Ranking and Selecting Multi-Hop Knowledge Paths to Better Predict Human Needs

Proceedings of NAACL-HLT 2019, pages 3671–3681 Minneapolis, Minnesota, June 2 - June 7, 2019. c⃝2019 Association for Computational Linguistics 3671 Ranking and Selecting Multi-Hop Knowledge Paths to Better Predict Human Needs Debjit Paul Research Training Group AIPHES Institute for Computational Linguistics Heidelberg University paul@cl.uni-heidelberg.de Anette Frank Research Training Group AIPHES Institute for Computational Linguistics Heidelberg University frank@cl.uni-heidelberg.de Abstract To make machines better understand senti- ments, research needs to move from polarity identification to understanding the reasons that underlie the expression of sentiment. Cate- gorizing the goals or needs of humans is one way to explain the expression of sentiment in text. Humans are good at understanding sit- uations described in natural language and can easily connect them to the character’s psycho- logical needs using commonsense knowledge. We present a novel method to extract, rank, fil- ter and select multi-hop relation paths from a commonsense knowledge resource to interpret the expression of sentiment in terms of their underlying human needs. We efficiently inte- grate the acquired knowledge paths in a neural model that interfaces context representations with knowledge using a gated attention mech- anism. We assess the model’s performance on a recently published dataset for categorizing human needs. Selectively integrating knowl- edge paths boosts performance and establishes a new state-of-the-art. Our model offers inter- pretability through the learned attention map over commonsense knowledge paths. Human evaluation highlights the relevance of the en- coded knowledge. 1 Introduction Sentiment analysis and emotion detection are es- sential tasks in human-computer interaction. Due to its broad practical applications, there has been rapid growth in the field of sentiment analysis (Zhang et al., 2018). Although state-of-the-art sentiment analysis can detect the polarity of text units (Hamilton et al., 2016; Socher et al., 2013), there has been limited work towards explaining the reasons for the expression of sentiment and emo- tions in texts (Li and Hovy, 2017). In our work, we aim to go beyond the detection of sentiment, toward explaining sentiments. Such explanations can range from detecting overtly expressed expla- nations or reasons for sentiments towards specific aspects of, e.g., products or films, as in user re- views to the explanation of the underlying reasons for emotional reactions of characters in a narrative story. The latter requires understanding of stories and modeling the mental state of characters. Re- cently, Ding and Riloff (2018) proposed to catego- rize affective events with categories based on hu- man needs, to provide explanations of people’s at- titudes towards such events. Given an expression such as I broke my leg, they categorize the reason for the expressed negative sentiment as being re- lated to a need concerning ‘health’. In this paper we focus on the Modelling Naive Psychology of Characters in Simple Common- sense Stories dataset of Rashkin et al. (2018), which contains annotations of a fully-specified chain of motivations and emotional reactions of characters for a collection of narrative stories. The stories are annotated with labels from multiple the- ories of psychology (Reiss, 2004; Maslow, 1943; Plutchik, 1980) to provide …

Content truncated to 500 words