2017/02/28 by Kevin Jasberg, Jasberg, Kevin, Sergej Sizov +1
Computer Science · #Data Stream Mining Techniques #Data Visualization and Analytics #FOS: Computer and information sciences #Human-Computer Interaction (cs.HC) #Time Series Analysis and Forecasting
paper · pdf · doi:10.48550/arxiv.1702.08826
openalex publication_date 2017/02/28 · openalex created_date 2022/10/04 · openalex updated_date 2026/07/28
In many areas of data mining, data is collected from humans beings. In this\ncontribution, we ask the question of how people actually respond to ordinal\nscales. The main problem observed is that users tend to be volatile in their\nchoices, i.e. complex cognitions do not always lead to the same decisions, but\nto distributions of possible decision outputs. This human uncertainty may\nsometimes have quite an impact on common data mining approaches and thus, the\nquestion of effective modelling this so called human uncertainty emerges\nnaturally.\n Our contribution introduces two different approaches for modelling the human\nuncertainty of user responses. In doing so, we develop techniques in order to\nmeasure this uncertainty at the level of user inputs as well as the level of\nuser cognition. With support of comprehensive user experiments and large-scale\nsimulations, we systematically compare both methodologies along with their\nimplications for personalisation approaches. Our findings demonstrate that\nsignificant amounts of users do submit something completely different (action)\nthan they really have in mind (cognition). Moreover, we demonstrate that\nstatistically sound evidence with respect to algorithm assessment becomes quite\nhard to realise, especially when explicit rankings shall be built.\n