Aspect-based opinion summarization of product reviews on social media |
Asgeir Ogmundarson
|
| Abstract | With the increase of opinions shared by people on social media the importance of algorithms capable of analyzing them for sentiment to produce value has grown. All companies face hidden costs in the form of lost revenue due to unhappy customers not returning and potential customers snubbing a product due to bad publicity. This thesis develops an aspect-based opinion summarization system capable of extracting opinion features from product reviews, determining the sentiment of the reviews the opinion features appeared in and producing a summary of the most frequent features discussed with regards to sentiment. The results from the summary system are promising but could still be improved to produce even more value. |
| Type | Master's thesis [Academic thesis] |
| Year | 2015 |
| Publisher | Technical University of Denmark, Department of Applied Mathematics and Computer Science |
| Address | Richard Petersens Plads, Building 324, DK-2800 Kgs. Lyngby, Denmark, compute@compute.dtu.dk |
| Series | DTU Compute M.Sc.-2015 |
| Note | DTU supervisor: Ole Winther, olwi@dtu.dk, DTU Compute |
| Electronic version(s) | [pdf] |
| Publication link | http://www.compute.dtu.dk/English.aspx |
| BibTeX data | [bibtex] |
| IMM Group(s) | Intelligent Signal Processing |