Research

My research is focusing on augmenting human creativity: How can we build computer systems that help people to come up with better ideas? Within this framework, I'm currently involved in three projects:

Creativity Enhancing Interventions in a Crowdsourced Ideation Context

Although crowdsourced ideation (obtaining ideas by employing large numbers of people to generate them, typically via an online platform) has shown great success in the past. One caveat of this method of idea generation is that participants tend to produce ideas that are repetitive, expected and of varying quality.

One approach to tackle these challenges are creativity enhancing interventions (CEI). CEIs refer to stimuli (e.g. images, or texts), that are shown to the user during an ideation session, aimed to enhance the creativity of ideas submitted by the user (where creativity is usually operationalized as a combination of novelty and usefulness). Research has shown that providing these interventions (also called inspirations) to users potentially benefits the average novelty and diversity of the ideas provided. Therefore, one goal of my current research is to advance the state of the art in CEI systems.

An Idea Knowledge Graph

In order to enhance and augment crowd-powered large scale ideation, we envision an idea-based knowledge graph, leveraging semantic meaning about ideas to model multi-dimensional relations between ideas. Furthermore, by linking not only relationships between different ideas, but also relationships between different versions of the same idea, we envision a computationally accessible Idea Lifecycle, capturing iterations, refinement and different branches of ideas created during an innovation process.

When applied in a crowd ideation scenario, this idea-based knowledge graph could provide information about categories explored by the ideators and lead ideators to not well explored categories, help select tailored inspirations depending on information gathered for each ideator and the overall problem context, and aid in interactive visualization of the ideation outcome.

One promising use case for this idea-based knowledge graph is the computational analysis of inspiring example ideas, shown to crowd workers during ideation: In preliminary studies, we could show that inspirations are integrated into the ideators state of mind either directly (re-using the same term), indirectly (using a super-class of a term), or via the transfer of the use of a term.

Information extraction via semantic annotation

In order to get to the idea knowledge graph described above, we need ways to obtain a structured representation of ideas. One way to construct those representation is via Semantic Annotation. The annotation of words within documents furthermore enables the extraction of knowledge stored in general knowledge graphs. I am working on this topic within two projects:

neonion

neonion is user-centered, web application for the collaborative annotation of texts. It was developed as a research artifact to understand how he human user and the machine work hand in hand to make your annotation tasks effectiv, efficient, and satisfying. You can find more about the project here.

Interactive Concept Validation

To enable semantic annotation in a large scale ideation concept, we developed a software, that for a given input text (e.g. the input idea in an online brainstorming) searches for concepts for all terms in the idea using a general knowledge graph. Based on the answers from the knowledge graph, the software than asks the user to disambiguate their terms and safes the resulting annotations.

You can find more about the ICV software either on the github page , or by reading the related publication titled "Discovering the Sweet Spot of Human-Computer Configurations: A Case Study in Information Extraction".

Publications

2020

  1. Designing Crowdsourcing Innovation effectively: The Effects of Problem Formulation on Idea Originality, Feasibility, and Quantity

    Marc Karahan, Maximilian Mackeprang, Michael Tebbe

    The International Society for Professional Innovation Management (ISPIM)

2019

  1. Graduate Symposium: Understanding and Augmenting Ideation Processes

    Maximilian Mackeprang

    The 12th ACM Conference on Creativity & Cognition

    [link]
  2. Leveraging General KnowledgeGraphs in Crowd-poweredInnovation

    Maximilian Mackeprang, Abderrahmane Khiat, Claudia Müller-Birn

    Workshop on Designing Crowd-powered Creativity Support Systems

    [link]
  3. Discovering the Sweet Spot of Human-Computer Configurations: A Case Study in Information Extraction

    Maximilian Mackeprang, Claudia Müller-Birn, Maximilian Timo Stauss

    The 22nd ACM Conference on Computer-Supported Cooperative Work and Social Computing

    [link]

2018

  1. Concept Validation during Collaborative Ideation and Its Effect on Ideation Outcome

    Maximilian Mackeprang, Abderrahmane Khiat, Claudia Müller-Birn

    Extended Abstracts of the 2018 CHI Conference on Human Factors in Computing Systems

    [link]
  2. Generating Structured Data by Nontechnical Experts in Research Settings

    Andre Breitenfeld, Maximilian Mackeprang, Ming-Tung Hong, Claudia Müller-Birn

    i-com Journal of Interactive Media

    [link]

2017

  1. Enabling Structured Data Generation by Nontechnical Experts

    Andre Breitenfeld, Maximilian Mackeprang, Ming-Tung Hong, Claudia Müller-Birn

    Mensch und Computer 2017

    [link]
  2. Semantic Annotation for Enhancing Collaborative Ideation.

    Abderrahmane Khiat, Maximilian Mackeprang, Claudia Müller-Birn

    Proceedings of the 13th International Conference on Semantic Systems. ACM, 2017

    [link]