Last modified: 30.06.2017
Please Introduce yourself...
I’m Wafa, the only female student researcher in TANT-lab, and I’m working on a pet-project that involves a comparative study of social tagging (user generated keywords) and automatic tagging (automatic extraction of keywords) within a practical manner. The idea for the project originates from my former project, where I tried to analyse political recommendations and keywords written and tagged by 425 citizens from eight countries in Europe as a part of an EU-project called ASSET. During that time, I struggled to make sense out of the user generated keywords due to the idiosyncratic use of words, misspellings, the use of too many keywords, poor keywords etc. At that point, I decided to do a manual analytical tagging of the user generated keywords in order to produce a more legible visualisation of them, which was rather difficult and time consuming. I also noticed that the user generated keywords that I manually tagged, were partly similar to the automatic extracted keywords that CorTexT (a software platform) extracted from the political recommendations, which made me question the usability and the entire purpose of incorporating user generated keywords in such projects. Today, I am still considering whether I would recommend others, or myself in that case, to educate the users on how to comprehend and be better at tagging, to tag it yourself or leave it to the different automatic tagging algorithms. I’m investigating this further through a practical study of the differences and similarities of social tagging and automatic tagging, and their value in projects that mainly focuses on the content rather than the linguistic usage. The goal of this pet-project is therefore to gain a wider understanding of 1) how different automatic tagging algorithms extract terms, 2) the use of different tagging approaches and what they can provide you with, and 3) to share my experiences with others/you.
What is your field of research?
The difference between user-generated terms and automatic extracted terms
A comparative study of social tagging (user generated keywords) and automatic tagging (automatic extraction of keywords) within a practical manner.