TY - BOOK

T1 - Usability and User Experience Information in Reviews

AU - Hedegaard, Steffen

PY - 2014

Y1 - 2014

N2 - Product reviews available online contain information on users' interactionwith the product in the form of narratives. These narratives describe theusers' experiences with using the product in the wild, that is, in real usagesituations. The large amount of reviews available makes for a potentiallyinteresting source of information related to product use.This thesis addresses the following research questions about the informationavailable in reviews:1. How do users describe their interaction with products in narrativessuch as product reviews?2. Can such information be used to improve future versions of the prod-uct?The results show that users write about product use in terms related tostandard and popularly researched aspects of usability and user experience,e.g. effiiency, effectiveness, enjoyment, frustration. The frequency withwhich different aspects are depicted in reviews differs significantly betweenproduct domains. We also find that reviews contain descriptions of morepersistent usability issues. I devise automatic methods for classifying sentenceswith regard to dimensions of both usability and user experience andusability problems and perform a linguistic analysis of the content.To assist with the automatic classification tasks, I compared supervisedmachine learning algorithms with semi-supervised learning (SSL) algorithmsfor text classification on several standard corpora. This comparison showsthat support vector machines normally are the best choice for text classification.I also find that traditional feature vectors consisting of counts of functionwords can be used for identifying the author of a translated document but themethod can be augmented by adding semantic information from documents.Overall the results presented in this thesis help clarify the role of reviewsin relation to understanding both users and different aspects of product use.

AB - Product reviews available online contain information on users' interactionwith the product in the form of narratives. These narratives describe theusers' experiences with using the product in the wild, that is, in real usagesituations. The large amount of reviews available makes for a potentiallyinteresting source of information related to product use.This thesis addresses the following research questions about the informationavailable in reviews:1. How do users describe their interaction with products in narrativessuch as product reviews?2. Can such information be used to improve future versions of the prod-uct?The results show that users write about product use in terms related tostandard and popularly researched aspects of usability and user experience,e.g. effiiency, effectiveness, enjoyment, frustration. The frequency withwhich different aspects are depicted in reviews differs significantly betweenproduct domains. We also find that reviews contain descriptions of morepersistent usability issues. I devise automatic methods for classifying sentenceswith regard to dimensions of both usability and user experience andusability problems and perform a linguistic analysis of the content.To assist with the automatic classification tasks, I compared supervisedmachine learning algorithms with semi-supervised learning (SSL) algorithmsfor text classification on several standard corpora. This comparison showsthat support vector machines normally are the best choice for text classification.I also find that traditional feature vectors consisting of counts of functionwords can be used for identifying the author of a translated document but themethod can be augmented by adding semantic information from documents.Overall the results presented in this thesis help clarify the role of reviewsin relation to understanding both users and different aspects of product use.

UR - https://soeg.kb.dk/permalink/45KBDK_KGL/fbp0ps/alma99122153181905763

M3 - Ph.D. thesis

BT - Usability and User Experience Information in Reviews

PB - Department of Computer Science, Faculty of Science, University of Copenhagen

ER -