How to Build a Private ChatGPT Using Open-Source Technology? Download our free white paper.

Start your NLP project easily with Lettria’s Python SDK

Discover how to easily start your NLP project in Python using Lettria's SDK, and learn how to do sentiment analysis, entity extraction, and emotion analysis with minimal effort.


Natural Language Processing (NLP) is a vital branch of artificial intelligence that focuses on the interaction between computers and human languages. By developing algorithms and models to process, analyze, and understand human language, NLP has become indispensable across various industries. It enhances customer support through chatbots and sentiment analysis, boosts productivity by automating tasks like content generation and summarization, optimizes search and information retrieval in large databases, and facilitates data extraction and analysis from unstructured text data.

NLP in Python

Diving into a Natural Language Processing (NLP) project can be quite an undertaking, as it usually involves numerous components and tools.

When starting an NLP project in Python, you often need to gather various tools such as NLTK for basic text processing and analysis, SpaCy for advanced NLP tasks, Gensim for topic modeling, TextBlob for simpler NLP tasks, and TensorFlow or PyTorch for deep learning applications. Additionally, you may need to consider paid tools and libraries like Snowflake for scalable data warehousing, as well as text annotation and labeling tools like Prodigy or LightTag for creating high-quality training data.

This extensive list of tools, each with its specific purpose, can be overwhelming and daunting for newcomers. This complexity can deter many from exploring the captivating world of NLP.

Challenges of legacy NLP workflows in Python

Navigating Natural Language Processing (NLP) projects with multiple moving parts can present a series of challenges that hinder progress and increase complexity.

Complexity of Integrating Multiple Tools

One of the main challenges in traditional NLP workflows in Python is integrating multiple tools and libraries. Each tool has its unique features, API, and data structures, making it difficult to combine them for end-to-end processing. This increases the learning curve for developers and results in time-consuming integrations.

Inefficient Error Handling

In traditional NLP workflows, errors may occur at any stage, from data preprocessing to model training and evaluation. Identifying the source of these errors and debugging them can be a complex and time-consuming process due to the intricate interactions between various tools and libraries.

Data Privacy Concerns

Data privacy is a critical concern in NLP applications, especially when handling sensitive information. Ensuring that third-party libraries and tools adhere to data protection regulations can be challenging, and non-compliance may result in legal and financial repercussions.

Increased Development and Maintenance Costs

Implementing traditional NLP workflows in Python often requires significant computational resources and time investment. Additionally, maintaining and updating the codebase as new tools or library versions become available can be labor-intensive and costly.

A simple solution to NLP in Python

But there's a simpler solution: Lettria and our Python SDK. With Lettria, you can streamline the entire process and replace all those different tools with just one comprehensive solution, making NLP projects more accessible and enjoyable than ever before.

Whether you’re looking to do sentiment analysis, entity recognition or parts of speech tagging, you’ll be able to do it in a much easier way with Lettria.

In this article we’ll uncover how to use Lettria in Python to get your NLP project off the ground way faster.

Lettria's No-Code Solution

Lettria offers a comprehensive solution to streamline your NLP projects, consisting of two key components:

  • A no-code platform to set up your NLP effortlessly
  • API and Python SDK for seamless integration of your NLP solution into production

Advantages of Lettria

Lettria's all-in-one approach brings several benefits to your NLP projects:

  • By consolidating everything into one single tool, you can achieve results much more quickly, making your NLP projects more efficient.
  • With a unified solution, optimizing your NLP system becomes faster and more straightforward, allowing for continuous improvement.
  • Errors can be traced and fixed more easily, reducing downtime and ensuring the smooth operation of your NLP applications.

Lettria simplifies the NLP process, allowing you to focus on delivering high-quality, impactful solutions for your organization.

Why Python?

Python has become the language of choice for many NLP practitioners, and for good reason:

  • It's easy to get started with, thanks to its intuitive and human-readable syntax, making it an excellent choice for beginners and seasoned developers alike.
  • Creating a coding environment and learning NLP is simple through tools like Google Colab or Jupyter notebooks, which enable users to write and execute code in an interactive manner.
  • Python allows for easy web app development using frameworks like Streamlit, empowering developers to quickly turn their NLP projects into user-friendly applications.
  • Furthermore, Python is versatile enough to be used in production environments, ensuring that your NLP solutions can be scaled and deployed seamlessly.

With its numerous advantages, Python serves as the perfect foundation for your NLP journey, providing a flexible and robust platform for development and deployment.

What can you do with Lettria in Python?

