Research gaps in neuro-symbolic computing
Although the field of NSAI has grown significantly in research interest and popularity, it is still maturing. NSAI is proving to be rather promising and is seeing major financial investment from strong research and technical institutions such as MIT and IBM. In this section, we will briefly highlight the future NSAI research direction and study avenues of interest:
- Standardized datasets and comprehensibility tests for benchmarking:
Like the existing CLEVR and CLEVRER datasets, the domain of NSAI would benefit from common datasets that can be used to benchmark existing and new techniques. These standardized resources are critical for comparative research.
- Strategies for NN inference using symbolic propositions.
- More comparative evaluations on the NSAI paradigm to determine the best NN architecture to use.
- Context-aware semantics:
Further investigation on how semantics can be constructed and influenced by...
