This paper presents the ANR TALAD project that aims to demonstrate the added value of Discourse Analysis (DA) and Natural Language Processing (NLP) for collaborative work. It focuses on “nominations” - a concept commonly investigated in DA. This concept refers to the way in which speakers name things with a meaning that contains some semantic appreciation (like an opinion) - typically by reusing denominations that already exist (for instance "migrant" versus "refugee") but whose meaning is somehow updated by their context of appearance. The new meaning entirely depends on the speaker’s deictics and provides access to his ideology. The project aims, on the one hand, at examining existing literature in DA to help characterize the concept and, on the other, to show what NLP contributions could be proposed to easily detect and analyze nominations.