No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Humor as circuits in semantic networks
Abstract
This work presents a first step to a general implementation of the Semantic-Script Theory of Humor (SSTH). Of the scarce amount of research in computational humor, no research had focused on humor generation beyond simple puns and punning riddles. We propose an algorithm for mining simple humorous scripts from a semantic network (Concept-Net) by specifically searching for dual scripts that jointly maximize overlap and incongruity metrics in line with Raskin's Semantic-Script Theory of Humor. Initial results show that a more relaxed constraint of this form is capable of generating humor of deeper semantic content than wordplay riddles. We evaluate the said metrics through a user-assessed quality of the generated two-liners.
... However, having new combinations does not necessarily result in funny jokes. Many humor generation systems (Valitutti et al. 2016;Stock and Strapparava 2003;Labutov and Lipson 2012), do not use audience responses to reformulate their products and to produce humorous utterances that are possible and reachable, valid, appreciated, and also of high quality. There is a fine line between humor and nonsense. ...
... Automatically generated two-liner jokes like (Labutov and Lipson 2012;Manurung et al. 2008) are a specific example of standalone texts that are are able to place the joke within a frame of reference. The set-up of the joke provides a context, and the punchline provides the element of surprise, creating thus the humorous effect. ...
Humor - as well as language in general - is by nature social and tied to a context. To better engage with context , computational humor could draw inspiration from the concept of intersubjectivity: the sharing of perspectives. This paper focuses on discussing the possible advantages of utilizing the concept of intersubjectivity to contextualize computational humor. Intersubjectivity in humor generation system design is discussed as a possible means of evaluation of the creative product, as well as a potential approach to generating more impressive humoristic content. Firstly, evaluation of computational humor has been wanting for more effective and versatile methods. To this problem, an implementation of sharing perspectives between the system and its users offers a viable solution. Secondly, approaches to humor generation are contrasted with interactive dialogue systems, to analyze how they contextualize humor. The comparisons show that well defined interactive design and evaluation methods that enable perspective sharing between the producer and the press would greatly benefit humor-generating systems. The final section theorizes on the possible foundations for modeling intersubjectivity in computational humor.
... The ConceptNet semantic network was exploited for the generation of verbal humour, including riddles (Labutov and Lipson, 2012). For this purpose, different, but overlapping, paths between the same two concepts are aligned with a surface template that maximises inter-path incongruity. ...
... Yet, features should be as representative as possible of the concept. For instance, asking for the hyponym of a word with too many hyponyms (e.g., person, plant, instrument) decreases solvability, an aspect considered by Labutov and Lipson (2012). So, the rank was penalised according to the number of other words related the same way as the concept is to the used features. ...
... According to the SSTH, there is an obvious script and an underlying script in a joke; these two scripts must be incongruous and also overlapping in meaning. The set-up of the joke elicits the obvious script, while the punchline of the joke switches the interpretation (i.e., the obvious script) to the underlying script (Labutov & Lipson, 2012;Raskin, 1985). That is, the perception of humour is largely dependent on the subversion of the previous interpretation. ...
As a research topic, humour has drawn much attention from multiple disciplines including linguistics. Based on Engelthaler & Hills’ (2018) humour scale, this study developed a measure named Humour Index (HMI) to quantify the degree of humour of texts. This measure was applied to examine the diachronic changes in the degree of humour of American newspapers and magazines across a time span of 118 years (1900-2017) with the use of texts from Corpus of Historical American English (COHA). Besides, the study also discussed the contributions of different types of words to the degree of humour in the two genres. The results show significant uptrends in the degree of humour of both newspapers and magazines in the examined period. Moreover, derogatory and offensive words are found to be less frequently used than other categories of words in both genres. This study provides both theoretical and methodological implications for humour studies and claims or hypotheses of previous research, such as infotainment and linguistic positivity bias.
... Although not all riddles have a humorous intent, they can be adopted as a template for jokes, especially when formulated as question-answer, with the latter working as a punchline. An example are punning riddles (Binsted & Ritchie, 1997), produced by rules that exploited a lexicon with syntactic and semantic information, or others generated by mining incongruous circuits in a common-sense knowledge base (Labutov & Lipson, 2012). Besides the latter, riddles have been automatically generated from a knowledge base of famous characters and their properties (Guerrero et al., 2015), or by mixing word categories from a thesaurus and associated modifiers (Galvan et al., 2016). ...
Creativity is an inherently human skill, and thus one of the goals of Artificial Intelligence. Specifically, linguistic computational creativity deals with the autonomous generation of linguistically-creative artefacts. Here, we present four types of text that can be tackled in this scope—poetry, humour, riddles, and headlines—and overview computational systems developed for their generation in Portuguese. Adopted approaches are described and illustrated with generated examples, and the key role of underlying computational linguistic resources is highlighted. The future of such systems is further discussed together with the exploration of neural approaches for text generation. While overviewing such systems, we hope to disseminate the area among the community of the computational processing of the Portuguese language.
