No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
A Neural Approach to Pun Generation
... Only a few humor-generation systems utilized neural-based solutions (Luo et al., 2019;Yu et al., 2018;Ren and Yang, 2017). None of them tackled a questionanswering scenario or involved a zero-shot setting with prompt engineering, as we present here. ...
... To avoid distortion originated by a system having only a handful of responses, we also measure diversity. We use the measures proposed by Yu et al. (2018) to compare the different methods' diversity (see Table 20). Results show that the GLM approach is more diverse than the templates one. ...
... Diversity is classified according to (Amin and Burghardt, 2020). Automated diversity is measured using the ratio of distinct unigrams and bigrams divided by the total number of words in the whole set, according to (Yu et al., 2018). N = 1000 for all methods except for humans (N = 33). ...
... Previous AI dialogue system studies concerning humor have been carried out on various disciplines [5], and the findings showed that it was challenging to model humor from the structure and semantic understanding of Natural Language Processing [5]. Until now, the humor studies have focused on humor generation and recognition instead of humor response [6,7]. ...
... Previous AI dialogue system studies concerning humor have been carried out on various disciplines [5], and the findings showed that it was challenging to model humor from the structure and semantic understanding of Natural Language Processing [5]. Until now, the humor studies have focused on humor generation and recognition instead of humor response [6,7]. ...
... Cross-culture dimensions have a non-negligible impact on the performance of empathetic response [9] and humorous interactions [5] in AI. Users from various cultural backgrounds express their emotions and humor tend to confirm to their social homogeneity, internal sanctioning system, and cultural norms [3,5]. ...
... Due to the lack of sizeable training data, existing approaches are heavy-weighted in order to not rely on pun sentences for training. For example, (Yu et al., 2018) † Work done when the author is interning at UCLA. 1 We focus on generating homographic puns where two or more meanings of a word form an intended humorous effect. The sentence is ungrammatical. ...
... Neural Pun Yu et al. (2018) propose the first neural approach to homographic puns based on a constrained beam search algorithm to jointly decode the two distinct senses of the same word. ...
... ;Yu et al. (2018) to calculate distinct unigram and bigrams as the diversity(Li et al., 2016) in terms of sentence level and corpus level. We also Gen AMBIPUN My calculator is unjust and ::::: illogic. ...
... Humor generation. Humor generation aims to generate humorous texts, including question-answer jokes (Hong and Ong 2009;Labutov and Lipson 2012) and narrative jokes (Sjöbergh and Araki 2009;Yu, Tan, and Wan 2018). Main approaches are divided into template-based and neural model-based methods. ...
... Neural model-based methods adopt neural networks, such as sequence-to-sequence models, language models, to generate humorous texts. For example, Li, Liu, and Wang (2022) and Yu, Tan, and Wan (2018) use neural approaches for pun generation. Although these methods can generate high level of creative output, the generated text may not be an appropriate humorous sentence from the perspective of human evaluation. ...
Humor is a crucial part of human communication. Understanding humor and generating humorous responses in dialogue can provide natural and empathic human-computer interactions. However, most existing pre-trained language models (PLMs) perform unsatisfactorily in humor generation. On one hand, the serious shortage of humor corpus and datasets pose challenges for constructing models that can understand and generate humorous expressions. On the other hand, humor generation relies on rich knowledge and commonsense, which is often tacit and unspoken. In this paper, we construct the largest Chinese Explainable Humor Response Dataset to date with chain-of-humor and humor mind map annotations, which can be used to comprehensively evaluate as well as improve the humorous response ability of PLMs. We further design humor-related auxiliary tasks to further enhance PLMs' humorous response performance. Extensive evaluations demonstrate that our proposed dataset and auxiliary tasks effectively help PLMs to generate humorous responses, laying the groundwork for future humor research.
... Another approach explored for humor generation is through lexical replacement on input texts, for instance the methods by Valitutti et al. (2016) and He et al. (2019). Lastly the most recent systems use some modern Natural Language Generation (NLG) techniques to create punning jokes, for example via Neural Networks (Yu et al., 2018). ...
... This system is especially challenging, as it relies on some specific resources for phonological, lexical, and semantic analyses that should have their counterparts in Portuguese. Other techniques are the ones by Yu et al. (2018), a recurrent neural network to create homographic puns, and He et al. (2019), which is based on probabilistic models of surprise to edit input texts to create punning humor. ...
