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

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

Authors

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.

Original languageEnglish
Title of host publicationComputational Processing of the Portuguese Language : 10th International Conference, PROPOR 2012, Coimbra, Portugal, April 17-20, 2012. Proceedings
EditorsHelena Caseli, Aline Villavicencio, Antonio Teixeira, Fernando Perdigao
Number of pages11
Place of PublicationBerlin, Heidelberg
PublisherSpringer Verlag
Publication date2012
Pages146-156
ISBN (print)978-3-642-28884-5
ISBN (electronic)978-3-642-28885-2
DOIs
Publication statusPublished - 2012
Externally publishedYes
EventInternational Conference on Computational Processing of Portuguese - Coimbra, Portugal
Duration: 17.04.201220.04.2012
Conference number: 10
https://aclweb.org/portal/content/10th-international-conference-computational-processing-portuguese-propor-2012

Recently viewed

Activities

  1. Presentation of the paper entitled: "Combining a PI Controller with an Adaptive Feedforward Control in PMSM"
  2. Presentation of the paper entitled: "Case Study: Aspects of Fuzzy Controller Implementation in Embedded Systems"
  3. Combining flatness based feedforward action with a fractional PI regulator to control the intake valve engine
  4. All Surface: Blobs and the Liquefaction of Architecture
  5. Comparing Two Voltage Observers in a Sensorsystem using Repetitive Control
  6. Feedback in the context of digital media: The effectiveness of a mathematics teaching-learning platform and its usage and perception by students
  7. Probabilistic and discrete methods for the computational study of coherent behavior in flows
  8. Can the ability to identify criteria explain why some selection procedures work? Results and unresolved issues
  9. A Dynamic Signal Analyzer. Analysis and Synthesis of Speech at the Biological Computer Laboratory
  10. Optimal trajectory generation using MPC in robotino and its implementation with ROS system
  11. Keynote speech entitled: "A Stabilizing Control Strategy for a Bank System using State Space and Sliding Mode Control Approach with an Extended Kalman Filter"
  12. Enhancing metacognition by using flipping classroom with geogebra
  13. An MPC for an Aggregate Actuator with a Self-Tuning Feedforward Control
  14. From Projects and Formats to Communities
  15. A sufficient asymptotic stability condition in generalised model predictive control to avoid input saturation
  16. Revisiting the concept of the script in institutional theory
  17. Revitalizing the script as a concept to understand structure and agency in institutional theory

Publications

  1. Learning Rotation Sensitive Neural Network for Deformed Objects' Detection in Fisheye Images
  2. A multi input sliding mode control for Peltier Cells using a cold-hot sliding surface
  3. Digital Control of a Camless Engine Using Lyapunov Approach with Backward Euler Approximation
  4. Unidimensional and Multidimensional Methods for Recurrence Quantification Analysis with crqa
  5. Different approaches to learning from errors: Comparing the effectiveness of high reliability and error management approaches
  6. Evaluating the construct validity of Objective Personality Tests using a multitrait-multimethod-Multioccasion-(MTMM-MO)-approach
  7. Cross-document coreference resolution using latent features
  8. The Use of Genetic Algorithm for PID Controller Auto-Tuning in ARM CORTEX M4 Platform
  9. Methodologies for Noise and Gross Error Detection using Univariate Signal-Based Approaches in Industrial Application
  10. Binary Random Nets I
  11. Methodologies for noise and gross error detection using univariate signal-based approaches in industrial applications
  12. Evolutionary generation of dispatching rule sets for complex dynamic scheduling problems
  13. Ant colony optimization algorithm and artificial immune system applied to a robot route
  14. Development of a Didactic Graphical Simulation Interface on MATLAB for Systems Control
  15. Knowledge Graph Question Answering Using Graph-Pattern Isomorphism
  16. Modified dynamic programming approach for offline segmentation of long hydrometeorological time series
  17. Using Euler Discrete Approximation to Control an Aggregate Actuator in Camless Engines
  18. Random measurement and prediction errors limit the practical relevance of two velocity sensors to estimate the 1RM back squat
  19. Framework for the Parallelized Development of Estimation Tasks for Length, Area, Capacity and Volume in Primary School - A Pilot Study
  20. Using protochirons for three-dimensional coding of certain chemical structures.
  21. Substructure, subgraph, and walk counts as measures of the complexity of graphs and molecules.
  22. Multidimensional recurrence quantification analysis (MdRQA) for the analysis of multidimensional time-series
  23. Evaluation of Time/Phase Parameters in Frequency Measurements for Inertial Navigation Systems
  24. Application of non-convex rate dependent gradient plasticity to the modeling and simulation of inelastic microstructure development and inhomogeneous material behavior
  25. An MPC for an Aggregate Actuator with a Self-Tuning Feedforward Control
  26. Model inversion using fuzzy neural network with boosting of the solution
  27. A model predictive control for an aggregate actuator with a self-tuning initial condition procedure in combustion engines
  28. A discrete approximate solution for the asymptotic tracking problem in affine nonlinear systems
  29. Neural network-based adaptive fault-tolerant control for strict-feedback nonlinear systems with input dead zone and saturation
  30. Unity and diversity in the law of state responsibility
  31. On the Nonlinearity Compensation in Permanent Magnet Machine Using a Controller Based on a Controlled Invariant Subspace
  32. Control condition design and implementation features in controlled trials