... Computational linguistics tends to leverage neural systems, template-based systems, or a hybrid of the both for humor generation that rarely benefits from those theory-driven impulses. For example, Labutov and Lipson (2012) explored to mine simple humorous scripts from a semantic network (ConceptNet). They claimed that this may generate humor beyond simple puns and punning riddles Binsted and Ritchie (1997). ...
Language is the principal tool for human communication, in which humor is one of the most attractive parts. Producing natural language like humans using computers, a.k.a, Natural Language Generation (NLG), has been widely used for dialogue systems, chatbots, machine translation, as well as computer-aid creation e.g., idea generations, scriptwriting. However, the humor aspect of natural language is relatively under-investigated, especially in the age of pre-trained language models. In this work, we aim to preliminarily test whether NLG can generate humor as humans do. We build a new dataset consisting of numerous digitized Chinese Comical Crosstalk scripts (called C$^3$ in short), which is for a popular Chinese performing art called `Xiangsheng' since 1800s. (For convenience for non-Chinese speakers, we called `crosstalk' for `Xiangsheng' in this paper.) We benchmark various generation approaches including training-from-scratch Seq2seq, fine-tuned middle-scale PLMs, and large-scale PLMs (with and without fine-tuning). Moreover, we also conduct a human assessment, showing that 1) large-scale pretraining largely improves crosstalk generation quality; and 2) even the scripts generated from the best PLM is far from what we expect, with only 65% quality of human-created crosstalk. We conclude, humor generation could be largely improved using large-scaled PLMs, but it is still in its infancy. The data and benchmarking code is publicly available in \url{https://github.com/anonNo2/crosstalk-generation}.
... After we attempted the two methods, we concluded that the Retrieval-Based method is not feasible to generate self-mockery for chatbots due to the sparsity of the appropriate self-oriented humor candidate sentences. Moreover, we could not find available Natural Language Processing (NLP) methods that can generate usable humorous utterances for chatbots on a large scale [3,27,36] to fulfill the demand of the Retrieval-Based method. Therefore, we adopted the ...
... Previous studies have attempted to model this knowledge using ad-hoc manually created databases and labeled training examples. Examples of works on humor generation include dirty joke-telling robots (Sjöbergh & Araki, 2008), a model that generates two-liner jokes (Labutov & Lipson, 2012), and a model that generates punning riddles (Binsted & Ritchie, 1994). In the past, most works used supervision in some form; Sjöbergh and Araki (2008) used human jokes collected from various sources. ...
Although humor enriches human lives, some jokes fail to amuse people because of a lack of morality. In this paper, we propose a mechanism capable of selecting humor based on moral criteria. To this end, we first construct a model based on an N-gram corpus and generate joke candidates using various template patterns. We then employ a moral judgement classifier based on a recurrent neural network and utilize the trained model for humor selection. The experimental results obtained from best–worst scaling demonstrate that this scheme is able to generate jokes with moral category labels. We confirmed that jokes about the classifier categorized as Loyalty and Authority, which are regarded as good in our study, are funnier than jokes about Fairness, Purity, Harm, Cheating, and Degradation. Although we did not confirm that there was a difference in the funny level between good and bad moral jokes, the results demonstrate that moral categories of humor can affect the funny level.
... Meanwhile, the generation of humor is also one of the focuses of this study. Labutov and Lipson (2012) adopted a human rating method to evaluate the effectiveness of the generated humorous sentences based on SSTH theory. Valitutt and Toivonen (2013) generated the humor texts by replacing the words and proposed a human rating method to evaluate the humor rating. ...
Homographic pun has been developed into a new research area as an important branch of humor research, being a common source of humor in jokes and other comedic works. Pun word is the key to better understand homographic pun. However, in order to construct automatic model for locating the pun from homographic pun, it remains difficult challenges because of the ambiguity and confusion. In this paper, we firstly introduce several multi-dimensional semantic relationships of homographic pun based on the relevant theory and then employ a novel effective un-supervised semantic similarity match approach MSRLP that depending on the multi-dimensional semantic relationships to locate the pun in a homographic pun. Performance evaluation demonstrates that our presented approach significantly achieves the state-of-the-art performance on the public SemEval2017 Task7 dataset, outperforming a number of strong baselines by at least 3.67% in F1-score measure.
... Another cluster of work has considered the generation of humor, mostly via fixed templates such as acronyms , puns (Binsted and Ritchie 1997;Ritchie et al. 2007), two-liners (Labutov and Lipson 2012), or cross-reference ambiguity (Tinholt and Nijholt 2007). ...
