Interactivity, Game Creation, Design, Learning, and Innovation. 5th International Conference, ArtsIT 2016, and First International Conference, DLI 2016, Esbjerg, Denmark, May 2–3, 2016, Proceedings

Research Article

Analysing Emotional Sentiment in People’s YouTube Channel Comments

Download
202 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-55834-9_21,
        author={Eleanor Mulholland and Paul Mc Kevitt and Tom Lunney and Karl-Michael Schneider},
        title={Analysing Emotional Sentiment in People’s YouTube Channel Comments},
        proceedings={Interactivity, Game Creation, Design, Learning, and Innovation. 5th International Conference, ArtsIT 2016, and First International Conference, DLI 2016, Esbjerg, Denmark, May 2--3, 2016, Proceedings},
        proceedings_a={ARTSIT \& DLI},
        year={2017},
        month={3},
        keywords={360-MAM-Affect 360-MAM-Select Affective computing Brit Lab EmoSenticNet Gamification Google YouTube API Head Squeeze Machine learning Natural language processing Recommender system Sentiment analysis YouTube YouTube EDU},
        doi={10.1007/978-3-319-55834-9_21}
    }
    
  • Eleanor Mulholland
    Paul Mc Kevitt
    Tom Lunney
    Karl-Michael Schneider
    Year: 2017
    Analysing Emotional Sentiment in People’s YouTube Channel Comments
    ARTSIT & DLI
    Springer
    DOI: 10.1007/978-3-319-55834-9_21
Eleanor Mulholland1,*, Paul Mc Kevitt1,*, Tom Lunney1,*, Karl-Michael Schneider2,*
  • 1: Ulster University
  • 2: Google Ireland Ltd.
*Contact email: mulholland-e9@email.ulster.ac.uk, p.mckevitt@ulster.ac.uk, tf.lunney@ulster.ac.uk, karlmicha@gmail.com

Abstract

Online recommender systems are useful for media asset management where they select the best content from a set of media assets. We are developing a recommender system called 360-MAM-Select for educational video content. 360-MAM-Select utilises sentiment analysis, emotion modeling and gamification techniques applied to people’s comments on videos, for the recommendation of media assets. Here, we discuss the architecture of 360-MAM-Select, including its sentiment analysis module, 360-MAM-Affect and gamification module, 360-Gamify. 360-MAM-Affect is implemented with the YouTube API [9], GATE [5] for natural language processing, EmoSenticNet [8] for identifying emotion words and RapidMiner [20] to count the average frequency of emotion words identified. 360-MAM-Affect is tested by tagging comments on the YouTube channels, Brit Lab/Head Squeeze [3], YouTube EDU [28], Sam Pepper [22] and MyTop100Videos [18] with EmoSenticNet [8] in order to identify emotional sentiment. Our results show that and are the most frequent emotions across all the YouTube channel comments. Future work includes further implementation and testing of 360-MAM-Select deploying the Unifying Framework [25] and Emotion-Imbued Choice (EIC) model [13] within 360-MAM-Affect for emotion modelling, by collecting emotion feedback and sentiment from users when they interact with media content. Future work also includes implementation of the gamification module, 360-Gamify, in order to check its suitability for improving user participation with the Octalysis gamification framework [4].