Language Scent: Exploring Cross-Language Information Navigation

Abstract

While multilingual users often switch between languages when seeking information, this process remains undersupported by current systems where information is typically siloed by language. Our formative study reveals that users select their search language based on its perceived value for their current information need, a concept we formalize as language scent. Language scent extends Pirolli and Card's information foraging theory - which explains how users navigate among already-encountered sources - to the multilingual case, where users' choice of query language determines which sources they can encounter in the first place. Building on this insight, we designed Niffler, a multilingual information seeking system that provides proximal cues for gauging the language scent of different languages. Finally, we conducted a lab study with 16 multilingual speakers to understand Niffler's utility, usage patterns and application contexts.