Mohammadjavad Vakili

Mohammadjavad Vakili

Physicist, Data Scientist

An experiment with named entity recognition

1 minute read

Published:

Named entity recognition (NER) is the practice of recognizing names in a given body of text and identifying the entity category associated with them: persons, locations, etc.

Overview

Model applied to the comment section of New York Times articles

There are many pretrained NER models in the literature. We make use of the spaCy en_core_web_sm model. This model is a CNN model trained on the OntoNotes data. This CNN model carries out multiple tasks such as Part of the Speech Tag and recognition of named entities.

We use the New York Times API to fetch the comments of a given nytimes article whose link is provided by the user. Once the comments are feteched, we use the en_core_web_sm model to identify the named entities in all the comments.

Sentiment

Then we use a rule-based sentiment analayzer to find the sentiment polarity associated with the most common named entities mentioned in the comments.

Serving on a browser

We use streamlit to serve the model in an app that can be hosted on the browser. Let’s have a look at how the results look like:

In the example shown above, we see a comment in which three named entities are recognized: Action (a person), Ohio (GPE, a geographical location) and New York (GPE).

In the figure shown above, we see the distribution of sentiment scores associated with the named entity Acton in the comments. The scores are bounded between -1 and 1. The higher (lower) score values correspond to more positive (negative) sentiments.

You May Also Enjoy

Organizing a directory workflow for a Semantic Segmentation task

2 minute read

Published:

Semantic segmentation is the task of assigning labels to different pixels in an image. So you can think of it as a binary or multi-label classification. That means the targets have the same size as their corresponding images. How do we organize the directory for such a task and how can we make use of the ImageDataGenerator? Let’s start with a simple example.

Many flavors of feature importance

5 minute read

Published:

I recently read this excellent book on financial machine learning which has a whole chapter dedicated to feature importance and its importance! It offers nice guidelines and some of the best practices for investigating feature importance in problems where we would like to know the extent to which different features contribute to the outcome of a machine learning model. Let’s imagine you are trying to predict whether you need to sell or buy stocks of a certain commidity in the market. Given the historical time evolution of the price, you can hand-engineer a large number of features. This includes:

Second order optimization methods and Neural Networks

1 minute read

Published:

Optimization of deep Neural Networks is often done using Gradient-based methods such as mini-batch gradient descent and its extensions such as Momentm, RMSprop, and Adam. Second order optimization methods such as Newton, BFGS, etc are widely used in different areas of statsitics and Machine Learning. Why are these methods are not popular in deep learning?

Mixture density networks for hard to evaluate conditional probabilities in cosmology

1 minute read

Published:

One of the chalenges of cosmological parameter estimation is marginalizing out the nuisance parameters such as the parameters that model the connection between galaxies and dark matter. For such marginalizations we are required to evaluate this conditional probability p(galaxy properties | dark matter halo properties). The list of galaxy properties consists of stellar mass, star formation rate, etc and the list of dark matter halo properties consists of mass, maximum circular velocity, etc.