Data Science Insider: August 11th, 2023

In This Week’s SuperDataScience Newsletter: Meta's Protein Folding Model Team Laid Off. Crucial Role of Data Visualization. Streamline Data Cleaning and Joins with Tableau Prep. Master Sentiment Analysis. Bots Surpass Humans in CAPTCHA Tests. Cheers, - The SuperDataScience Team P.S. Have friends and colleagues who could benefit from these weekly updates? Send them to [this link]( to subscribe to the Data Science Insider. --------------------------------------------------------------- [Meta's Protein Folding Model Team Laid Off]( brief: Meta, the company behind the protein folding model which we have extensively covered in these SuperDataScience newsletters, has reportedly laid off its team responsible for the model's development. The ESMFold team used AI to develop a database of over 600 million protein structures and has now been disbanded. The team's project had successfully trained a language model to predict protein structures from biological data, benefiting drug development. This move reflects Meta's shift towards revenue-generating projects over pure scientific endeavours. Despite the team's small size in comparison to Meta's extensive AI workforce, this decision highlights the company's emphasis on commercial AI ventures. The restructuring aligns with Meta's transformation efforts to enhance profitability and growth, with a focus on generative AI products. Why this is important: This move highlights the dynamic nature of AI research and the potentially worrying evolving priorities of tech companies, underscoring the need for data scientists to align their expertise with business goals and the potential impact on research directions. [Click here to learn more!]( [Crucial Role of Data Visualization]( brief: Data visualization is the graphical representation of data to simplify complex information, enabling easier interpretation and analysis. This Builtin article underscores its significance in illuminating patterns, relationships, and insights within datasets, aiding decision-making across industries. The article also discusses the role of aesthetics and interactivity in creating compelling visualizations and highlights how storytelling through data can foster understanding. It does this whilst emphasizing the importance of data scientists mastering visualization techniques to effectively convey their analyses and discoveries to both technical and non-technical stakeholders. Visualizations, such as charts, graphs, and interactive dashboards, are essential tools for data scientists to master as they enhance data communication by making information accessible to diverse audiences. Why this is important: Understanding data visualization enables data scientists to transform intricate data into understandable insights for various audiences. Mastery of visualization techniques enhances collaboration, supports effective decision-making, and empowers researchers to effectively communicate the value of their analyses and inform strategic directions based on data-driven insights. [Click here to read on!]( [Streamline Data Cleaning and Joins with Tableau Prep]( In brief: This Medium article praises the virtues of Tableau Prep, arguing that it is a crucial tool in data analysis, offering streamlined data pipeline solutions for data analysts and scientists. Tableau Prep simplifies the data cleaning and joining processes, enabling effortless data manipulation, including various joins like left, inner, right, and outer, facilitating a comprehensive understanding of data behaviour. The application excels in handling primary key-based data joins, surpassing even Microsoft Access with its efficient algorithm. It simplifies data cleaning through text manipulation and supports easy conversion of data types, ensuring consistency. While aiding data refinement, Tableau Prep introduces geographic updates and fuzzy matching. This tool empowers analysts by facilitating quick data cleaning, enhancing insights, and creating efficient data pipelines. Why this is important: Data scientists benefit greatly from Tableau Prep as it eliminates the need for complex programming, aiding in faster data preparation and insightful analysis. Understanding its capabilities enhances data professionals' efficiency, enabling them to deliver accurate insights promptly. [Click here to discover more!]( [Master Sentiment Analysis]( In brief: In this advanced and interesting data science tutorial, sentiment analysis is employed using NLTK and Altair to gauge positivity in news headlines from the UCI News dataset. The analysis centres on valence scores, employing the Vader Sentiment Intensity Analyzer from NLTK. The approach involves lexicon-based sentiment scoring, classifying headlines into positive, neutral, and negative categories. The tutorial demonstrates visualizing sentiment distribution via Altair, revealing insights into article categories and publisher sentiment trends. Finally, the significance of effective communication through interactive reports is emphasized, utilizing the Datapane library to package visualizations and insights for comprehensive understanding. The tutorial underscores sentiment analysis, interactive visualization, and clear reporting as essential skills for data scientists. Why this is important: Data scientists must master sentiment analysis for extracting insights from unstructured text. Proficiency in creating interactive visualizations, as demonstrated with Altair, enhances data exploration and communication. Crafting context-rich reports, like those enabled by Datapane, empowers professionals to present findings effectively. These skills are indispensable for robust data analysis and decision-making. [Click here to see the full picture!]( [Bots Surpass Humans in CAPTCHA Tests]( In brief: Recent research indicates that bots outperform humans in solving CAPTCHA tests (those annoying ones that ask you to select all images with cars in them!), designed to differentiate between human users and bots on websites. This revelation raises questions about the continued efficacy of such security measures, considering the inconvenience that they pose to many users. The study conducted by Gene Tsudik and his team at the University of California, Irvine, found that bots exhibited higher accuracy and speed in solving CAPTCHAs across a variety of popular websites. The researchers suggest that the effort invested in maintaining CAPTCHAs may be unwarranted, advocating for the use of intelligent algorithms to identify and deter bot interactions instead of relying solely on traditional tests. Why this is important: This research underscores the advancements bots have made, necessitating innovative approaches to web security. Data scientists, like you, must engage with intelligent algorithms and ML techniques to fortify online platforms against bot-driven threats and find more effective ways to distinguish between human users and bots. [Click here to find out more!]( [Super Data Science podcast]( this week's [Super Data Science Podcast]( episode, Chris Wiggins talks with Jon Krohn about the power dynamics of data, the transformation of the field of biology through data-driven approaches to genetic sequencing, and the New York Times’ data science team’s cutting-edge approach to accommodating its tech stack. [Click here to find out more!]( --------------------------------------------------------------- What is the Data Science Insider? This email is a briefing of the week's most disruptive, interesting, and useful resources curated by the SuperDataScience team for Data Scientists who want to take their careers to the next level. Want to take your data science skills to the next level? Check out the [SuperDataScience platform]( and sign up for membership today! Know someone who would benefit from getting The Data Science Insider? Send them [this link to sign up.]( # # If you wish to stop receiving our emails or change your subscription options, please [Manage Your Subscription]( SuperDataScience Pty Ltd (ABN 91 617 928 131), 15 Macleay Crescent, Pacific Paradise, QLD 4564, Australia