Data Science for Librarians: Transforming Information into Insight

By Jason Miller

Discover the transformative potential of data science in the world of libraries with this comprehensive guide tailored specifically for librarians seeking to enhance their professional expertise. Delving into the intersection of information science and cutting-edge data analytics, this book equips readers with the knowledge and skills needed to harness the power of data for informed decision-making and innovative service delivery.

From understanding the fundamentals of data science to implementing advanced techniques like machine learning and text mining, each chapter offers practical insights and real-world examples that illuminate the path forward. Readers will learn how to collect, clean, and analyze data effectively, uncovering valuable insights that can drive strategic initiatives and optimize library resources.

But this book is more than just a technical manual—it's a roadmap for librarians navigating the complexities of the digital age. With a focus on ethical considerations, privacy protection, and staying ahead of emerging trends, it empowers librarians to leverage data responsibly and ethically, ensuring that their practices uphold the core values of librarianship.

Whether you're a seasoned professional looking to expand your skill set or a newcomer eager to explore the possibilities of data science, this book is your indispensable companion on the journey to unlocking the full potential of libraries in the 21st century.

• Language: English
• Publisher: SD
• Release date: May 7, 2024
• ISBN: 9798224905171

Jason Miller

Jason Miller (Inominandum) has devoted more than twenty years to studying practical magic in its many forms. The author of the now classic Protection and Reversal Magick as well as The Sorcerer's Secrets, Financial Sorcery, and Sex, Sorcery & Spirit, he runs the Strategic Sorcery training course Take Back Your Mind Program and The Sorcery of Hekate Training. He lives with his wife and children in the New Jersey Pine Barrens. For more information, visit StrategicSorcery.net.

Book preview

1

introduction to data science and its relevance in libraries

understanding the fundamentals of data science

Draw a circle. This is the data lifecycle, the heart of data science. We can think of data science as peeling the layers of an onion. We see what's inside the data and the intricacies that govern the different actions you can carry out on data.

So what exactly is data science? It is the process of discovering knowledge and understanding of data. One of the best ways to think about data science is as a crime scene investigation. Imagine you are a detective and arrive at the scene of a crime. You collect the physical and digital traces of the crime.

Now, all of these traces are of data type. You collect data about the footprints, the fingerprints, the personal witness accounts, and the CCVT footage. Each of these bits of data is a clue. A data scientist is no different from a detective, except that they have to analyze extremely large amounts of data. They try to understand the patterns, trends, associations, and correlations in this large body of data.

One of the most important concepts in data science is the data lifecycle. Data has a lifecycle, just as a human organism has a lifecycle. It is born, it grows, reaches maturity, and eventually dies. The first step in this lifecycle is data collection, taking raw inputs from multiple sources. For example, the circulation log in the case of a library, the results of the user survey for a website, or the web traffic available at the site.

This is followed by the preprocess data, in which the raw data is cleaned up, modeled, and arranged so it is ready to be analyze. Next comes exploratory data analysis. The data scientist uses statistical techniques and visualization tools to explore the data, to find out which interesting knowledge can be extracted from the data.

Last, knowledge extracted from the data is used in decision making. Data is to data science what language is to linguists. With reference to the schematic, language to a linguist is a material that needs to be analyzed; similarly, data is a material to be processed for the data science.

Ok, now let's complect our conversation a bit. Now you may ask, what does complect mean? I would describe it as making our conversation richer, more saturated with explanation and supporting data. It's like weaving the different strands of threads together to form a whole woven fabric. In data science we often deal with complex systems, where multiple interacting factors are influencing one another in complex ways.

So, making the conversation more complect would mean explaining all the dependencies and acknowledging that fact A is related to factors B, C, and D. And in fact, relationship of A with B is not the same as relationship of A with C. This concept of complect I think would be a good model for much of the impact assessment that we often do in our field (like assessing what is the impact of book borrowing data on literacy or achievement scores, or learning what is the effect of library services on students success). If we are up to understand the reasons behind the book borrowing patterns we could also complect in the demographic data. We might want to consider the borrowing behavior for each age bracket or for each different socio-economic category.

Or we might want to consider metadata about the books themselves. Did borrowers tend to choose books from different genres depending on the time of year? Or did books by one particular author become significantly more popular? What were the trends in borrowing behavior relative to the publication date of the book, and the length of time it has spent on the shelf? This complects the problem considerably.

All of these considerations provide us with a lot more insight and additional meaning that we could add to our raw data. And that difference of insight and meaning is really the point of the discussion thus far. There is, of course, no substitute for practical experience. You need to be able to mix it with the real world in order to work out every nuance and feature of what you are looking at. You can't devise a grand, notional schema for every project you're working on, and whilst this is true, making the discussion richer and thicker will inevitably give you a stronger insight. The model is, of course, in the end, an abstraction.

But it will be an abstraction that is much more close to the reality of the thing you are describing. It's an abstraction that goes one step further, just beyond the surface of the water. In short, as you're talking to your colleagues, or raising a problem with stakeholders, you should always be looking to complect your conversation. Not just focus upon the data; but the situation.

why data science is essential for modern librarianship