Welcome to our QTM 350 - Data Science Computing Final Project! Our team is Simran Mallik, Rishika Shah, Yash Bhatia, Lia Rubel, Ceci Chen, and Senna Kim. We are excited to share our comparative analysis of AWS Translate and Google Translate with you.

In this project, we assess the quality of AWS Translate by comparing it to Google Translate through both qualitative and quantitative measures. We also analyze the services through a formal vs informal framework. In this study, we use Spanish, French, and Chinese as our test languages. Ultimately, we conclude that users may find AWS Translate more helpful for “informal” language and Google Translate more helpful for “formal” language.

To walk through our project, please read our blog which outlines our process in its entirety. The blog includes the following sections:

• Background and Motivation

• Hypothesis

• Data

• Method

• Analysis

• Results

• Discussion

The data section in the blog also includes an architecture diagram of the following AWS resources used in the project:

• Amazon Datasync Securely move data from local machine and network to S3 bucket

• Amazon S3 Buckets A source bucket to store data for analysis

• Amazon Sagemaker Deploy Amazon Translate ML API in Jupyter Notebook environment

For a deeper dive into our analysis (data cleaning, visualization, statistics), refer to the following Jupyter Notebooks:

• Final AWS notebook Analysis of AWS Translate through qualitative and quantitative metrics

• Final Google Notebook Analysis of Google Translate through qualitative and quantitative metrics

• Final Project Analysis notebook Comparison of analysis between AWS and Google Translate with visualizations