Redefining New York Public Library’s Digital Search Experience with the Enhancement of AI
Redefining New York Public Library’s search experience with AI: If your team was capable of implementing Artificial Intelligence, how does that affect the design in redefining the search experience of a website?
Over the past week, I focused my efforts on looking into New York Public Library as my fictional client to improve their digital experience. I started my research on the topic of AI and discovered 3 main ways a big public organization can benefit it’s users. More specifically, the evolution of deep learning: has the ability to increase our capacity to be more inclusive, efficient and relevant in helping our users achieve their objectives.
How?
- Speech and image input that also accounts for multi languages
- Predictive analytics engine
- Recommendations
Even with the restricted time frame, I still felt the need to do a specific user persona. Although I was not able to take into account for a wide demographic of interviews working alone, my approach got me to understand 2 users personal involvement, objectives and challenges that they faced within their past decade being NYPL members.
For one of my users, some of her main objectives , was the need to borrow children books for curriculum, to renting movies for movie night with her friends and most frequently, to find a quiet working space to get productive during those college days.
I decided to investigate as to why 1 of my users that I interviewed insisted on going into the library to seek out resources rather than search online. When trying out the current search engine to conduct a simple heuristic analysis, I started realizing some issues that other users may find difficulty with, and connect the dots as to her minimal involvement with the website:
- Current search experience
- Filtering wizard
- Advanced searchProblem statement:
Current search Bar was not the most dynamic to user intentions. Currently, it is placed in the secondary navigation and need to be clicked in order for it to be used. On top of that, the search filter that assists in user’s search fill overloaded and gives off a messy first impression that isn’t friendly to narrowing out the search.
As for the recommendation and advance search engine, it was similar in terms design being not user friendly. Currently, users could type in an additional keyword to narrow his search, and the other factors to narrow down search are restricted to searching by collection, format and location, But this doesn’t explore the variety of options within a particular format.
So something I wanted to solve for in my redesign was the accessibility of the site and making it more intuitive, bearing in mind that this website is a representation of 92 locations and caters to one of the most diverse ethnic groups.
So I explored and expanded on these 3 main concepts, in order for users to help them in their search, while unfolding more of their stories that I had felt friction while learning more about what they do, and laying it out in a continuous scroll so users stay on the main landing page and still get a good overall picture of what they do and how they can get involved with their main activities and be touched with a meaningful engagement.
Firstly, I had the idea of NYPL welcoming the user with a warm greeting, based on pull quotes from famous literary and public figures that could possibly inspire a user to borrow a book from the landing page, and reinforced that with placing popular books that are trending right now.
I simplified the main navigation into 4 main options, of user’s frequent habits and options, and removed the secondary navigation to and have it panned out on the website to be explored while staying on the page.
Search engines can leverage on predictive analytics to generate more accurate, relevant search results specific to users. They’re also used to offer alternative search terms that could be used for user’s query for guidance.
The advanced search, was replaced with themes to pick from, and open ended relevant factors that can help users identify with. The site also takes into account the native speakers language, allowing him to search in his own native tongue or type it in his language, accounting for the ethnic diversity it represents, as well as users who might have taken a photo reference.
Use of images opens up opportunities for users to locate resources that they might have forgotten titles but have taken a photo of it.
I realize that the current site’s secondary navigation, each header had to be clicked in order to find out more, and decided to showcase that information by simply hovering over the picture.
With reference to a map, users can immediately locate their nearest library branch if they had intentions of visiting, read reviews and find their contact to reach out for more information.
Through the redesign of their site, I had to dive pretty deep to find so many literally hidden gems within the site, that includes weekly podcasts from 2 different stations, library talks and me in. Throughout this time, I was listening to the real Lolita: one of their latest podcast episodes that revealed to me the importance for users to have an immediate encounter with the site if given the option, which is made playable from the latest blog updates.
Calendar schedule was a great way to summarize all ongoing events in chronological order that considers users who had a time frame being in the city. A good example would be a tourist visiting over the weekend and wanted to engage in one of NYPL activity. Currently, it was hard to align all the activities visually especially for a user looking for upcoming events.
I changed the layout of the current events and talks into a larger resolution to be attention grabbing, and create a carousel scroll at the bottom to browse through other events.
Last but not least, learning capabilities is another aspect I thought could benefit current members: by proposing new books to borrow, users are encourage to stay in the loop with New York Public Library, having the next relevant book recommended to users based on their preferences.
