Some cunning NLP research which has found that really simple statistical models (e.g. does the sentence contain "not"?) is sufficient to answer a good number of text comprehension datasets: https://thegradient.pub/nlps-clever-hans-moment-has-arrived/
miniblog.
Related Posts
I'm utterly torn between ido and helm. I like that ido has a small popup, preserving window layout, but helm is great for large datasets.
Elasticsearch gets better with every release. 1.2 has been very reliable on our write-heavy cluster with big datasets (1.1 struggled).
One of the fun things about machine learning is you have large datasets. A naive algorithm often isn't fast enough and forces optimisation.