Intro (feel free to skip this)

I feel like we are living in the golden age of Natural Language Processing (NLP). The last 20 years have seen the rise of Neural Language models, the creation and subsequent vast application of Word Embeddings and innovative Network architectures for Sequence Modeling and, last but not least, transformers.

If you did not get the detail of the previous paragraph, please bear with me. The motivation of this post is showing how a non-specialist (like me) can today build effective software that is able to “use” or interpret text written by humans. To our grandparents, google would look like magic.

Here, I just scratch the surface of NLP models, by commenting my colab notebook which demostrates how to build a sentiment movie review app from the ground up exploiting Keras’ TextVectorization.

Code

for i in range(3):
    print(i)




print("hello")

Code below

print("hello world!")

Footnotes how to

$$\int_{a}^{b} x^2 dx$$

Here’s a simple footnote,¹ and here’s a longer one.². Here another one with a mismatched numbering ³

Inline eqn

The following is an inline equation $$\int x^3 dx$$

Other publications

• Dynamic Word Embeddings