Authors :
Sai Rushyanth Vattikonda
Volume/Issue :
Volume 8 - 2023, Issue 6 - June
Google Scholar :
https://bit.ly/3TmGbDi
Scribd :
https://tinyurl.com/43svz9yb
DOI :
https://doi.org/10.5281/zenodo.8059970
Abstract :
User-generated content, such as reviews and
comments, contains valuable information about products
as well as the opinions expressed by users. With the rise
of internet usage, there has been an influx of usergenerated data in the form of reviews and comments.
Individuals share their experiences, opinions, sentiments,
and emotions by writing reviews and comments about
products they have purchased online or their
impressions after engaging with various media, such as
movies or books. These user-generated data often
encompass emotions such as happiness, sadness, and
surprise, which can serve as crucial indicators for
recommending new items based on users' emotional
preferences.
In this study, we propose a method to extract
emotions from user-generated data by utilizing lexical
ontology, specifically WordNet, in conjunction with
insights from the field of psychology. These extracted
emotions can then be leveraged to enhance
recommendations. To evaluate the effectiveness of our
approach, we compare our emotion prediction model
with the traditional rating-based item similarity model,
as well as explore the impact of emotional fuzziness in
the feature space.
User-generated content, such as reviews and
comments, contains valuable information about products
as well as the opinions expressed by users. With the rise
of internet usage, there has been an influx of usergenerated data in the form of reviews and comments.
Individuals share their experiences, opinions, sentiments,
and emotions by writing reviews and comments about
products they have purchased online or their
impressions after engaging with various media, such as
movies or books. These user-generated data often
encompass emotions such as happiness, sadness, and
surprise, which can serve as crucial indicators for
recommending new items based on users' emotional
preferences.
In this study, we propose a method to extract
emotions from user-generated data by utilizing lexical
ontology, specifically WordNet, in conjunction with
insights from the field of psychology. These extracted
emotions can then be leveraged to enhance
recommendations. To evaluate the effectiveness of our
approach, we compare our emotion prediction model
with the traditional rating-based item similarity model,
as well as explore the impact of emotional fuzziness in
the feature space.
