Voyant Tools 2.0
Voyant Tools is a data visualisation and analysis tool used for documents, articles, and even books. It gathers the data from the inputted text and displays it in a variety of ways, using links, termberries and bubble lines. It allows the user to analyse data by finding the most used words or phrases and their frequency throughout the text.
Geoffrey Rockwell, a Professor in the University of Alberta, Canada and Stéfan Sinclair, an Associate Professor in McGill University developed the site in 2014. It has since evolved, with improved usability, increased features and most importantly, it is now open sourced. It also shares its ancestry with HyperPro, which was limited to small blocks of text no more than a couple of megabytes in size. In contrast, Voyant Tools 2.0, released in 2016, has a much greater capacity and can readily analyse entire novels, as it was designed with scale in mind.
Another difference between Voyant Tools and its previous versions is the license it uses. Its code is under a GPL3 license and the content of the web application is under a Creative Commons by Attribution license. This gives users the right to share, use and build upon the tool. Neither Voyant Tools 1.0 or HyperPro gave this authority to its users.
Voyant Tools is available online with no download needed and supports a variety of both import and export formats, e.g. HTML, XML, TXT. Users can copy and paste text, upload documents, or examine existing corpus. There is a wide variety of pre-uploaded texts available, such as William Shakespeare’s plays or Jane Austen’s novels. One of Voyant Tools most useful features, is its ability to analysis more than one document at any given time. You can upload files or paste many URL’s and then correlate the data. This is especially helpful when comparing and contrasting similar articles. It allows users to see which keywords the articles have in common and the average words per sentence of each document.
Voyant Tools has a strong community supporting it due to this versatility. It’s available in ten different languages and popular all over the world. Hermeneuti.caM1 proved this in 2016 when it reported the site had 80,000 views from over 150 different countries. Its code is available on GitHub and it’s constantly improving. Yet, there are still many issues need resolving. For example, many skins and tools are not yet usable and tool help buttons are small, difficult to find, and not always understandable.
Although the help button is unnoticeable on the Voyant Tools homepage, their website offers a more readable help section. It has many guides, each with individual drop-down options to allow users to quickly and easily find what they’re looking for.
Once a user clicks on a link, it will appear in a separate tab. This is a small but extremely useful feature that prevents users from losing where they are on a particular page. This section gives a guide on how to start, an overview on available tools and general information about Voyant Tools.
This tool is well documented and has many research papers and reviews based on it. For example:
- The University of Colorado Boulder2 reviewed this tool as part of the “Information Literacy Commons” project. In it they declare that “most text analysis tools are not designed for average humanities scholars”. They go on to say that “fortunately, these complaints regarding ease of use are generally unfounded with Voyant Tools”. They praise it’s usability and uncomplicated dashboard. The only negative remark the University had was it’s “occasionally prolonged text-loading time”. However, Voyant Tools are currently addressing this issue and expect to resolve it soon.
- Beth Platte, language and technology professor at Reed College, Oregon, conducted a more in-depth review of the website3. She starts the report by immediately declaring “Voyant Tools is one of my favourite text analysis tools because it is fast and easy to use”. She expands this by referencing its interface and usability, before going into further detail. This is a prevailing theme throughout these reviews. It is clear that Dr. Platte has a great respect for this tool, both as a humanist and as a computer scientist.
- Missouri University conducted an article which focused on the effect of digital tools on librarians6. They examined many websites, one of which was Voyant Tools, reporting how it could benefit libraries globally. “Voyant is more sophisticated than other web-based text analysis tools and easier to use”. They found it the perfect middle ground between expensive software and basic technology. It allows librarians to digitise and examine vast collections of novels and documents all over the world.
Yet despite these praising reviews, Voyant Tools is not the only popular data visualisation tool.
Another commonly used site is TAPoR, a set of text analysis tools that examines inputted text. It offers more options than Voyant Tools, with the website divided into HTML, XML, and plain text sections for the user to choose from. However, compared to Voyant Tools visually, it is sub-par and less experienced users may struggle to find what they’re looking for. As well as this, it is only possible to examine one document at a time, limiting its versatility and increasing the time it takes to use the site.
A more user-friendly tool is RAWGraphs, an open source data visualisation network. It was developed to provide the missing link between spreadsheet and graphic editors. Once again, it favours examining one document rather than many, but what it does display, it displays exceptionally. It offers dozens of charts and graphs to portray data. It allows users to create customisation diagrams, which greatly surpasses Voyant Tools.
Textalyser is a text mining tool that provides detailed statistics of inputted documents. It provides many results from word count to average syllables per word to the most frequent words used. It has a simple, basic design with minimal colours, making it pleasing to look at. Voyant Tools doesn’t offer as many details as Textalyser, but does provide more visually appealing results.
Despite Voyant Tools being the most advanced data visualisation tool, there are still areas I believe need improving. More language options are necessary, as well as a scroll bar to navigate them as some options are currently inaccessible at the bottom of the page. It doesn’t allow users to pause projects and later amend them, as there is no way to save the information once you leave the website. These are issues that Voyant Tools should address sooner rather than later to increase it’s user-base.
Based on my research and use of this tool, I recommend it wholeheartedly for researchers, academics and students alike. It displays a user-friendly interface while analysing and displaying texts in great detail. Its tools are straightforward to use, making it remarkably versatile for less-experienced users. It is an invaluable tool for digital humanists and others aiming to interestingly yet clearly display data. It allows users to provide evidence of patterns in various texts and contextualize trends in the usage of certain words. Voyant Tools 2.0 is without a doubt a digital humanities tool I will continue to use in the future.
- Voyant Facts, http://hermeneuti.ca/VoyantFacts
- Welsh, Megan E. (2014) “Reviewing Voyant Tools,” Collaborative Librarianship: Vol. 6 : Iss. 2 , Article 8. https://digitalcommons.du.edu/cgi/viewcontent.cgi?article=1105&context=collaborativelibrarianship
- Platte, B. (2017) “Text Analysis using Voyant Tools”, https://blogs.reed.edu/ed-tech/2017/03/text-analysis-using-voyant-tools/
- Miller, A.(2018) “Text Mining Digital Humanities Projects: Assessing Content Analysis Capabilities of Voyant Tools, Journal of Web Librarianship”, 12:3, 169-197, DOI: 1080/19322909.2018.1479673
- Gallant, K., Lorang, E., Ramierz, A. (2014) “Tools for Digital Humanities: A librarian’s guide”, https://mospace.umsystem.edu/xmlui/bitstream/handle/10355/44544/ToolsForTheDigitalHumanities.pdf?sequence=1
- Calado, F., (2018) “Using Voyant-Tools to Formulate Research Questions for Textual Data”, https://digitalfellows.commons.gc.cuny.edu/2018/11/01/using-voyant-tools-to-formulate-research-questions-for-textual-data/