Can Semantic Search be more interpretable? COLBERT, SPLADE might be the answer but is it enough?

Warning : I am not a information retrieval researcher, so take my blog post with a pinch of salt In my last blog post, I described a simplified description of a framework for infomation retrieval from the paper -

tl;dr : metrics-obsessed capitalism is the reason why we can’t have nice things – and watch @tante’s talk Back in Paris, I was semi-joking with friends about how the rate of technological innovations seem to have slowed down or come to a halt. The last “big” innovation that we could think about were smartphones, as carrying the internet permanently in our pockets was quite a change.

Wenn sich im Juni 2024 die Bibliotheksszene des DACH-Raums in Hamburg auf der BiblioCon treffen wird, ist auch die Redaktion der LIBREAS. Library Ideas dabei (zumindest zum Teil). Gerne treffen wir dort auf unsere Leser*innen und Autor*innen. Offenes Treffen (05.06.2024) Am Abend des Mittwoch trifft sich die Redaktion LIBREAS.

As academic search engines and databases incorporate the use of generative AI into their systems, an important concept that all librarian should grasp is that of retrieval augmented generation (RAG).   You see it in use in all sorts of "AI products" today from chatbots like Bing Copilot, to Adobe's Acrobat Ai assistant that allow you to chat with your PDF.

Enoch’s Hammer of Luddite fame, at the Tolson Memorial Museum. I gave a talk at the Pint of Science Festival – an international science festival that takes places at local pubs and cafes across the world – this evening.

Let's be clear here, Google Scholar is ill designed for use for systematic reviews . I am not trying to argue otherwise. (Obligatory warning, I am not a real systematic review librarian) But why exactly?

The pews of the Internet Archive back in 2018. tl;dr: Posts on this blog are now automatically archived, indexed and full-text searchable through The Rogue Scholar . The jury might still be out on whether the small or indie web will make a comeback, but I’ve personally enjoyed posting more on my blog here in recent months.

The dynamic footer of my website has been powered by a little aggregation of some of my personal data for about 5 1/2 years by now. Until recently, all the data related to my activity and physiology (steps, heart rate, sleep, body) came from an Oura Ring.

In the last blog post , I argued that despite the advancements in AI thanks to transformer based large language models, most academic search still are focused mostly in supporting exploratory searches and do not focus on optimizing recall and in fact trade off low latency for accuracy.