Humor is an essential human trait. Efforts to understand humor have called out links between humor and the foundations of cognition, as well as the importance of humor in social engagement. As such, it is a promising and important subject of study, with relevance for artificial intelligence and human– computer interaction. Previous computational work on humor has mostly operated at a coarse level of granularity, e.g., predicting whether an entire sentence, paragraph, document, etc., is humorous. As a step toward deep understanding of humor, we seek fine-grained models of attributes that make a given text humorous. Starting from the observation that satirical news headlines tend to resemble serious news headlines, we build and analyze a corpus of satirical headlines paired with nearly identical but serious headlines. The corpus is constructed via Unfun.me, an online game that incentivizes players to make minimal edits to satirical headlines with the goal of making other players believe the results are serious headlines. The edit operations used to successfully remove humor pinpoint the words and concepts that play a key role in making the original, satirical headline funny. Our analysis reveals that the humor tends to reside toward the end of headlines, and primarily in noun phrases, and that most satirical headlines follow a certain logical pattern, which we term false analogy. Overall, this paper deepens our understanding of the syntactic and semantic structure of satirical news headlines and provides insights for building humor-producing systems.
... Another cluster of work has considered the generation of humor, mostly via fixed templates such as acronyms , puns (Binsted and Ritchie 1997;Ritchie et al. 2007), two-liners (Labutov and Lipson 2012), or cross-reference ambiguity (Tinholt and Nijholt 2007). ...
Humor is an essential human trait. Efforts to understand humor have called out links between humor and the foundations of cognition, as well as the importance of humor in social engagement. As such, it is a promising and important subject of study, with relevance for artificial intelligence and human-computer interaction. Previous computational work on humor has mostly operated at a coarse level of granularity, e.g., predicting whether an entire sentence, paragraph, document, etc., is humorous. As a step toward deep understanding of humor, we seek fine-grained models of attributes that make a given text humorous. Starting from the observation that satirical news headlines tend to resemble serious news headlines, we build and analyze a corpus of satirical headlines paired with nearly identical but serious headlines. The corpus is constructed via Unfun.me, an online game that incentivizes players to make minimal edits to satirical headlines with the goal of making other players believe the results are serious headlines. The edit operations used to successfully remove humor pinpoint the words and concepts that play a key role in making the original, satirical headline funny. Our analysis reveals that the humor tends to reside toward the end of headlines, and primarily in noun phrases, and that most satirical headlines follow a certain logical pattern, which we term false analogy. Overall, this paper deepens our understanding of the syntactic and semantic structure of satirical news headlines and provides insights for building humor-producing systems.
... More recently, several researchers have looked into automatic common-sense knowledge construction and expansion using common-sense inferences (Tandon et al., 2011;Bordes et al., 2011;Socher et al., 2013;Angeli and Manning, 2014). Several works have looked into combining NLP with commonsense (Gerber et al., 2010;Gordon et al., 2011;LoBue and Yates, 2011;Labutov and Lipson, 2012;Gordon et al., 2012). Most relevant to our work is a SemEval-2012 task (Gordon et al., 2012), looking into common-sense causality identification prediction. ...
Automatic satire detection is a subtle text classification task, for machines and at times, even for humans. In this paper we argue that satire detection should be approached using common-sense inferences, rather than traditional text classification methods. We present a highly structured latent variable model capturing the required inferences. The model abstracts over the specific entities appearing in the articles, grouping them into generalized categories, thus allowing the model to adapt to previously unseen situations.
... Current engineering efforts in the study of humor are primarily focused on humor generation, a task I argue is both smaller in scope and more prone to incomplete, mechanistic solutions than humor recognition. Several of the most effective of these methods still work by generating randomized jokes en masse for a human to choose from [22] [17], or tightly applying human-created templates [14]. ...
Humor is a creative, ubiquitous, and powerful communication strategy, yet it is currently challenging for computers to correctly identify instances of humor, let alone understand it. In this thesis I develop a computational model of humor based on error identification and resolution, as well as methods for understanding the mental trajectory required for successful humor appreciation. An infrastructure for constructing humor detectors based on this theory is implemented in the context of a general error handling and investigation system for the Genesis story-understanding system.
The computational model consists of a series of Experts that quantify important story elements such as allyship, harm to characters, character traits, karma, morbidity, contradiction, and unexpected events. Due to the homogeneous structure of their interactions, Experts using different methodologies such as simulation, Bayesian reasoning, neural nets, or symbolic reasoning can all interact, share findings of interest, and suggest reasons for each other's issues through this system.
This system of Experts can identify the resolvable narrative flaws that drive humor, therefore they are also able to discover unintentional problems within narratives. I have additionally demonstrated successful quantification of indicators of effective human engagement with narrative such as suspense, attention span length, attention density, and moments of insight. Variations in Expert parameters account for different senses of humor in individuals. This new scope of understanding allows Genesis to help authors search their narratives to determine if higher level narrative mechanics are well executed or not, a crucial role usually reserved for a human editor. By successfully demonstrating a framework for computational recognition and comprehension of humor, I have begun to show that computers are capable of sharing an ability previously considered an exclusively human quality.
... There has been considerable research on the topic of discovering unique, interesting or surprising documents. Researchers have studied different dimensions of this problem in terms of humor identification [12,4,10], text aesthetics [15,16,7], and document diversity [1,8]. [12] studies a computational approach for humor recognition by utilizing a set of humorspecific stylistic features such as alliteration, antonymy, and adult slangs. ...