... Although He et al. (2019) propose a retrieve-andedit approach to incorporate surprise, their error analysis shows that the proposed retrieval methods are often unsuccessful. Moreover, existing works on pun generation are split up in terms of generating homographic puns, wherein the same written word has two or more meanings (Mittal et al., 2022;Yu et al., 2020Yu et al., , 2018, and homophonic puns, where two words that sound similar have different meanings (Luo et al., 2019;He et al., 2019;Hashimoto et al., 2018). There lacks a unified generation framework for both types of puns. ...
... Existing works on pun generation often rely on naive intuitions of semantic ambivalence. For example, Yu et al. (2018) and Luo et al. (2019) promote the ambivalence of the pun word via a constrained language model and reinforcement learning; others find related words to support semantic ambiguity (Yu et al., 2020;Mittal et al., 2022). However, these systems lack serious theoretical backbones and therefore none could evaluate their generated results with regards to the proposed intuitions. ...
... the research community (He et al., 2019;Yu et al., 2018Yu et al., , 2020Mittal et al., 2022;Horri, 2011). As one of the most important ways to communicate humor (Abbas and Dhiaa, 2016), puns can help relieve anxiety, avoid painful feelings and facilitate learning (Buxman, 2008). ...
... Specifically, in heterographic pun generation, systems generate puns using a pair of homophones involving a pun word and an alternative word (He et al., 2019;Yu et al., 2020;Mittal et al., 2022). Alternatively, in homographic pun generation, systems generate puns that must support both given senses of a single polysemous word (Yu et al., 2018;Luo et al., 2019;. Despite the great progress that has been made under such experimental settings, real-world applications for pun gener- If you sight a whale, it could be a fluke. ...
Previous work on pun generation commonly begins with a given pun word (a pair of homophones for heterographic pun generation and a polyseme for homographic pun generation) and seeks to generate an appropriate pun. While this may enable efficient pun generation, we believe that a pun is most entertaining if it fits appropriately within a given context, e.g., a given situation or dialogue. In this work, we propose a new task, context-situated pun generation, where a specific context represented by a set of keywords is provided, and the task is to first identify suitable pun words that are appropriate for the context, then generate puns based on the context keywords and the identified pun words. We collect CUP (Context-sitUated Pun), containing 4.5k tuples of context words and pun pairs. Based on the new data and setup, we propose a pipeline system for context-situated pun generation, including a pun word retrieval module that identifies suitable pun words for a given context, and a generation module that generates puns from context keywords and pun words. Human evaluation shows that 69% of our top retrieved pun words can be used to generate context-situated puns, and our generation module yields successful puns 31% of the time given a plausible tuple of context words and pun pair, almost tripling the yield of a state-of-the-art pun generation model. With an end-to-end evaluation, our pipeline system with the top-1 retrieved pun pair for a given context can generate successful puns 40% of the time, better than all other modeling variations but 32% lower than the human success rate. This highlights the difficulty of the task, and encourages more research in this direction.
... Although He et al. (2019) propose a retrieve-andedit approach to incorporate surprise, their error analysis shows that the proposed retrieval methods are often unsuccessful. Moreover, existing works on pun generation are split up in terms of generating homographic puns, wherein the same written word has two or more meanings (Mittal et al., 2022;Yu et al., 2020Yu et al., , 2018, and homophonic puns, where two words that sound similar have different meanings (Luo et al., 2019;He et al., 2019;Hashimoto et al., 2018). There lacks a unified generation framework for both types of puns. ...
... Existing works on pun generation often rely on naive intuitions of semantic ambivalence. For example, Yu et al. (2018) and Luo et al. (2019) promote the ambivalence of the pun word via a constrained language model and reinforcement learning; others find related words to support semantic ambiguity (Yu et al., 2020;Mittal et al., 2022). However, these systems lack serious theoretical backbones and therefore none could evaluate their generated results with regards to the proposed intuitions. ...
We propose a unified framework to generate both homophonic and homographic puns to resolve the split-up in existing works. Specifically, we incorporate three linguistic attributes of puns to the language models: ambiguity, distinctiveness, and surprise. Our framework consists of three parts: 1) a context words/phrases selector to promote the aforementioned attributes, 2) a generation model trained on non-pun sentences to incorporate the context words/phrases into the generation output, and 3) a label predictor that learns the structure of puns which is used to steer the generation model at inference time. Evaluation results on both pun types demonstrate the efficacy of our model over strong baselines.
... One of the only other notable examples of neural-based humour generation comes from Yu et al. [93], who focus specifically on pun generation [4]. To do this, the authors aimed to maximise incongruity, training a neural network using a Seq2Seq model and Wikipedia data. ...
... To do this, the authors aimed to maximise incongruity, training a neural network using a Seq2Seq model and Wikipedia data. The model is given as input a polysemic word (a word with multiple possible meanings), and two of its definitions [93]. The model is then used to generate two sentences using this word -one for each of the two meanings provided as input. ...
