Modeling information relevance aims to construct a conceptual understanding of information significant for users’ goals. Today, myriad relevance estimation methods are extensively used in various systems and services, mostly using behavioral signals such as dwell-time and click-through data and computational models of visual or textual correspondence to these behavioral signals. Consequently, these signals have become integral for personalizing social media, search engine results, and even supporting critical decision making. However, behavioral signals can only be used to produce rough estimations of the actual underlying affective states that users experience. Here, we provide an overview of recent alternative approaches for measuring and modeling more nuanced relevance based on physiological and neurophysiological sensing. Physiological and neurophysiological signals can directly measure users’ affective responses to information and provide rich data that are not accessible via behavioral measurements. With these data, it is possible to account for users’ affective experience and attentional correlates toward information. Modeling information relevance aims to construct a conceptual understanding of information significant for users' goals. Today, myriad relevance estimation methods are extensively used in various systems and services, mostly using behavioral signals such as dwell-time and click-through data and computational models of visual or textual correspondence to these behavioral signals. Consequently, these signals have become integral for personalizing social media, search engine results, and even supporting critical decision making. However, behavioral signals can only be used to produce rough estimations of the actual underlying affective states that users experience. Here, we provide an overview of recent alternative approaches for measuring and modeling more nuanced relevance based on physiological and neurophysiological sensing. Physiological and neurophysiological signals can directly measure users' affective responses to information and provide rich data that are not accessible via behavioral measurements. With these data, it is possible to account for users' affective experience and attentional correlates toward information.