User experiences can be made more engaging by incorporating surprise. For example, online shoppers may like to view unique products. In this paper we propose an approach for detecting surprising documents, such as product titles. As the concept of surprise is subjective, there is currently no principled method for measuring the surprisingness score of a document. We present such a method; an unsupervised approach for automatically discovering surprising documents in an unlabeled corpus. Our approach is based on a probabilistic model of surprise, and a construction of effective distributional word embeddings, which can be adapted to the semantic context in which the word appears. As the performance of our model does not degrade with the length of the document, it is particularly well suited for very short documents (even a single sentence). We evaluate our model both in supervised and unsupervised settings, demonstrating its state-of-the-art performance on two real-world data sets: a collection of e-commerce products from eBay, and a corpus of NSF proposals. These experiments show that our surprisingness score exhibits high correlation with human annotated labels.
... Early work on computational humour focused more on humour generation in specific contexts, such as punning riddles (Binsted and Ritchie, 1994;Ritchie et al., 2007), humorous acronyms (Stock and Strapparava, 2003), or jokes in the form of "I like my X like I like my Y" (Petrovic and Matthews, 2013). Labutov and Lipson (2012) offered a slightly more generalized approach using Semantic Script Theory of Humour. ...
... Both Computational Humor and Text Visualization as fields have seen extensive activity lately, but tend to work on separate topics. Computational Humor deals a lot with the modeling and detection of incongruities within text and many attempts have recently been made attempting to detect or generate jokes using computers (Labutov and Lipson 2012;Mihalcea et al. 2010;Petrovic and Matthews 2013;Ritchie 2003;Taylor and Mazlack 2007;Valitutti et al. 2013;Taylor and Raskin 2012), but no attempt focused on visualization has been made. On the other hand, studies on Text Visualization tend to focus on other topics such as identifying the central topic within a text. ...
This chapter presents a visual text mining approach to modeling humor within text. It includes algorithms for visualizing and discovering shifts in text interpretation as intelligent agents parse meaning from garden path jokes. Three successful text visualization methods are described to identify discrimination features for humorous and non-humorous texts. These visualization methods include Collocated Paired Coordinates, Heat maps, and two-dimensional Boolean plots.
... Similarly, the Sematic Script Theory of Humor (SSTH) says that a joke emerges when it can be interpreted according to two different, generic scripts, one of which is less obvious (Attardo and Raskin, 1991). Labutov and Lipson make a first step at exploiting the SSTH theory to automatically generate two-line jokes (Labutov and Lipson, 2012). ...
... Compared to humor recognition, humor generation has received quite a lot attention in the past decades (Stock and Strapparava, 2005;Ritchie, 2005;Hong and Ong, 2009). Most generation work draws on humor theories to account for humor factors, such as the Script-based Semantic Theory of Humor (Raskin, 1985;Labutov and Lipson, 2012) and employs templates to generate jokes. For example, Ozbal and Strapparava (2012) created humorous neologism using WordNet and ConceptNet. ...
The introductory chapter reviews the literature on humour theories. While the traditional theories of humour attend to the competence of humour, this study examines the performance of humour in political cartoons. Identifying a lacuna in the field of visual humour, the study also furthers a theoretical model for perceiving the dynamic movement of visual humour in political cartoons. The methodology followed in the study is primarily that of a constructivist interpretative framework. The study sets the theoretical framework through the selection of Jakobson, Barthes, Lakoff and Johnson. The model is designed using these theoreticians as they become the connecting thread in the fields of meaning construction.
Humor is essential to establish more natural and enjoyable human–computer interactions, and researchers have been working on developing a way to automatically generate humor. This study aims to explore a better way to automatically generate Japanese common humor “soramimi”, and understand how the humor occurs. Soramimi is a type of parody song in which the original lyrics are replaced by different words that have similar pronunciations. Although a previous study proposed an algorithm to replace input text with homophonic soramimi text, the mechanism of soramimi humor is still unclear. Based on the incongruity-resolution model, we hypothesized that phonological similarity between the parody and the original lyrics enhances humor in soramimi. A subjective experiment was conducted in which the phonological similarity and humor of fifteen soramimi parody lyrics were evaluated. The results indicated that the phonological similarity of soramimi was positively correlated with its humorousness. Exploring other factors that affect humorousness and the development of an automatic generation system for soramimi lyrics based on the identified factors are topics for our future research.
The goal of this paper is to investigate the use of visualization as an approach to modeling humor within text. In particular, we developed algorithmic and automated approaches to visualizing and detecting shifts in interpretation as intelligent agents parse meaning from garden path jokes. Garden path jokes can occur when a reader’s initial interpretation of an ambiguous text turns out to be incorrect, leading them down the wrong path to a semantic dead end. Given new information, semantic incongruities arise that require resolution, often triggering a humorous response. This is a work of visual text mining, that is visualizing texts in order to detect patterns and features associated with various text based phenomena such as humor. In this paper we describe three successful approaches to text visualization conducive to identifying distinguishing features given humorous and non humorous texts. These are the use of paired collocated coordinates, heat maps, and two-dimensional Boolean plots. The proposed methodology and tools offer a new approach to testing and generating hypotheses related to theories of humor as well as other phenomena involving incongruity-resolution and shifts in interpretation including non-verbal humor.
