The world’s leading publication for data science, AI, and ML professionals.

5 Ideas For Your Next NLP Project

Generating Some Ideas

Natural Language Processing Notes

Photo by Stanley Dai on Unsplash
Photo by Stanley Dai on Unsplash

Natural Language Processing (NLP) is a branch of Artificial Intelligence (AI) that is concerned with the interactions made between computers and natural language. Essentially, by analyzing and representing natural language computationally, computers are capable of understanding natural language and responding in a way similar to a human.

As a beginner learning the ropes of any new technology, getting your hands dirty is an important part of the learning process. Although I believe theoretical knowledge is very crucial, I don’t believe it’s effective in isolation as the theory doesn’t always translate into real-world scenarios. Taking a practical approach is by far the greatest way to keep testing yourself whilst gaining experience of what it’s like to work in a real-world environment.

Without further ado, here are 5 Ideas for your next NLP project:

1 Conversational AI

Chatbots and voice assistants fall into the category of Conversational AI. Conversational AI permits Artificial Intelligence (AI) to interact with natural language in a human-like way.

There are various use cases for chatbots in enterprises. Some examples include:

  • Customer Support – Online chatbots are superb at answering frequently asked questions, hence they are replacing many human agents along the customer journey.
  • Healthcare – Babylon Health in the UK is a ** great example** of how conversational AI is making health care services more affordable and accessible to patients.
  • HR Processes – With the Covid-19 lockdown, many people are working from home but companies are still hiring, meaning employee training and onboarding are still necessary. These processes can be optimized with Conversational AI.

2 Autocomplete and Autocorrect

Autocomplete and autocorrect have both come to my aid on numerous occasions. I’m sure we’ve all typed "omw" on our smartphone for it to automatically be changed to "on my way!" or were on the verge of misspelling a word an autocorrect chimes in to save the day.

Google’s search engine is a great example of both autocomplete and autocorrect. When we type something into its search engine mechanism, google would scour the internet and return the most relevant results.

Gif By Author
Gif By Author

These technologies help users find much more accurate results and the crazy thing about it is the majority of us use google everyday hence we expect these features to be there once we click on a search box. To continuously keep up with customer expectations, both autocomplete and autocorrect are valuable applications of NLP to integrate with many businesses therefore making it a really great project to implement.

3 Text Generation

Technically speaking, text generation falls under Natural Language Generation (NLG) which is a subfield of Natural Language Processing (NLP). Wikipedia describes NLG as "a software process that transforms structured data into natural language. It can be used to produce long-form content for organizations to automate custom reports, as well as produce custom content for a web or mobile application."[Source: Wikipedia].

4 Named Entity Recognition

Named-entity recognition is a subtask of information extraction that seeks to locate and classify named entities mentioned in unstructured text into pre-defined categories such as person names, organizations, locations, medical codes, time expressions, quantities, monetary values, percentages, etc. [Source: Wikipedia].

Named-entity Recognition (NER) is extremely valuable in any situation in which a high-level overview of a large document is required. By implementing NER you can easily gather the gist of a body of text and begin to group the documents based on certain criteria such as the relevance or similarity to other text. Some use cases include:

  • Customer Assistance – Customer’s requests would be categorized into suitable headings and keywords would be extracted to determine the priority of the request to helps speed up response times.
  • Recruitment – Instead of having humans parse numerous CVs, we can use NER to summarize applicant’s resumes and begin to categorize them into suitable hires and unsuitable.

5 Sentiment Analysis

In Getting Started with Sentiment Analysis, I stated that "The goal of sentiment analysis is to interpret and classify subjective data using natural language processing and machine learning.". Many examples of this type of problem are performed using customer feedback and twitter data but you could use any type of text where people voice their opinions or experiences.

I’ve covered Sentiment analysis quite extensively in previous blogs and my first NLP project was a sentiment analysis project. Check out the Github repository below.

kurtispykes/twitter-sentiment-analysis

Wrap Up

Projects are probably more viable than a CV in this day and age. If you wish to learn a new technology or break into a new field, getting hands-on with a project is the best way to get ahead. Be sure to learn from your mistakes and don’t be afraid to ask for help when you’re stuck on something.

Thanks for reading! Connect with me on LinkedIn and on Twitter to stay up to date with my posts on Data Science, Artificial Intelligence, and Freelancing.

Related Articles

Sentiment Analysis: Predicting Whether A Tweet Is About A Disaster

7 Free Online Resources For NLP Lovers

Tips To Land A Data Science Role in 2021


Related Articles