Engineering

Diving into AI with Azure Cognitive Services

Diving into AI with Azure Cognitive Services

Artificial intelligence (AI) is one of the core technologies for digital transformation that is helping modern businesses scale up. To make AI possible, we need to find answers for the following three issues:

  1. Big computing
  2. Powerful algorithms
  3. Massive data

Thanks to the advancement of cloud computing, we can more easily solve these problems. Microsoft Azure is one of the top three cloud providers, and they also offer AI services – Azure Cognitive Services – which are well worth exploring.

What Are Azure Cognitive Services?

Azure Cognitive Services are cloud-based artificial intelligence services that help developers build cognitive intelligence into applications without having direct AI or data science skills or knowledge.

These services can be categorized into four main pillars:

  • Vision: Identify and analyze content within images and videos.
  • Speech: Improve application experiences through speech services.
  • Language: Understand unstructured text and conversations and analyze them.
  • Decision: Speed decision making.

Let’s go deeper with an example showing how Azure Cognitive Services might help to support a travel agency.

Travel Agency Example

Consider the following scenario: You work for a travel agency that has amassed a large amount of hotel data from around the world. Now, you want to create a service that will help your customers easily find the right places to stay or activities to do during their trip.

Azure Cognitive Services can help make this possible in several ways. Let’s explore how.

Data Collection

Hotel data is collected from various sources and stored in a variety of formats, including structured formats (relational databases, CSV) and unstructured formats (PDFs, audio, videos, images, etc.) It is estimated that around 80% of business data is unstructured. Consequently, we are all faced with a similar challenge: How do we effectively apply our information to our products and business?

Building a Search Solution with Azure Cognitive Search

Azure Cognitive Search is a cloud search service that provides developers with the infrastructure, APIs, and tools needed to create rich search experiences from private, heterogeneous data collections and then leverage them in web applications, mobile apps, and enterprise solutions.

Cognitive services interpret unstructured data in a human-like way. In turn, when search operations are coupled with artificial intelligence, together they enable us to create solutions capable of finding more value in data.

The picture below shows a sample architecture in which we could leverage Microsoft Azure infrastructure to build a search solution:

Data Flow

  1. Azure Cognitive Search consumes unstructured data from Azure Blob Storage, including documents and images.
  2. Document cracking begins the indexing procedure by extracting images and text from the data, followed by content enrichment. The enrichment steps in this procedure depend on the selected data and skill type.
  3. Built-in skills based on the Computer Vision and Language Service APIs enable AI enrichments including image optical character recognition (OCR), image analysis, text translation, entity recognition, and full-text search.
  4. Custom skills are useful in scenarios that necessitate more complex AI models or services. Forms Recognizer, Azure Machine Learning models, and Azure Functions are a few examples.
  5. The indexer saves enriched documents in a search index. This index supports full-text and other queries.
  6. The enriched documents can also project into a knowledge store, which downstream apps like knowledge mining or data science can use.
  7. Queries access the enriched content in the search index. The index supports custom analyzers, fuzzy search queries, filters, and a scoring profile to tune search relevance.
  8. Any application that connects to Blob Storage or to Azure Table Storage can access the knowledge store.

Keyword Search and Semantic Search

Pictured below are visualizations of how full-text and semantic search work:

Full-text search

Semantic search

Conclusion

I used to believe that artificial intelligence was difficult to learn and apply. But now, with the help of Azure Cognitive Services, I am confident that I can harness the power of AI and incorporate it into my work. 

Become Part of Our Talented Team

If you are looking for an incredible opportunity and a great company, Wizeline may be the perfect place for you. I invite you to follow us on LinkedIn Wizeline and WizelineAcademy to find information about upcoming bootcamps for IT professionals who want to keep expanding their skills. 

Also, why not allow yourself to become a Wizeliner? We have open roles for engineers in Colombia, Mexico, Vietnam, and Spain. Apply here today if you’re interested in joining our global team.

References

https://www.mongodb.com/basics/full-text-search

https://www.youtube.com/watch?v=1rrwL3qczRU

https://www.youtube.com/watch?v=d_6ZNyV1MvA

https://learn.microsoft.com/en-us/azure/architecture/solution-ideas/articles/cognitive-search-with-skillsets

 


Ana Karen Aguilar

Posted by Ana Karen Aguilar on October 31, 2022