1. Abstract
2. Introduction
3. Neural Networks: What are they?
4. Parsing a sentence is a process that occurs through time.
5. Some examples of NN for language processing
5.1 Rummelhart and McClelland
5.2 Gasser and Lee
5.3 Elman
5.4 Koncar and Guthrie
5.5 Stolcke (1990)
5.6 Allen
5.7 Munro, Cosic, and Tabasko
5.8 Jain
5.9 Miikkulainen
5.10 Nenov and Dyer
6. Important issues
6.1 Number of hidden nodes
6.2 Topology
6.3 Incremental Learning
6.4 Learning Algorithm
6.4.1 Recurrent Backpropagation
6.4.2 Backpropagation Through Time (BPTT)
6.4.3 Backpropagation with Momentum
6.4.4 Quickprop
6.4.5 Resilient backpropagation
6.4.7 Teacher Forcing
6.5 Semantical grounding
7. Genetic Algorithms
8. Scope and direction of Research
8.1 Description of the Neural Network.
8.1.1 The Input Nodes.
8.1.2 The Output Nodes.
8.1.3 The Hidden Layer.
8.2 Training the network.
8.3 Evaluating the performance.
8.4 Evaluating the results.
8.4.1 Validation.
8.4.2 Additional Genetic Runs.
8.4.3 Observing the Sentence Processing.
8.5 Running the experiment.
9 Conclusion
Bibliography
Appendix A Words in the Vocabulary (35 in total)
Appendix B Sentences in the Language
Appendix C Definition of the Genotype
Appendix D List of possible relationships between phrases of the sentences
Appendix E Input Layer
Appendix F Output Layer
|