The Practical Value Aspect of Semantically Structured Information Environments: Investigating the implied logical reasoning potential of semantic layers through the design of persuasive recommender systems for data-driven businesses
Translated title
Det Praktiske Værdiaspekt af Semantisk Strukturerede Informationsområder
Authors
Poulsen, Nicklas Thorup ; Overgaard, Joachim August
Term
4. Term
Education
Publication year
2021
Submitted on
2021-06-01
Pages
177
Abstract
Mange datadrevne virksomheder bruger analyser, men overser muligheden for at give data en fælles, tydelig betydning—et semantisk lag—fordi værdien ikke er åbenlys. Denne afhandling undersøger værdien af at tilføje et semantisk lag til forretningsstrukturer, og hvordan det kan styrke anerkendelsen af denne del af informationsarkitekturen i professionelle miljøer. For at gøre værdien konkret præsenteres et koncept: en designproces for et overtalende anbefalingssystem. Konceptet kombinerer en formel ontologi (en maskinlæsbar model af centrale begreber og deres relationer) med principper for overtalende kommunikation i et målrettet informationsmiljø. Målet er at give de mest logiske indholdsanbefalinger til hver type bruger i et domæne og formidle dem på en mere overtalende og risikobevidst måde. Tilsammen viser disse elementer det praktiske potentiale i semantiske anbefalinger og hvordan målrettet information kan give nye, logiske indsigter.
Many data-driven businesses use analytics but overlook the chance to give their data a shared, explicit meaning—a semantic layer—because its benefits are not obvious. This thesis examines the value of adding such a semantic layer to business structures and how it can raise the profile of this part of information architecture in professional practice. To make the value tangible, it presents a concept: a design process for a persuasive recommender system. The concept combines a formal ontology (a machine-readable model of key concepts and their relationships) with principles of persuasive communication, applied in a targeted information environment. The goal is to produce the most logical content recommendations for each type of domain user and to present them in a more persuasive and risk-aware way. Together, these elements demonstrate the practical potential of semantic recommendation and show how targeted information can yield new, logical insights.
[This summary has been rewritten with the help of AI based on the project's original abstract]
Keywords
Documents