A major goal of collaborative ideation is improving the creativity of the ideas generated. Recent approaches enhance creativity by showing users similar ideas during productive ideation and diverse ideas when they reach an impasse. However, related work either demands a higher mental effort from users to assess similarity or yields only a limited number of similarity values. Furthermore, idea relationship is only considered in one dimension (similarity). In our research in progress, we introduce a new approach called concept validation. It enables us to (1) capture the conceptualization of users' ideas and (2) assess multi-dimensional relationships between all ideas in near real-time. We conducted a study with 90 participants to validate the suitability of our approach. The results indicate that adding the extraction of semantic concepts to the ideation process has no negative impact on number and creativity of ideas generated. This signifies an important step towards our vision of an idea-based knowledge graph used by an interactive system to improve computer-supported human creativity.