Is discovery the researcher’s dream?

A special issue of UKSG Insights on discovery (Discovery is the Researcher’s Dream)  sounded promising, but it turned out simply to be a collection of 27 earlier articles published in the journal over the last nine years or so. Can anything useful be said about a grouping that was never designed to be read together? Rather than read all 27 articles, I looked to the introduction (by Magaly Bascones and Rebekah Cummings) for some kind of summary. In only 800 words (not including the abstract, which repeats the first paragraph of the article) you can’t expect anything comprehensive, but the main points seemed to come across.

The real issue is how library catalogues have been replaced by Google Scholar for initial searching for articles. The Renew Consultants Report (How Readers Discover Content in Scholarly Publications (Inger and Gardner, 2021) shows how Google and Google Scholar together represent by far the most common starting point for researchers discovering content: together, they are more than used more than three times as much as “library web pages” (which I assume means the online library catalogue).

So how should libraries respond? Let’s be frank: despite the title, Discovery is not “the researcher’s dream”. Only search-engine specialists dream about discovery. Researchers dream about what they discover, not the means by which they discover it. Incidentally, this is one of the maddening (but fulfilling) things about UX: your users can’t understand why you are so concerned. For them, the world is full of inadequate websites, and we muddle through somehow.

The dream of a straightforward search that allows information seekers to find the content they are looking for and, more importantly, relevant content they do not yet know about.

Bascones and Cummings reveal the two main aspects of discovery: search and serendipity. Libraries would like to provide both, as long as they don’t get in each other’s way.

The introduction describes solving discovery problems as “motivated by rationales that vary from the noble goals of knowledge creation and sharing to profit-driven commercial grounds”. That’s an odd contrast, suggesting that profit-driven commercial grounds are somehow not noble. I’d like to think you can promote the noble goals of knowledge and make a profit at the same time.

One thing that has not changed is the centrality of discovery to the academic user journey. Where they search might have changed, but, remarkably, a large proportion of discovery of content appears to be via searching. But, as the editors state, what has changed is how traditional library catalogues compete with or are replaced by search engines. Ever since Google came onto the scene, as Edward Chamberlain states, “discovery of academic material is increasingly happening outside of the library search domain”.  

You could always argue (although you might be accused of changing the subject) that users don’t use search engines properly. Some of the articles mentioned here reveal how users show poor search strategies (Janyk, 2014). One article commented:

According to the collected data, searchers rarely used Boolean terms to expand or narrow their search. Individuals were more inclined to begin a new search with different search terms rather than rerunning the same search and incorporating Boolean terms. This behaviour is consistent with other findings that have shown when students are unable to find a relevant item on the first page of search results, they reword the search terms and rerun the search (Cassidy, 2014)

You could say this is just sour grapes. In all likelihood, you would probably find in a survey that students didn’t use the online library catalogue properly either – or perhaps they migrated away because they were not satisfied with what information was available.

The big question is, given this state of affairs, what should the library do? The librarians’ answer was initially defensive: “some [users] were missing the simplicity of library catalogue searches and were feeling overwhelmed by the long list of results and access options.” There are various other suggestions:

• Provide users with more information literacy, or talk with course providers about reading lists.
• The University of Bremen amended the library interface to take usage into account. More popular items are ranked higher. This shocked me: certainly Amazon, Netflix, and so on, all have search based on usage, but an academic library? Possibly if students on a specific course could recommend items from the reading list they found useful – but this is not what Bremen described.
• Open the library sites for indexing by the big search engines. After all, “Aggregate to disseminate” – the bigger the collection, the more likely users are to make use of your resource.
• Edward Chamberlain suggests adding schema.org tags to web pages.
• Or that we make use of the how OCLC is “leading the way” in making library collections more widely accessed. Unfortunately, my experience of trying to make use of OCLC’s WorldCat interface were less than successful.
• More radical is to abandon the library OPAC website altogether (Kortekaas and Kramer, 2014). That seems radical indeed, in fact their article is entitled “thinking the unthinkable”, and you want to look to see if all the other libraries have taken up the suggestion and done likewise – but they haven’t.
• Social search: The two editors believe that knowing why material is useful to you equates to recommendations and social interactions. They talk about a “national-level aggregation” – but who would run such a system? And who decides what is a trusted source?
• One suggestion seems by the editors seemed to be that libraries should simply push their data out to Facebook, where researchers are busy looking for academic content. I don’t use Facebook much, but even if Facebook were widely used by academics, I don’t think it is the site where many of them will initiate their search for content.

So, any conclusions? I note that the six (presumably human-generated) keywords provided for the introduction “metadata”, but not the word “Google” or “Google Scholar”. In my opinion, a concept-based (or full-text) search would identify this article better than by using these keywords – that seems to tell you something about what makes content discoverable.  

I tried a university library catalogue myself, and made my own suggestions for how to improve them. Certainly, I’m not surprised that users go to the simplest, easy-to-use interface available: Google and Google Scholar. But I still don’t think that researchers have dreams about Google – and nor would Archimedes, if he were alive today.