[Curator's note: there are numerous other technical challenges  addressed with alternative prescriptions throughout the paper. These highlights are narrative-centric, and I invite you to review the paper if you are a keen technologist looking for answers to the following:- Zero- and Few-Shot Learning - Response Generation- Arith... See more

This represents a fundamentally different way of thinking about IR systems. Within the index-retrieve-then-rank paradigm, modeling work (e.g., query understanding, document understanding, retrieval, ranking, etc.) is done on top of the index itself. This results in modern IR systems being comprised of a disparate mix of heterogeneous models (e.g., ... See more

- Leveraging Document and Corpus Structure-  Scaling to Multiple Languages]

- The problem is one of content. The misconception is that without deep content, design is reduced to pure style, a bag of dubious tricks. In graphic-design circles, form-follows-function is reconfigured as form-follows-content. If content is the source of form, always preceding it and imbuing it with meaning, form without content (as if that were ... See more

If all of these research ambitions were to come to fruition, the resulting system would be a very early version of the system that we envisioned in the introduction. That is, the resulting system would be able to provide domain expert answers to a wide range of information needs in a way that neither modern IR systems, question answering systems, o... See more

When experiencing an information need, users want to engage with a domain expert, but often turn to an information retrieval (IR) system, such as a search engine, instead. Classical information retrieval systems do not answer information needs directly, but instead provide references to (hopefully authoritative) answers.

[...] Today’s cutting edge IR systems are not fundamentally different than classical IR systems developed many decades ago. Indeed, a majority of today’s systems boil down to: (a) building an efficient queryable index for each document in the corpus, (b) retrieving a set of candidates for a given query, and (c) computing a relevance score for each ... See more

To replace indexes with a single, consolidated model, it must be possible for the model itself to have knowledge about the universe of document identifiers, in the same way that traditional indexes do. One way to accomplish this is to move away from traditional LMs and towards corpus models that jointly model term-term, term-document, and document-... See more