Recent Events

Archive | Internet Librarian

Getting Smart About Search

Monday afternoon in the Discovery, Navigation, and Search Track ended with a discussion between a veteran searcher, Marydee Ojala (that would be me), and a current library school student, Sheridan Reid. Sheridan also works with 5th graders. We started from the position that search technology does not stand and that obsolescence is inevitable. Sheridan noted that her generation and the children in elementary school are phone-centric. Marydee talked about the AI technologies, particularly machine learning and predictive analytics that are being used to determine relevance.  Boolean search strategies, the reliance on controlled vocabulary, and copyright restrictions have little meaning to younger people.  That’s not all bad, since Boolean doesn’t really work with web search engines and controlled vocabulary is giving way to contextual and conceptual analysis or word choice.  As for ignoring copyright, that may not be generational at all.

What’s clear is that the convenience factor of getting to information, whether it’s on the web or in a subscription database, is very important. Sheridan made the point that not everyone she knows is wedded to online information. Sometimes a print book is more appropriate than an ebook. Both speakers see a future filled with multimedia searching.

The session ended with thoughts on “magical thinking.” Everyone searches well. Everything you find on the internet is true (or nothing you find is true). Searching library databases provides comprehensive answers. All three are magical thinking.

 

Robots in the Library

The presence of robots in the library was addressed in four different environments. Cindy Hill talked about the Telepresence robot at the Federal Reserve Bank of San Francisco. One issue surrounding the robot are security, both in terms of roaming and hacking. Perception of robots as gimmicky was another issue, as was maintenance.  Cindy thought that ownership of the robot should fall under the library but it went to IT that, ironically, will loan out the robot. Interestingly, she did not find cost to be an issue.

Bonnie Roalsen, John Walsh, Rebecca Meehan, at Woburn Public Library, have a robotics literacy program where they teach coding and robotics. They assured the audience that robots can be cute, not scary, and noted that coding has become a lot easier now that it’s done in blocks instead of interminable lines of code.

Jason Griffey talked about strategies and perceptions regarding technologies that take control of a physical activity and put it into the hands of machines. He worries about how we teach people to interact with robots and thinks that job replacement will be an inevitable outcome of increased automation of library jobs.

Dewey is the programmable robot at Palo Alto Public Library. and they will soon add Misty. Dan Lou described the coding events at the library. Dewey has been taught to dance, do pushups, and give high-fives, largely because of requests from children learning to code. For those thinking of buying a robot, she warned attendees to be aware of the platform on which the robot runs (Dewey and Misty are on different platforms) and that robots, like computers, get updated and can become obsolete when a new version appears.

.

 

Smart Search: The New Frontier

Greg Notess may be retired from Montana State University Library but he’s not lost his interest in search technology. He reviewed some new directions that search is taking. Overall changes include search engines putting mobile first, discover and push emphasized over search, the devolution of searchbox to question mark , search becoming hard to find (the magnifying glass icon), and sites with no search capabilities, such a Springshare or no “about us” (looking at you, MarketWatch.com).

Search input can now be graphical, as with reverse image searching, or vocal, when you speak your search query. In the future, we may see gestures as controlling search. Machine learning is closely tied to your location, search history, profiled information. Many of the AR and AI technologies Greg showed seem to be more relevant to shopping than libraries, although the Google Maps new functionality that lets you see your route when walking has general interest.

 

Artificial Intelligence for Libraries

Daniel Lee and Gary Price started off the AI, Robots & Machine Learning track with their version of AI101, defining the basic technologies of AI. The overall goal of AI is for machines being able to replicate human thought processes. Natural Language Processing (NLP) is the ability of machines to read and understand language. Machine learning is all about algorithms and statistical modeling.

They have lots of ideas about the use of AI in libraries, among them self-checkout, which would free up circulation staff to do other, more challenging work. What about self-service that responds to your user’s profile and needs. To build the profile, you could use machine learning to mine previous requests.

Anyone thinking about how to onboard AI teammates?