With Lettria, you can do a variety of different NLP tasks and processes, with one single SDK.

  • Advanced tokenization: Breaks down text into sentences, words, and punctuation marks
  • Subsentence detection: Splits sentences into smaller parts for a deeper analysis
  • Part-of-Speech (POS) tagging: Identifies the role of each word in a sentence
  • Dependency parsing: Uncovers connections between words by labeling their relationships
  • Lemmatization: Finds the simplest form of each word for increased accuracy
  • Named Entity Recognition (NER): Detects elements such as people, places, and dates within the text
  • Natural Language Understanding (NLU): Interprets the meaning of words and phrases
  • Coreference analysis: Identifies links between different parts of the text for a cohesive view
  • Sentiment analysis: Assesses the sentiment conveyed in the text (positive, negative, or neutral). Read about what is sentiment analysis and how it works here.
  • Emotion analysis: Evaluates the emotions expressed in the text (positive, negative, or neutral). Learn how to analyze survey data with Emotion Analysis here.
  • Sentence Type classification: Detects the purpose of each sentence (question, statement, command, exclamation)
  • Language detection: Accurately identifies the language of the text for smooth processing

Quickstart Guide to Lettria in Python

To demonstrate how Lettria's NLP solution works, we'll perform a request on two sample documents and extract relevant information from the results. In this context, a "document" refers to any piece of text you'd like to analyze individually, such as a review from a dataset of online reviews.

Obtaining Your API Key

Before you can start making requests to Lettria's API, you'll need an API key. Don't worry, you can quickly create a free key to get started.

Read this knowledge base article or watch the video below to learn how to set up your first project and get your API key.

To demonstrate how Lettria's NLP solution works, we'll perform a request on two sample documents and extract relevant information from the results. In this context, a "document" refers to any piece of text you'd like to analyze individually, such as a review from a dataset of online reviews.

Installation and Setup of Lettria in Python

The first thing you need to do is install Lettria.

Then, import Lettria and set up the NLP class with your API key:

You’ll be using the variable ‘nlp’ for all of your analyses with Lettria.

Analyzing Data with Lettria

Next, define the documents you want to analyze. Lettria in Python takes documents in the form of a string or a list of strings.

We can make a list of strings of sentences that we want to analyze with Lettria as below.

Now, perform requests to Lettria's API for each document. The add_document() function performs a request to the Lettria API using your API key, to analyze the document added. The results are appended as an additional ‘Document’ instance to the ‘documents’ attribute.

The NLP analysis results are saved in the ‘nlp’ variable. You can save the results as a JSON file for future use as well.

Similarly, you can also load results from a JSON file with the load_results() function. To learn more, you can read the Python SDK documentation here.

Leveraging the Results

The NLP class is designed to give quick access to relevant data at the document, sentence and subsentence level. Once all the results are loaded into the NLP class, you can leverage them using different functions. When a response from the API is received, it is stored in a hierarchy of classes:

NLP (all data) → Document → Sentence → Subsentence → Token

At each level, direct access to lower levels is possible. For example, nlp.sentences gives access to a list of all the sentence objects in the current data, and nlp.documents[0].sentences will give the sentence objects of the first document object.

You can get a full list of functions that can be accessed through the NLP, document, sentence and subsentence classes below.

Method Description
vocabulary() Returns vocabulary from current data.
word_count() Returns word count from current data.
word_frequency() Returns word frequency of current data.
list_entities() Returns dictionaries of detected entities by type.
get_emotion() Returns emotion results at the specified hierarchical level
get_sentiment() Returns sentiment results at the specified hierarchical level
word_sentiment() Returns average sentiment for each word of the whole vocabulary
word_emotion() Returns average emotion for each word of the whole vocabulary
meaning_sentiment() Returns average sentiment for each meaning
meaning_emotion() Returns average emotion for each meaning
filter_polarity() Filters Sentence or Subsentence of the specified polarity
filter_emotion() Filters Sentence or Subsentence of the specified emotions
filter_type() Filters Sentence of the specified types
match_pattern() Returns matches from given patterns.

Closing Thoughts

If you want to learn more about the Python SDK, you can access the full documentation here to understand the different classes and functions. You can also read our customer review sentiment analysis tutorial article, to check out a real world use case of the Python SDK.

With this guide, you're now ready to harness the power of Lettria's NLP solution in your own Python projects. If you’re interested in learning more about Lettria, we have plenty of material on our blog to help you get started!

Want to learn how to build a private ChatGPT using open-source technology?


Build your NLP pipeline for free
Get started ->