DEV Community: Jasper Samuel J

Amazon QuickSight

Jasper Samuel J — Fri, 15 Mar 2024 15:36:18 +0000

It's often hard to analyze the data and learn from it for those who are not familiar with programming languages, that's where QuickSight comes in.

QuickSight is business analytics and intelligence service from AWS, business users can collect meaningful insights.
Unlike PowerBI this is not a huge application to download and play around, QuickSight is a web based application which makes it easy for us to work on our data without worrying about storage and security.

Features

Data Integration The biggest problem with most of the analytics and visualization applications are data integration. In QuickSight we can connect data from various sources which includes all the AWS services and third party services like MySQL, SQL Server etc.
ML Insights QuickSight will deliver deeper insights and trends without manual analysis based on the Machine learning algorithm which powers Amazon.com for the past two decades.
Data Visualization Visualization is paramount for any analytics application as it facilitates comprehension of data for business stakeholders, empowering them to make informed decisions effortlessly, QuickSight have the option to create wide range of options to create graphs, charts, heatmap etc.
Collaboration QuickSight reports are easy to share and embed on your reports without worrying about the security concerns.
Security QuickSight is secured by Role based access control of AWS's IAM this will help you to save your business data, reports from others without your permission.
Pricing QuickSight offers an unique pricing models where users charged on each sessions initiated, this will help you to save tons of money.
Login into the AWS Console and search for QuickSight

Click on New analysis and then new dataset

You can choose the data sources to import the data

For time being I'll navigate you around to quicksight using a sample data.

You can add new data by clicking data and add dataset.

You can drag and drop the fields from left to the graph to insights.

You can publish your report by clicking on publish

Amazon Comprehend

Jasper Samuel J — Mon, 02 Oct 2023 09:15:49 +0000

Insights from Documents

In the business world each and every word from the customers place a vital role in the development of the new products and improving the existing ones, in order to make this process simple and intuitive AWS launched Amazon Comprehend a service which we can use to get insights from documents, social media platforms, product reviews, support tickets etc.

Amazon Comprehend

Amazon Comprehend is a Natural Language Processing which uses Machine Learning to collect valuable insights from the texts, we can collect the information from Customer ratings, reviews, support tickets, social media platforms, surveys etc as Comprehend provides Custom Entity Recognition we can process all the data and get Custom Classification, Key phrase Extraction, Sentiment Analysis, Entity Recognition and much more

Features

Custom Entity Recognition allows you to identify the domain specific terms, using AutoML Comprehend will learn from small examples and then later we can train our own private custom ML model
Custom Classification API enables you to build custom models using your business specific labels without learning ML, Custom Classification will automatically classify the inbound requests by problem types, it will help you monitor the website comments, customer feedback etc.
Entity Recognition API helps you automatically categorize the entities based on the location, date, person name etc.
Sentimental Analysis API will help you to get the overall sentiment of the text inbounded and targeted sentiment will provide more granular insights by identifying sentiment towards entities within text.
PII Identification and Redaction Comprehend ML capabilities and helps you detect the PII data from emails, product reviews, support tickets etc. you can redact the PII entities before documenting the solution, search solutions are free for PII entities.
Topic Modelling identifies a relevant term and topics from the collection of documents stored in S3 and map them as a group based on the relevant topic.
Multiple Language Support Comprehend can perform text analysis on multiple languages like German, Chinese, Hindi etc, if you want more languages support you can use Amazon Translate to translate them into the Comprehend supported languages.

Click on Launch Comprehend

You can find the sample input data in the text box, click on the Analyze
Scroll down in the results area you can find the data get classified.

You can try with your own custom data and check out the result.

Amazon Personalize

Jasper Samuel J — Thu, 01 Sep 2022 13:53:05 +0000

Personalization

Ever wondered while scrolling a shopping site you came across most of the ads and products you searched or purchased recently, that’s called personalization.

When you are using a social media application you can see most of the posts which are shown in your feed are related to the contents you interacted, liked and commented, after sometime when your interaction changes your feed contents will also got changed, behind the scenes Personalization is happening.

Amazon Personalize

Amazon Personalize makes the developers to create personalized applications with the power of Machine Learning, Amazon Personalize will help us to built applications which is providing personalized product recommendations to the end users and customized direct marketing.

Amazon Personalize will fine-tune and adopt automatically when the end-users usage differs and varies time to time, Meanwhile the developers have the access to deploy custom ML models to deliver customized recommendations across various industries.

As, Amazon Personalize doesn’t need any ML experience to work on, anyone who knows the basics of AWS can easily make their applications or websites tailormade for all the end users.

Features

Real-time recommendations The ML models which is powering Amazon.com for two decades is used in Amazon Personalize which makes tailored recommendations by responding the changing intent of the users in real-time.
New user and New item recommendations Amazon Personalize helps the new users to find the correct recommendations and makes the new products easily seen by all the end users.
Similar Item recommendations Improve the discoverability of all your products by showing the similar items to all the users.
Automated ML You don’t need any ML experience to work on Amazon Personalize, once you provide the data on Amazon S3 or other real-time integration, Amazon Personalize can automatically load and inspect the data, let you select the right algorithms, train a model, provide accurate metrics and generate personalized recommendations.

Adding Personalization to your application

Click and get started and name your dataset group
Choose the domain for your application or website

Name your dataset

Choose the data source and give it a name
Give the data location pathway and click finish

After importing the data and give your business rules Amazon Personalize will do it’s job.