Entropy-guided feature generation for structured learning of Portuguese dependency parsing

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Standard

Entropy-guided feature generation for structured learning of Portuguese dependency parsing. / Fernandes, Eraldo R.; Milidiú, Ruy L.
Computational Processing of the Portuguese Language: 10th International Conference, PROPOR 2012, Coimbra, Portugal, April 17-20, 2012. Proceedings. ed. / Helena Caseli; Aline Villavicencio; Antonio Teixeira; Fernando Perdigao. Berlin, Heidelberg: Springer, 2012. p. 146-156 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7243 LNAI).

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Harvard

Fernandes, ER & Milidiú, RL 2012, Entropy-guided feature generation for structured learning of Portuguese dependency parsing. in H Caseli, A Villavicencio, A Teixeira & F Perdigao (eds), Computational Processing of the Portuguese Language: 10th International Conference, PROPOR 2012, Coimbra, Portugal, April 17-20, 2012. Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7243 LNAI, Springer, Berlin, Heidelberg, pp. 146-156, International Conference on Computational Processing of Portuguese, Coimbra, Portugal, 17.04.12. https://doi.org/10.1007/978-3-642-28885-2_17

APA

Fernandes, E. R., & Milidiú, R. L. (2012). Entropy-guided feature generation for structured learning of Portuguese dependency parsing. In H. Caseli, A. Villavicencio, A. Teixeira, & F. Perdigao (Eds.), Computational Processing of the Portuguese Language: 10th International Conference, PROPOR 2012, Coimbra, Portugal, April 17-20, 2012. Proceedings (pp. 146-156). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7243 LNAI). Springer. https://doi.org/10.1007/978-3-642-28885-2_17

Vancouver

Fernandes ER, Milidiú RL. Entropy-guided feature generation for structured learning of Portuguese dependency parsing. In Caseli H, Villavicencio A, Teixeira A, Perdigao F, editors, Computational Processing of the Portuguese Language: 10th International Conference, PROPOR 2012, Coimbra, Portugal, April 17-20, 2012. Proceedings. Berlin, Heidelberg: Springer. 2012. p. 146-156. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-28885-2_17

Bibtex

@inbook{f536d12ded6d493bb6a28d5057fd8d2a,
title = "Entropy-guided feature generation for structured learning of Portuguese dependency parsing",
abstract = "Feature generation is a difficult, yet highly necessary, subtask of machine learning modeling. Usually, it is partially solved by a domain expert that generates complex and discriminative feature templates by conjoining the available basic features. This is a limited and expensive way to obtain feature templates and is recognized as a modeling bottleneck. In this work, we propose an automatic method to generate feature templates for structured learning algorithms. The method receives as input the training dataset with basic features and produces a set of feature templates by conjoining basic features that are highly discriminative together. We denote this method entropy guided since it is based on the conditional entropy of local decision variables given the feature values. We illustrate our approach on the Portuguese dependency parsing task and report on experiments with the Bosque corpus. We show that the entropy-guided templates outperform the manually built templates used by MSTParser, which was the best performing system on the Bosque corpus up to now. Furthermore, our approach allows an effortless inclusion of two new basic features that automatically generate additional templates. As a result, our system achieves a per-token accuracy of 92.66%, what represents a reduction by more than 15% on the previous smallest error rate for Portuguese dependency parsing.",
keywords = "dependency parsing, entropy-guided feature generation, machine learning, structured learning, Informatics, Business informatics",
author = "Fernandes, {Eraldo R.} and Milidi{\'u}, {Ruy L.}",
year = "2012",
doi = "10.1007/978-3-642-28885-2_17",
language = "English",
isbn = "978-3-642-28884-5",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer",
pages = "146--156",
editor = "Helena Caseli and Aline Villavicencio and Antonio Teixeira and Fernando Perdigao",
booktitle = "Computational Processing of the Portuguese Language",
address = "Germany",
note = "International Conference on Computational Processing of Portuguese, PROPOR 2012 ; Conference date: 17-04-2012 Through 20-04-2012",
url = "https://aclweb.org/portal/content/10th-international-conference-computational-processing-portuguese-propor-2012",

}

RIS

TY - CHAP

T1 - Entropy-guided feature generation for structured learning of Portuguese dependency parsing

AU - Fernandes, Eraldo R.

AU - Milidiú, Ruy L.

N1 - Conference code: 10

PY - 2012

Y1 - 2012

N2 - Feature generation is a difficult, yet highly necessary, subtask of machine learning modeling. Usually, it is partially solved by a domain expert that generates complex and discriminative feature templates by conjoining the available basic features. This is a limited and expensive way to obtain feature templates and is recognized as a modeling bottleneck. In this work, we propose an automatic method to generate feature templates for structured learning algorithms. The method receives as input the training dataset with basic features and produces a set of feature templates by conjoining basic features that are highly discriminative together. We denote this method entropy guided since it is based on the conditional entropy of local decision variables given the feature values. We illustrate our approach on the Portuguese dependency parsing task and report on experiments with the Bosque corpus. We show that the entropy-guided templates outperform the manually built templates used by MSTParser, which was the best performing system on the Bosque corpus up to now. Furthermore, our approach allows an effortless inclusion of two new basic features that automatically generate additional templates. As a result, our system achieves a per-token accuracy of 92.66%, what represents a reduction by more than 15% on the previous smallest error rate for Portuguese dependency parsing.

