Day 76 – Predicting Similarity Between Words Using spaCy Natural Language Processing – Part IV

Predicting similarities between two words used in much daily life e-g YouTube recommendations, Production search recommendation, NetFlix movie recommendation, etc., based on our watchlist or history.

We will look into a similar kind of simple code to predict similarities between two words.

We have seen the below posts on spaCy in our previous blogs.

• What is spaCy, usage, and features?
• Sentence Analyzer Using spaCy Natural Language Processing – Part I
• Detail Sentence Analyzer Using spaCy Natural Language Processing – Part II
• Named Entity Recognition Using spaCy Natural Language Processing – Part III

In this blog, we will predict similarities between two words and score them.

I am using Google Colab for this named entity recognition to check, and you can find a link here.

Image Source: spaCY

bicyle car 0.695734679932836
movie cinema 0.6450227275334257

We can see similarity between bicycle and car score 0.69 and similarity between movie and cinema 0.64.

Let me know your comments what you like on spaCy similarity model and how we can use to built robust systems.