With AI celebrating its 60 th anniversary, questions arise of when (not even if) a computational system will be able to understand humor. These questions open up interesting opportunities, but point out areas of research that yet are insufficient for informal human computer communication. This paper looks at computational humor as a way of verifying computational understanding of text (written or verbal). In particular, we treat ontology as a knowledge representation mechanism and natural language as a vehicle delivering this knowledge. A true ontology should provide a world model for the described domain, identifying its main concepts and tying them together with all relevant contentful properties. The question is how to get this model from text accurately? Assuming, as we do, that there is an accurate and unambiguous way of getting explicitly stated information from text, a lot of information is, in fact, implicit and yet crucial to the world model that we are creating. This implicit information has to be made explicit at the reasoning stage if we hope to come up with the results similar to human reasoning or understanding. In this paper, we will look at various ways, requiring optimal human-computer hybrid collaboration, in which ontology helps text understanding for humor processing, and text helps with dynamic ontology development. We hypothesize that such communication will be helpful for interaction with any computational system in a human-friendly way in general, and for robots in particular.
Humor generation is a very hard problem. It is difficult to say exactly what makes a joke funny, and solving this problem algorithmically is assumed to require deep semantic understanding, as well as cultural and other contextual cues. We depart from previous work that tries to model this knowledge using ad-hoc manually created databases and labeled training examples. Instead we present a model that uses large amounts of unannotated data to generate I like my X like I like my Y, Z jokes, where X, Y, and Z are variables to be filled in. This is, to the best of our knowledge, the first fully unsupervised humor generation system. Our model significantly outperforms a competitive baseline and generates funny jokes 16% of the time, compared to 33% for human-generated jokes.
This paper presents an open domain monologue generation method for non-task-oriented dialogue systems to be able to speak their opinions and impressions as a speaker. To generate monologues, we acquire suitable utterances that contain a given topic from Twitter. Our method determines whether utterances have cohesion or not using the support vector machine and concatenate them in a row. It scores the utterance sequences from the aspect of their humor, unexpectedness and speciality in the given topic. We acquire the utterance sequences that ranks high as monologues. Results of an experiment demonstrate that our method can generate amusing and semantically appropriate monologues.
In this paper, we present our work of humor recognition on Twitter, which will facilitate affect and sentimental analysis in the social network. The central question of what makes a tweet (Twitter post) humorous drives us to design humor-related features, which are derived from influential humor theories, linguistic norms, and affective dimensions. Using machine learning techniques, we are able to recognize humorous tweets with high accuracy and F-measure. More importantly, we single out features that contribute to distinguishing non-humorous tweets from humorous tweets, and humorous tweets from other short humorous texts (non-tweets). This proves that humorous tweets possess discernible characteristics that are neither found in plain tweets nor in humorous non-tweets. We believe our novel findings will inform and inspire the burgeoning field of computational humor research in the social media.
The present study derives hypotheses from the General Theory of Verbal Humor (GTVH) and tests them on a sample of 534 subjects. Subjects are presented with three sets of jokes, each consisting of an anchor joke and comparison jokes in which variations in one and only one of the six
Knowledge Resources (KR), script Opposition (SO), logical mechanism (LM), Situation (SI) target (TA), narrative strategy (NS), andlanguage (LA) occurred. Subjects rated the degree of similarity between the anchor joke and the six comparison jokes. The results support the hypothesis that the extent to which the similarity judgment is affected depends on the type of the KR manipulated. Also, there generally is a decreasing trend in similarity between the KRs LA and SO. Whereas there was a significant difference between all consecutive KRs, äs predicted by the hierarchy postulated by the GTVH, SI and LM were not in the right order. Possible explanations for thisfact are discussed.
This article presents and discusses a study which empirically Supports some of the Claims of the General Theory of Verbal Humor (GTVH). After introducing the theory, the article will present the hypotheses derived from the theory that were tested and finally the results of the investigation.
We report on the design of a system that enables the user to perform retrieval of items of interest both by browsing an automatically constructed very large hypertext and by spreading activation on the hypertext network itself. It is this second option that is particularly interesting since it enables the user to retrieve items that have not been visited, but that are similar to items flagged as relevant during the browsing. The spreading of activation from relevant items to similar items is achieved using a form of "constrained spreading activation". This is a kind of spreading activation controlled by heuristic rules that limit and direct the activation towards the most promising links and node of the hypertext.
The past 15 years has seen the development of a number of programs which perform tasks in the area of humour, but these have been exploratory research prototypes, usually on a very small scale, and none of them interacted with users. Amongst those which actually created humorous texts, the JAPE program was probably the most substantial, but even it was far from being useful for any practical pur-pose. We have developed a fully engineered riddle genera-tor, inspired by the ideas in the JAPE system, which uses a large-scale multimedia lexicon and a set of symbolic rules to generate jokes. It has an interactive user interface, spe-cially designed for children with complex communication needs (CCN), so that users can make choices to guide the riddle generator. The software is robust, stable, and re-sponds sufficiently promptly that naive users can interact without difficulty. It has been tested over with real users (children with CCN), with highly positive results, and is publicly available for free download.
