Using Word Associations for Humour Recognition

PhD Thesis Proposal Defence


Title: "Using Word Associations for Humour Recognition"

by

Mr. Andrew CATTLE


Abstract:

As natural language interfaces become more prevalent, the ability for 
computers to both understand and create humour becomes more important. 
Humour is a ubiquitous part of human communication. It can be used to make 
one’s self more likeable, to defuse a tense situation, or just for pure 
entertainment. As modern digital virtual assistants such as Alexa, 
Cortana, Google Now, and Siri become more human-like, the ability to 
effectively recognize, interpret, and even produce humour becomes more 
important.

What makes humour such an exciting challenge is that it requires not only 
linguistic dexterity but also world/domain knowledge. Syntax, phonology, 
and semantics all play a role in making a joke funny. However, existing 
humour recognition works have typically taken a fairly basic view of joke 
semantics, treating jokes as unordered bags-of-words and computing word 
embedding similarities between all word pairs. This bears little 
resemblance to the way humans actually interpret humour.

In this proposal we motivate the use of semantic relatedness based on word 
associa- tions for humour recognition. Furthermore, we present evidence 
that word associations outperform Word2Vec similarity on both humour 
classification and humour ranking tasks across several datasets. Word 
associations’ focus on relatedness over similarity offers an increased 
flexibility and the ability to capture weaker, more tangential 
relationships between concepts. Word associations also better represent 
the way humans store their mental lexicons.

We introduce two methods for extracting word association features. The 
first is a graph-based method which is efficient to calculate but suffers 
from coverage issues. The second is a sophisticated word association 
strength prediction model capable of predicting association strengths 
between arbitrary word pairs. In addition to documenting the creating of 
this prediction model we also evaluate its performance both on an 
association prediction task and on a humour classification task.


Date:			Thursday, 3 May 2018

Time:                  	3:00pm - 5:00pm

Venue:                  Room 4472
                         (lifts 25/26)

Committee Members:	Dr. Xiaojuan Ma (Supervisor)
  			Dr. Yangqiu Song (Chairperson)
 			Prof. Fangzhen Lin
 			Prof. Dit-Yan Yeung


**** ALL are Welcome ****