Explore text analytics

In this exercise, you’ll use AI natural language processing functionality to analyze text. The goal of this exercise is to explore common applications of text analysis techniques.

This exercise should take approximately 15 minutes to complete.

Prepare for text analysis

We’ll use a browser-based application that’s based on a simplified subset of the text analysis capabilities of the Microsoft Foundry language playground.

  1. Download and extract text.zip at https://aka.ms/ai-text. This archive contains multiple text documents that you’ll use in this exercise.
  2. In a web browser, open the Text Analyzer at https://aka.ms/text-analyzer. This application provides the text analysis functionality you’ll use.

Analyze sentiment

Sentiment analysis is a common NLP task. It’s used to determine whether text conveys a positive, neutral or negative sentiment; which makes it useful for categorizing reviews, social media posts, and other subjective documents.

Note: The sentiment analysis functionality in the app is a simplified implementation of the AFINN-165 scoring technique, in which a set of words with scores between -5 (negative) and +5 (positive) are used to calculate an aggregate sentiment score for the text.

  1. In the Text Analyzer app, ensure the Analyze sentiment tile is selected.
  2. Upload document-1.txt from the folder where you extracted the downloaded text files.

  3. Run the analysis and observe the resulting sentiment prediction.

    Screenshot of sentiment analysis results in text Analyzer.

  4. Repeat the analysis for document-2.txt and document-3.txt.

    You can experiment further by uploading your own text files or by typing text into the Text Analyzer interface. The results may vary due to the simplified implementation of sentiment anslysis in this lightweight app.

Extract key phrases

Key phrase extraction can be a useful first step in identifying the main topics in a document.

Note: The key phrase extraction technique used in the app is based on statistical analysis techniques such as word frequency and parts-of-speech tagging.

  1. In the Text Analyzer app, select the Extract key phrases tile.
  2. Upload document-1.txt from the folder where you extracted the downloaded text files.

  3. Run the analysis and observe the resulting key phrases.

    Screenshot of key phrase extraction results in text Analyzer.

  4. Repeat the analysis for document-2.txt and document-3.txt.

    You can experiment further by uploading your own text files or by typing text into the Text Analyzer interface. Again, the results may vary.

Extract named entities

Named entities are the people, places, dates, and other important items mentioned in text.

Note: The named entity recognition technique used in the app is based on pattern matching and a dictionary of common entity types.

  1. In the Text Analyzer app, select the Extract named entities tile.
  2. Upload document-1.txt from the folder where you extracted the downloaded text files.

  3. Run the analysis and observe the resulting named entities.

    Screenshot of named entity recognition results in text Analyzer.

  4. Repeat the analysis for document-2.txt and document-3.txt.

    You can experiment further by uploading your own text files or by typing text into the Text Analyzer interface.

Summarize text

Summarization is a way to distill the main points in a document into a shorter amount of text.

Note: The summarization technique used in the app is based on the TextRank algorithm.

  1. In the Text Analyzer app, select the Summarize text tile.
  2. Upload document-1.txt from the folder where you extracted the downloaded text files.

  3. Run the analysis and observe the resulting summarization.

    Screenshot of text summarization results in text Analyzer.

  4. Repeat the analysis for document-2.txt and document-3.txt.

    You can experiment further by uploading your own text files or by typing text into the Text Analyzer interface.

Summary

In this exercise, you explored the use of a AI to analyze text, using NLP functionality in a browser-based app.

While the statistical techniques in this exercise are sufficient to demonstrate the concepts, to perform high-quality language analytics at scale, you should use a cloud-based AI platform like Microsoft Foundry. Microsoft Foundry a wide range of language models, some of which are optimized for specific language processing tasks. Additionally, Azure Language in Microsoft Foundry tools offers a specialized service with APIs for common text analytics tasks.