Building a Sentiment Lexicon for Swedish
Paper i proceeding, 2016

In this paper we will present our ongoing project to build and evaluate a sentiment lexicon for Swedish. Our main resource is SALDO, a lexical resource of modern Swedish developed at Språkbanken, University of Gothenburg. Using a semi-supervised approach, we expand a manually chosen set of six core words using parent-child relations based on the semantic network structure of SALDO. At its current stage the lexicon consists of 175 seeds, 633 children, and 1319 grandchildren.


Bianka Nusko

Göteborgs universitet

Nina Tahmasebi

Göteborgs universitet

Olof Mogren

Chalmers, Data- och informationsteknik, Datavetenskap

Linköping Electronic Conference Proceedings

1650-3740 (eISSN)

Vol. 126 006 32--37-


Språkteknologi (språkvetenskaplig databehandling)