Word counts
|
|
Toggle faculties
Toggle degrees
Toggle languages
|
Character counts
|
|
Toggle faculties
Toggle degrees
Toggle languages
|
Page counts
|
|
Toggle faculties
Toggle degrees
Toggle languages
|
Distinct words
|
|
Toggle faculties
Toggle degrees
Toggle languages
|
Faculties, thesis types and languages
|
|
Toggle faculties
Toggle degrees
Toggle languages
|
|
|
|
|
Background aobut the Thesis network
The Thesis Network has been collected out of public material from the HELDA database of theLibrary of the University of Helsinki. The context analysis of the network has been performed for those theses, where an abstract in English could be found. This material consists of 265 theses in total, which covers about 57 % of those conferred in the Conferment Ceremony of the Faculty of Philosophy in 2017.
Based on the data from thes abstracts, an analysis on keywords and their relative weights was performed using the term frequency-inverse document frequency (tl-idf) method. Even though the method is not necessarily optimal for short datasets[1], it was still enough to produce reasonable results with our material and use of more advanced algorithms can be left for further studies. We also used the Porter stemmer algorithm to combine the different variations of the same words, which in turn unifies the language in the material. A value for closeness between the theses was defined as the sum of the weights of the mutual tf-idf keywords. The values for closeness were filtered to cover the strongest 8.6 % out of all the possible connections. A vast majority of the 70 000 calculated connections are relatively weak and presenting this amount of data in a web browser is fairly slow.
The dots (nodes) of the network reflect the theses of the conferred and the bows (edges) between them the connections. The colors of the nodes show the (current-day) faculty faculties of each, their sizes the type of thesis and the weight of the bow the closeness of the theses.
The Thesis Network is able to show most of the faculties as their own clusters, with the facutlies of Biological and Environmental Sciences and Pharmacy being a notable exception by merging into one. The Faculty of Science and the bio-pharmaceutical amalgation is mostly separated from the faculties of Arts and Educational Sciences, even though there are conntector-theses also bridging between these faculties.
In the interactive network, the nodes can be hovered or clicked with a mouse to inspect the connections of a single these and to see more information in the left side bar. This infucrmation includes the abstract of the thesis. It is also possible to perform searches by the names of the authors.
Source: [1] https://helda.helsinki.fi/handle/10138/37924