexicon dan perubahan morfologi kata dalam penentuan tagset yang tepat untuk suatu kata. Aturan dengan pendekatan morfologi kata seperti awalan, akhiran, dan sisipan biasa disebut sebagai lexical rule. Penelitian ini menerapkan lexical rule hasil learner dengan menggunakan algoritma Brill Tagger. Bahasa Madura adalah bahasa daerah yang digunakan di Pulau Madura dan beberapa pulau lainnya di Jawa Timur. Objek penelitian ini menggunakan Bahasa Madura yang memiliki banyak sekali variasi afiksasi dibandingkan dengan Bahasa Indonesia. Pada penelitian ini, lexicon selain digunakan untuk pencarian kata dasar Bahasa Madura juga digunakan sebagai salah satu tahap pemberian POS Tagging. Hasil ujicoba dengan menggunakan lexicon mencapai akurasi yaitu 86.61% sedangkan jika tidak menggunakan lexicon hanya mencapai akurasi 28.95 %. Dari sini dapat disimpulkan bahwa ternyata lexicon sangat berpengaruh terhadap POS Tagging.

scholarly journals The Impact of Arabic Part of Speech Tagging on Sentiment Analysis: A New Corpus and Deep Learning Approach

2021 ◽  
Vol 184 ◽  
pp. 148-155
Author(s):  
Abdul Munem Nerabie ◽  
Manar AlKhatib ◽  
Sujith Samuel Mathew ◽  
May El Barachi ◽  
Farhad Oroumchian
Keyword(s):  
Deep Learning ◽  
Sentiment Analysis ◽  
Learning Approach ◽  
Part Of Speech Tagging ◽  
Part Of Speech ◽  
The Impact ◽  
Speech Tagging
Sign in / Sign up

Export Citation Format

Share Document