Neuromorphic Natural Language Processing Workgroup Results

Make a page for each project below, and edit titles as needed:

Overall Summary:

1. learning word meanings
- distributed representations modified for spiking architectures
- learning word relations with hebbian learning
2. reasoning about words
- word similarities
- word analogies
- Happiness, or predicting sentiment on TrueNorth
3. understanding sequences of words
- question classification via SKIM
- Spike-Based Recurrent NN for Question Classification
4. reasoning with memory
- Spike-based parsing with cell assemblies
- Q&A with working memory in Nengo
- question answering system
- Semantic spreading activation network