AB - Feature generation is a difficult, yet highly necessary, subtask of machine learning modeling. Usually, it is partially solved by a domain expert that generates complex and discriminative feature templates by conjoining the available basic features. This is a limited and expensive way to obtain feature templates and is recognized as a modeling bottleneck. In this work, we propose an automatic method to generate feature templates for structured learning algorithms. The method receives as input the training dataset with basic features and produces a set of feature templates by conjoining basic features that are highly discriminative together. We denote this method entropy guided since it is based on the conditional entropy of local decision variables given the feature values. We illustrate our approach on the Portuguese dependency parsing task and report on experiments with the Bosque corpus. We show that the entropy-guided templates outperform the manually built templates used by MSTParser, which was the best performing system on the Bosque corpus up to now. Furthermore, our approach allows an effortless inclusion of two new basic features that automatically generate additional templates. As a result, our system achieves a per-token accuracy of 92.66%, what represents a reduction by more than 15% on the previous smallest error rate for Portuguese dependency parsing.

KW - dependency parsing

KW - entropy-guided feature generation

KW - machine learning

KW - structured learning

KW - Informatics

KW - Business informatics

UR - http://www.scopus.com/inward/record.url?scp=84858599304&partnerID=8YFLogxK

UR - https://d-nb.info/1019948167

U2 - 10.1007/978-3-642-28885-2_17

DO - 10.1007/978-3-642-28885-2_17

M3 - Article in conference proceedings

AN - SCOPUS:84858599304

SN - 978-3-642-28884-5

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 146

EP - 156

BT - Computational Processing of the Portuguese Language

A2 - Caseli, Helena

A2 - Villavicencio, Aline

A2 - Teixeira, Antonio

A2 - Perdigao, Fernando

PB - Springer

CY - Berlin, Heidelberg

T2 - International Conference on Computational Processing of Portuguese

Y2 - 17 April 2012 through 20 April 2012

ER -

Recently viewed

Publications

  1. Evaluation of Time/Phase Parameters in Frequency Measurements for Inertial Navigation Systems
  2. A discrete approximate solution for the asymptotic tracking problem in affine nonlinear systems
  3. Neural network-based adaptive fault-tolerant control for strict-feedback nonlinear systems with input dead zone and saturation
  4. Multi-Parallel Sending Coils for Movable Receivers in Inductive Charging Systems
  5. The Use of Factorization and Multimode Parametric Spectra in Estimating Frequency and Spectral Parameters of Signal
  6. Perfect anti-windup in output tracking scheme with preaction
  7. Control of the inverse pendulum based on sliding mode and model predictive control
  8. Enhancing Performance of Level System Modeling with Pseudo-Random Signals
  9. Using Complexity Metrics to Assess Silent Reading Fluency
  10. Continuous 3D scanning mode using servomotors instead of stepping motors in dynamic laser triangulation
  11. Digital Control of a Camless Engine Using Lyapunov Approach with Backward Euler Approximation
  12. Analyzing different types of moderated method effects in confirmatory factor models for structurally different methods
  13. Using the flatness of DC-Drives to emulate a generator for a decoupled MPC using a geometric approach for motion control in Robotino
  14. Dynamic Lot Size Optimization with Reinforcement Learning
  15. On robustness properties in permanent magnet machine control by using decoupling controller
  16. Classical PI Controllers with Anti-Windup Techniques Applied on Level Systems
  17. A model predictive control in Robotino and its implementation using ROS system
  18. Introducing parametric uncertainty into a nonlinear friction model
  19. Stepwise-based optimizing approaches for arrangements of loudspeaker in multi-zone sound field reproduction
  20. A geometric approach for controlling an electromagnetic actuator with the help of a linear Model Predictive Control
  21. A localized boundary element method for the floating body problem
  22. Mapping interest rate projections using neural networks under cointegration
  23. The Influence of Note-taking on Mathematical Solution Processes while Working on Reality-Based Tasks
  24. Robust Flatness Based Control of an Electromagnetic Linear Actuator Using Adaptive PID Controller
  25. Gaussian processes for dispatching rule selection in production scheduling
  26. Performance analysis for loss systems with many subscribers and concurrent services
  27. Comments on "Tracking Control of Robotic Manipulators With Uncertain Kinematics and Dynamics"
  28. A guided simulated annealing search for solving the pick-up and delivery problem with time windows and capacity constraints
  29. An analytical approach to evaluating bivariate functions of fuzzy numbers with one local extremum
  30. On the Nonlinearity Compensation in Permanent Magnet Machine Using a Controller Based on a Controlled Invariant Subspace
  31. An Orthogonal Wavelet Denoising Algorithm for Surface Images of Atomic Force Microscopy
  32. Stability analysis of a linear model predictive control and its application in a water recovery process
  33. Robust Control of Mobile Transportation Object with 3D Technical Vision System
  34. Data-Driven flood detection using neural networks
  35. Passive Peak Voltage Sensor for Multiple Sending Coils Inductive Power Transmission System
  36. A Gait Pattern Generator for Closed-Loop Position Control of a Soft Walking Robot
  37. A two-stage Kalman estimator for motion control using model predictive strategy
  38. A general structural property in wavelet packets for detecting oscillation and noise components in signal analysis
  39. A denoising procedure using wavelet packets for instantaneous detection of pantograph oscillations
  40. Simulation based comparison of safety-stock calculation methods
  41. Primary Side Circuit Design of a Multi-coil Inductive System for Powering Wireless Sensors
  42. Continuous and Discrete Concepts for Detecting Transport Barriers in the Planar Circular Restricted Three Body Problem
  43. Convolutional Neural Networks
  44. A New Framework for Production Planning and Control to Support the Positioning in Fields of Tension Created by Opposing Logistic Objectives
  45. Cognitive load and instructionally supported learning with provided and learner-generated visualizations
  46. Using cross-recurrence quantification analysis to compute similarity measures for time series of unequal length with applications to sleep stage analysis
  47. PI and Fuzzy Controllers for Non-Linear Systems
  48. Long-term memory predictors of adult language learning at the interface between syntactic form and meaning