Have summaries of our latest blogs delivered to your inbox, so you can stay up to date on the topics and current events that matter to your business.
Market disruptors may be nothing new, but that doesn’t mean we shouldn’t be excited about the emergence of a boundary-breaking new business. After all, many of the products and services that...
The tech space has boomed for decades, and most investment firms and financial institutions have invested consistently and aggressively in the area. As a result, tech stocks have led the pack throughout...
Did you know that manufacturing leaders are 2.5 times more likely to put data and technology investments ahead of cost reduction efforts? There’s a good reason: Making data-driven decisions—supported...
Every day, the content universe of Nexis Uni expands by millions of documents, ranging from news, company information and legal data to a wide variety of journals for academic research. And we’re...
How well do you know your individual competitors and the competitive landscape in which you operate? When it comes to putting together useful business intelligence, SWOT analysis is an important piece...
American philosopher and author Matthew Stewart has said that “One of the distinguishing features of anything that aspires to the name of science is the reproducibility of experimental results.” That fact doesn’t change just because the type of science does. So, how can universities educating tomorrow’s data scientists ensure they have academic research and analytics tools that meet the reproducibility standard? See how Nexis® Data Lab makes it easy.
Nexis Data Lab is the controlled environment that students need to get down to the business of mining and analyzing big data. But what really sets it apart at the start is the incredible source universe that students and faculty can tap into for data science projects. With decades of archival news data, plus the latest from news publishers around the world, data scientists have a massive collection of text-based data at their fingertips – and the data requires far less wrangling because of the processes and proprietary technologies LexisNexis has in place to normalize, classify, and enrich the data. Enrichments ranging from topic tags to relevancy scores, plus powerful pre- and post-search refinement features allow users to go from queries to relevant result sets quickly. Because we handle the hosting, universities save the time and expense of creating a suitable infrastructure for research and analytics tools; meanwhile, users can analyze up to 10,000 documents in minutes. Moreover, the volume of data available with Nexis Data Lab makes it easy to cross-validate experiments by testing models with different sets of training data.
Since the full text data lives within the Jupyter notebook environment, users realize enhanced speed in their research and analysis. But beyond those advantages, Nexis Data Lab offers open-sourced and pre-packaged LexisNexis-developed Python libraries, allowing students just getting started in data science to use existing solutions while more experienced users can choose to modify the code or embed their own custom code. With options to view analysis in either text or graphical formats, students get the hands-on experience needed to unlock and share compelling insights from academic research.
Regardless of users’ chosen approach to conducting academic research and analysis, the Jupyter notebook certainly supports reproducibility. Whether students are executing projects to earn a grade or faculty members are pursuing publication, Nexis Data Lab allows users to easily export analysis, notebook code and a document manifest to their personal laptops for later use. This ensures easy replication by students’ professors or peer reviewers at academic journals alike.
While reproducibility is a basic tenet of academic research—regardless of the field of study—being able to replicate results from data science offers other benefits too.
It enables collaboration. The ability to share results with fellow researchers—as well as future colleagues—is crucial in data science. No one person can know all the relevant modelling approaches or analysis techniques. Reproducibility enables others to capitalize and build on initial findings, creating more value from the data science.
It fosters creativity. Whether users are testing and refining a model against previous results or comparing a new model to a previous one to evaluate performance, the ability to replicate what’s been done is crucial.
It builds trust. Anything people don’t understand can generate mistrust—and that goes double for artificial intelligence, thanks to Hollywood’s propensity for mad scientists and evil robots. But it’s an important consideration because getting future colleagues to buy in on data-driven insights depends on proving results can be trusted. Reproducibility offers that type of proof.
Get started with reproducible academic research and analysis now. Ask about a free trial of Nexis Data Lab!