Recognizing the intent of a written utterance is a difficult task. To understand the intended meanings, it may be neces-sary to recognize all possible meanings of the utterance; and, then choose the most appropriate one for the situation. The choice may be made using heuristics. A method for recogni-tion of a humorous intent in a text is proposed. Computational recognition of a humorous intent can be broken down into two parts: recognition of a humorous text, and recognition of the intent to be humorous. To narrow the focus, we propose to recognize the humorous intent of short dialogs. A dialog will be considered humorous if the first part of the text can have two meanings; and, one of the meanings conflicts with the meaning of the punchline. The intent of the text is considered humorous if the schema related to the non-conflicting mean-ing of the setup has not been activated in the preceding text.
In a corpus of jokes, a human might judge two documents to be the "same joke" even if characters, locations, and other details are varied. A given joke could be retold with an entirely different vocabulary, while still maintaining its identity. Since most retrieval systems consider documents to be related only when their word content is similar, we propose joke retrieval as a domain where standard language models may fail. In particular, we consider the task of identifying the "same joke" to be a necessary component of any joke retrieval system, and we examine it in both ranking and classification settings. We exploit the structure of jokes to develop two domain-specific alternatives to the "bag of words" document model. In one, only the punch lines, or final sentences, are compared; in the second, certain categories of words (e.g., professions and countries) are marked up and treated as interchangeable. Each technique works well for certain jokes. By combining the methods using machine learning, we create a hybrid that achieves higher performance than any individual approach.
Computational humor is a subdiscipline of computational linguistics with applications in human computer interfaces, edutainment, affective computing, intelligent agents, and other areas. Based on ontological semantics, we develop the resources and algorithms the computer needs to understand and produce humor, in principle and on a detailed example. Our ultimate output will not be that of a toy system that fills in templates, previously the only available approach, but rather true natural language generation, based on the computer approximation of the human understanding of the joke. This paper shows how ontological semantics provides for and computes the full computer understanding of humor, a sine qua non of humor generation.
In this paper we present our continuing eorts to generate narrative using a character-centric approach. In particular we discuss the advantages of explicitly representing the emergent event sequence in order to be able to exert influence on it and generate stories that 'retell' the emergent narrative. Based on a narrative distinction between fabula, plot and presentation, we make a first step by presenting a model based on story comprehension that can capture the fabula, and show how it can be used for the automatic creation of stories.
If computers are ever going to communicate naturally and effectively with humans, they must be able to use humor. Moreover, humor provides insight into how humans process real, complex, creative language. By modeling humor generation and understanding on computers, we can gain a better picture of how the human brain handles not just humor but language and cognition in general. This installment of Trends & Controversies focuses on different aspects and applications of humor.
Computationally recognizing humor is an aspect of natural language understanding. Although there appears to be no complete computational model for recognizing verbal humor, it may be possible to recognize jokes based on statistical language recognition techniques. Computational humor recognition was investigated. A restricted set of all possible jokes that have wordplay as a component was considered. The limited domain of "knock knock" jokes was examined. A created wordplay generator produces an utterance that is similar in pronunciation to a given word, and the wordplay recognizer determines if the utterance is valid. Once a possible wordplay is discovered, a joke recognizer determines if a found wordplay transforms the text into a joke.
INTRODUCTION Humor is a multi-disciplinary field of research. People have been working on humor in many fields of research, such as psychology, philosophy and linguistics, sociology and literature. Especially in the context of computer science (or Artificial Intelligence) humor research aims at modeling humor in a computationally tractable way. Having computational models of humor allows interface designers to have the computer generate and interpret humor when interacting with users. Being able to recognize a user's frustration [5] can be useful; however, preventing it whenever possible by an adequate use of humor can be useful as well. There are many situations in human-human interaction where humor plays an important role. It does not only play an entertaining role. It helps in regulating conversations (social and business conversations), building trust between partners and facilitating self-disclosure and it is an important factor in social attraction. Nass has shown in differen
Riddles based on simple puns can be classified according to the patterns of word, syllable or phrase similarity they depend upon. We have devised a formal model of the semantic and syntactic regularities underlying some of the simpler types of punning riddle. We have also implemented this preliminary theory in a computer program which can generate riddles from a lexicon containing general data about words and phrases; that is, the lexicon content is not customised to produce jokes. Informal evaluation of the program's results by a set of human judges suggest that the riddles produced by this program are of comparable quality to those in general circulation among school children. 1 Introduction One very common form of humour is the question-answer joke, or riddle. A significant subset, probably the vast majority, of these jokes are based on some form of pun (questionanswer puns make up almost a third of the riddles in [Web78]). For example: What do you get when you cross a sh...
