direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content

Crowdsourcing and Open Data

Core Team:

  • Tim Polzehl [1]
  • Babak Naderi [2]
  • Rafael Zequeira Jiménez [3]
  • Neslihan Iskender [4]
  • Vinicius Woloszyn [5]


Besides others, we work on building state-of-the-art methods for conducting valid, reliable, and reproducible Subjective Tests using Crowdsourcing for different media: speech, video, gaming, text. Our group is actively participating in the standardization activities in ITU-T Study Group 12, leading and participating in different Working Items, including P.Crowd [6] (speech and crowdsourcing), P.CrowdV [7] (video and crowdsourcing), P.CrowdG [8](gaming and crowdsourcing).

Our research can be categorized as the following:



  • Fundamental Aspects of Crowdsourcing Platforms
  • High Quality Crowd-Workflow Design
  • Combination of Human Computation and AI
  • Open Data, and Open Science (open-science.berlin [9])


  • Motivation of Workers
  • Gamification in Volunteer Crowdsourcing
  • Personalized/adaptive Design to Increase Performance
  • Quality Control Mechanisms (Data Reliability, Agreement)

Application of Crowdsourcing

  • Speech Quality Assessment using Crowdsourcing approach
  • Video Quality Assessment using Crowdsourcing approach
  • Gaming QoE Assessment using Crowdsourcing
  • Text Quality and Complexity Assessment using Crowdsourcing
  • Crowd- and AI-based Hybrid Workflows (error correction and performance boost by adding crowd-services to support AI) 
  • Crowd- and AI-based Translation Workflows
  • Crowd- and AI-based Text Summarization Workflows
  • Crowd- and AI-based Knowledge Graph and Chatbot Supporting Workflows
  • Crowd- and AI-based Chatbot/ Dialog Flow Supervision and in-time Correction
  • Crowd- and AI-based Information Learning (autonomous knowledge base updates)
  • Internal Crowdsourcing (employee sourcing)
  • Text Simplification and Text Complexity: einfaches-wiki.de [10]


Our research extends to the following areas:

Building on Crowdsourcing

  • Mobile crowdsourcing (in the field)
  • Crowd assessments: Usability, UX, QoE
  • Privacy and confidentiality in crowdsourcing
  • Mobile street application, urban mobility, city guarding
  • Data collection in the field: crowd as (continuous) sensors
  • Data management (clean, index, verify, tag, label, translate, etc. )

Improving Crowdsourcing

  • Real-time interaction, human computation as a service, (HuaaS)
  • Privacy and security in crowdsourcing
  • Motivation in crowdsourcing, gamification
  • Quality control (pattern recognition, cheater detection, anomaly)
  • Automatic user segmentation (clustering)
  • Training, E-learning and building expert-crowds
  • Task complexity modeling
  • Crowd and user biases, subjective normalization
  • Scalable Crowdsourcing, Robustness, Reliability in Engineering
  • Quality in Crowdsourcing (quality of opinion, audio/video, reliability) 


  • Crowdee [11]:  High Quality Large Scale Crowdsourcing for Studies and AI-related Data Acquisition: (Start-Up from QU TU Berlin)


  • SMESS - Towards a Standardized Methodology for Evaluating the Quality of Speech Services using Crowdsourcing [12]
  • RUBYDemenz - Robot mit Begleitung (BMBF) [13]
  • Automated Chatlog Analysis for Self-Learning NLU and Dialog Update in Customer Support Domain (DFKI)
  • BOP - Berlin Open Science Platform for the Curation of Research Data (TU Berlin)
  • DEKA -  [15]Design und Entwicklung einer kollaborativen digitalen Arbeitsplattform für die Digitalisierung von Innovationsprozessen (BMBF) [16]

Past Projects:

  • BRIDGE - Data & Fact Driven Decision-Making for Skills Based Inclusion of Migrants (EIT-Digital)
  • DoNotFear  [18]- Perceived Security In Public Transport [19] (EIT-Digital) [20]
  • OurPuppet - Pflegeunterstützung mit einer interaktiven Puppe für informell Pflegende (BMBF) [21]
  • ICU - Internes Crowdsourcing in Unternehmen: Arbeitnehmergerechte Prozessinnovationen durch digitale Beteiligung von Mitarbeiter/innen (BMBF) [22]
  • ERICS – European Refugee Information and Communication Service (EIT-Digital, 1/2017- 12/2018, Project Lead) [23]
  • ALM-enabled Smart Maintenance: Low cost, Multi-purpose (ALM) IoT modules for fitting machinery/production plants and measuring real time parameters such as vibrations, energy consumption, temperature etc. based on innovative fiber optics and Nucleo STM microcontrollers (EIT-Digital, 1-12/2017) [24]
  • Privacy, Security and Trust in Crowdsourcing Confidential Enterprise Data (EIT Digital, Project Lead) [25]
  • CrowdMAQA [26] (Motivation and Quality Control in Crowdsourcing)
  • AUNUMAP [27] (Automated User Segmentation from Speech and Text for Market Research Applications)
  • Vocalytics & SWYM [28] (Fully Automated User Characterization and Personality Estmation)
  • Speaker Recognition and Speaker Characterization through different Communication Channels [29]

  • Affect-based Indexing [30]

  • Anomaly Detection and Early Warning Systems [31]
  • Multimedia Content Retrieval [32]
  • Predicting the Perceived Quality of Audiovisual Speech (Perc Qual AVS)
  • Recognition of Mobile and Rich Speech (MARS) [33]
  • Universal Telecommunications Interface [34]
------ Links: ------

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Copyright TU Berlin 2008