Figure 1. This plot shows an interactive donut chart of top 30 words describing data analysis responsibilities.

It is unsurprising that the most frequent word for data analysis job responsibilities is "data". Other key words that stand out include "business", "support", "processes", "management", "reports", "quality", and "customer". These words suggest that data analysts use data to support the business needs of their customers, which may involve tasks such as drafting reports, performing quality control over various data processes, and sometimes even becoming involved in the management of teams and projects.

Figure 2. This plot shows an interactive donut chart of top 30 words describing data science responsibilities.

Although there are some words in this chart that overlap with the responsibilities of a data analyst, there are also key differences that set data scientists apart. For instance, words such as "development", "models", "learning", "techniques", "predictive", "statistical", and "algorithms" do not frequently appear in the list of data analyst responsibilities. These words highlight the unique responsibilities of data scientists in terms of development of predictive machine learning models and algorithms, and use of statistical techniques to draw insights and information for their projects and teams.

Figure 3. This plot shows an interactive donut chart of top 30 words describing machine learning responsibilities.

Again, overlap between the responsibilities of data science and machine learning job types exist. However, "learning" is the word with the highest frequency in this chart, which highlights the main focus of machine learning: learning how to find the optimal interaction between data and machines. Other words such as "technical", "design", "engineering", "build", "software", "code", and "research" also stand out. These words suggest that machine learning roles may be more technical in nature, falling more frequently under the title of machine learning engineer, and having responsibilities such as conducting research, building and designing code using various software tools to create machine learning models.