This thesis describes a formal model of a subtype of humour, and the implementation of that model in a program that generates jokes of that subtype. Although there is a great deal of literature on humour in general, very little formal work has been done on puns, and none has been implemented. All current linguistic theories of humour are over-general and not falsifiable. Our model, whichisspecific, formal, implemented and evaluated, makes a significantcontribution to the field. Punning riddles are our chosen subtype of verbal humour, for several reasons. They are very common, they exhibit certain regular structures and mechanisms, and they have been studied previously by linguists. Our model is based on our extensive analysis of large numbers of punning riddles, taken from children's jokebooks. The implementation of the model, J A P E (Joke Analysis and Production Engine), generates punning riddles, from a humour independent lexicon. Pun generation requires much less world knowle...
Today millions of web-users express their opinions about many topics through blogs, wikis, fora, chats and social networks. For sectors such as e-commerce and e-tourism, it is very useful to automatically analyze the huge amount of social information available on the Web, but the extremely unstructured nature of these contents makes it a difficult task. SenticNet is a publicly available resource for opinion mining built exploiting AI and Semantic Web techniques. It uses dimensionality reduction to infer the polarity of common sense concepts and hence provide a public resource for mining opinions from natural language text at a semantic, rather than just syntactic, level. Copyright © 2010, Association for the Advancement of Artificial Intelligence. All rights reserved.
Most techniques that calculate the relatedness between two concepts use a semantic network, such as Wikipedia, WordNet, or ConceptNet, to find the shortest intermediate pathway between two nodes. These techniques assume that a low number of edges on the shortest pathway indicates conceptual similarity. Although this technique has proven valid in conforming to psychological data, we test the usefulness of additional pathway variables in ConceptNet, such as edge type and user-rated score. Our results show strong evidence for the application of additional pathway variables in calculating semantic similarity.
HAHAcronym has been the first project concerned with computational humor sponsored by the European Commission. The project was meant to establish the potential of computational humor through the demonstration of a working prototype and an assessment of the state of the art and of scenarios where humor can add something to existing information technologies. The main goal of HAHAcronym was the realization of an acronym ironic re-analyzer and generator as a proof of concept in a focalized but non-restricted context. In order to implement it, some general tools have been developed or adapted for the humorous context. For all tools, particular attention has been put on reusability. A fundamental tool is an incongruity detector/generator to detect/generate semantic mismatches between the known/expected "sentence" meaning and other interpretations, along some specific dimension.
The article proposes a general theory of verbal humor, focusing on verbal jokes as its most representative subset. The theory is an extension and revision of Raskin's script-based semantic theory of humor and of Attardo's five-level joke representation model. After distinguishing the parameters of the various degrees of similarity among the joke examples, six knowledge resources informing the joke, namely script oppositions, logical mechanisms, situations, targets, narrative strategies, and language, are put forward. A hierarchical organization for the six knowledge resources is then discovered on the basis of the asymmetrical binary relations, of the proposed and modified content / tool dichotomy, and, especially, of the hypothesized perceptions of the relative degrees of similarity. It is also argued that the emerging joke representation model is neutral to the process of joke production. The proposed hierarchy enables the concepts of joke variants and invariants, introduced previously by Attardo, to be firmed up, generalized, and augmented into a full-fledged taxonomy indexed with regard to the shared knowledge resource values (for example, two jokes may be variants on, that is, sharing, the same script oppositions and logical mechanisms). The resulting general theory of verbal humor is discussed in the light of its relations with various academic disciplines and areas of research as well as with the script-based semantic theory of humor, special theories of humor, and incongruity-based theories.
In this paper we present the methodology for Word Sense Disambiguation based on domain information. Domain is a set of words in which there is a strong semantic relation among the words. The words in the sentence contribute to determine the domain of the sentence. The availability of WordNet domains makes the domain-oriented text analysis possible. The domain of the target word can be fixed based on the domains of the content words in the local context. This approach can be effectively used to disambiguate nouns. We present the unsupervised approach to Word Sense Disambiguation using the WordNet domains. The model determines the domain of the target word and the sense corresponding to this domain is taken as the correct sense. We have used the WordNet domains 3.1.as lexical database.
This paper deals with computational detection of humor. It assumes that computational humor is a useful task for any number of reasons and in many applications. It discusses the computational linguistic/ semantic precondi¬tions for computational humor and an ontological semantic approach to the task of humor detection, based on direct and comprehensive access to meaning rather than on trying to guess it through statistical-cum-syntactical keyword methods. The pa¬per is informed by the experience of designing and imple¬menting a humor detection model, whose decent success rate confirmed some of the assumptions while its misses made other ideas prominent, including the necessity of full text comprehension. The bulk of the paper explains how the comprehensive representation of meaning and, most impor¬tantly how unstructured natural language text is automati¬cally translated into the ontologically defined text meaning representations that can be used then to de¬tect humor in them, if any, automatically.
In a world in which millions of people express their feelings and opinions about any issue in blogs, wikis, fora, chats and
social networks, the distillation of knowledge from this huge amount of unstructured information is a challenging task. In
this work we build a knowledge base which merges common sense and affective knowledge and visualize it in a multi-dimensional
vector space, which we call SenticSpace. In particular we blend ConceptNet and WordNet-Affect and use dimensionality reduction
on the resulting knowledge base to build a 24-dimensional vector space in which different vectors represent different ways
of making binary distinctions among concepts and sentiments.
Humor is one of the most interesting and puzzling aspects of human behavior. Despite the attention it has received in fields such as philosophy, linguistics, and psychology, there have been only few attempts to create computational models for humor recognition or generation. In this article, we bring empirical evidence that computational approaches can be successfully applied to the task of humor recognition. Through experiments performed on very large data sets, we show that automatic classification techniques can be effectively used to distinguish between humorous and non-humorous texts, with significant improvements observed over a priori known baselines.
Browsing is the foremost method in searching through information in a hypertext or hypermedia system. However, as the number of nodes and links increases, this technique is far from satisfactory, and other search mechanisms must be provided. Classical search techniques such as menu selection hierarchies, string matching, Boolean query, etc., are already available, but they treat nodes as independent entities rather than considering the link semantics between nodes. Moreover, in order to write a query the users often encounter many problems such as how to find the appropriate terms that describe the information needs, how to correctly write a query in a language using artificial syntax, etc. This paper describes an alternative based on a Bayesian network that structures the indexing terms and stores the user's information needs. In our approach, the user does not have to write a formal query because the computation required is accomplished automatically and without any prior information or constraint. Moreover, using a constrained spreading activation, our solution uses link semantics to search relevant starting points for browsing.
Many networks of interest in the sciences, including social networks, computer networks, and metabolic and regulatory networks, are found to divide naturally into communities or modules. The problem of detecting and characterizing this community structure is one of the outstanding issues in the study of networked systems. One highly effective approach is the optimization of the quality function known as “modularity” over the possible divisions of a network. Here I show that the modularity can be expressed in terms of the eigenvectors of a characteristic matrix for the network, which I call the modularity matrix, and that this expression leads to a spectral algorithm for community detection that returns results of demonstrably higher quality than competing methods in shorter running times. I illustrate the method with applications to several published network data sets.
• clustering
• partitioning
• modules
• metabolic network
• social network
We describe ConceptNet, a freely available semantic network presently consisting of over 250,000 elements of commonsense knowledge. Inspired by Cyc, ConceptNet includes a wide range of commonsense concepts and relations, and inspired by WordNet, it is structured as a simple, easy-to-use semantic network. ConceptNet supports many of the same applications as WordNet, such as query expansion and determining semantic similarity, but it also allows simple temporal, spatial, affective, and several other types of inferences. This paper is structured as follows. We first discuss how ConceptNet was built and the nature and structure of its contents. We then present the ConceptNet toolkit, a reasoning system designed to support textual reasoning tasks by providing facilities for spreading activation, analogy, and path-finding between concepts. Third, we provide some quantitative and qualitative analyses of ConceptNet. We conclude by describing some ways we are currently exploring to improve ConceptNet.
Humour is a valid subject for research in artificial intelligence, as it is one of the more complex of human behaviours. Although philosophers and others have discussed humour for centuries, it is only very recently that computational work has begun in this field, so the state of the art is still rather basic. Much of the research has concentrated on humour expressed verbally, and there has been some emphasis on models based on "incongruity". Actual implementations have involved puns of very limited forms. It is not clear that computerised jokes could enhance user interfaces in the near future, but there is a role for computer modelling in testing symbolic accounts of the structure of humorous texts. A major problem is the need for a humour-processing program to have knowledge of the world, and reasoning abilities.
- S Attardo
- V Raskin
S. Attardo and V. Raskin. 1991. Script theory revis (it)
ed: Joke similarity and joke representation model. Humor: International Journal of Humor Research; Humor: International Journal of Humor Research.
- K Binsted
- A Nijholt
- O Stock
- C Strapparava
- G Ritchie
- R Manurung
- H Pain
- A Waller
- D O'mara
K. Binsted, A. Nijholt, O. Stock, C. Strapparava,
G. Ritchie, R. Manurung, H. Pain, A. Waller, and
D. O'Mara. 2006. Computational humor. Intelligent
Systems, IEEE, 21(2):59-69.
Modularity and community structure in networks
- M E J Newman
M.E.J. Newman. 2006. Modularity and community
structure in networks. Proceedings of the National
Academy of Sciences, 103(23):8577-8582.
A practical application of computational humour
- G Ritchie
- R Manurung
- H Pain
- A Waller
- R Black
- D Omara
G. Ritchie, R. Manurung, H. Pain, A. Waller, R. Black,
and D. OMara. 2007. A practical application of computational humour. In Proceedings of the 4th. International Joint Workshop on Computational Creativity,
London, UK.
Toward computational recognition of humorous intent
- J Taylor
- L Mazlack
J. Taylor and L. Mazlack. 2005. Toward computational
recognition of humorous intent. In Proceedings of
Cognitive Science Conference, pages 2166-2171.