In recent years there has been substantial growth in the capabilities of systems designed to generate text that mimics the fluency and coherence of human language. From this, there has been considerable research aimed at examining the potential uses of these natural language generators (NLG) towards a wide number of tasks. The increasing capabilities of powerful text generators to mimic human writing convincingly raises the potential for deception and other forms of dangerous misuse. As these systems improve, and it becomes ever harder to distinguish between human-written and machine-generated text, malicious actors could leverage these powerful NLG systems to a wide variety of ends, including the creation of fake news and misinformation, the generation of fake online product reviews, or via chatbots as means of convincing users to divulge private information. In this paper, we provide an overview of the NLG field via the identification and examination of 119 survey-like papers focused on NLG research. From these identified papers, we outline a proposed high-level taxonomy of the central concepts that constitute NLG, including the methods used to develop generalised NLG systems, the means by which these systems are evaluated, and the popular NLG tasks and subtasks that exist. In turn, we provide an overview and discussion of each of these items with respect to current research and offer an examination of the potential roles of NLG in deception and detection systems to counteract these threats. Moreover, we discuss the broader challenges of NLG, including the risks of bias that are often exhibited by existing text generation systems. This work offers a broad overview of the field of NLG with respect to its potential for misuse, aiming to provide a high-level understanding of this rapidly developing area of research.
... Due to the lack of sizeable training data, existing approaches are heavy-weighted in order to not rely on pun sentences for training. For example, (Yu et al., 2018) train a constrained neural language model (Mou et al., 2015) from a general text corpus and then use a joint decoding algorithm to promote puns. He et al. † Work done when the author is interning at UCLA. 1 We focus on generating homographic puns where two or more meanings of a word form an intended humorous effect. ...
... Neural Pun Yu et al. (2018) propose the first neural approach to homographic puns based on a constrained beam search algorithm to jointly decode the two distinct senses of the same word. ...
In this paper, we propose a simple yet effective way to generate pun sentences that does not require any training on existing puns. Our approach is inspired by humor theories that ambiguity comes from the context rather than the pun word itself. Given a pair of definitions of a pun word, our model first produces a list of related concepts through a reverse dictionary. We then utilize one-shot GPT3 to generate context words and then generate puns incorporating context words from both concepts. Human evaluation shows that our method successfully generates pun 52\% of the time, outperforming well-crafted baselines and the state-of-the-art models by a large margin.
... After early research (Ritchie, 2005) pointed out ambiguity as a key in pun generation, numerous studies have investigated automatic pun generation regarding heterographic puns, which slackens the surface form identity requirement for each meaning of the pun (He et al., 2019;Yu et al., 2020;Mittal et al., 2022). Other research explored homographic pun generation which is based on multiple meanings of a polysemous word (Yu et al., 2018;Luo et al., 2019;Tian et al., 2022). Recently, Sun et al. (Sun et al., 2022) extended the pun generation problem to consider contextual cues. ...
... Automatic Evaluation To measure the model's ability to incorporate pun word pairs in the final generation, we utilize the structure success rate (Structure Succ.) as our primary automatic evaluation metric, following Sun et al. (2022a,b). To evaluate diversity, we follow Luo et al. (2019); Yu et al. (2018); Mittal et al. (2022) to calculate distinct unigrams (Dist-1) and bigrams (Dist-2) in terms of sentence level and corpus level. Average sentence length (Avg-Length) is also reported. ...
... Conversely, neural models promise more originality. For example, works by Li et al. [33] and Yu et al. [57] use such models for pun creation. However, humor isn't solely textual; images play a role, which our research emphasizes. ...
Humor, deeply rooted in societal meanings and cultural details, poses a unique challenge for machines. While advances have been made in natural language processing, real-world humor often thrives in a multi-modal context, encapsulated distinctively by memes. This paper poses a particular emphasis on the impact of multi-images on meme captioning. After that, we introduce the \textsc{XMeCap} framework, a novel approach that adopts supervised fine-tuning and reinforcement learning based on an innovative reward model, which factors in both global and local similarities between visuals and text. Our results, benchmarked against contemporary models, manifest a marked improvement in caption generation for both single-image and multi-image memes, as well as different meme categories. \textsc{XMeCap} achieves an average evaluation score of 75.85 for single-image memes and 66.32 for multi-image memes, outperforming the best baseline by 3.71\% and 4.82\%, respectively. This research not only establishes a new frontier in meme-related studies but also underscores the potential of machines in understanding and generating humor in a multi-modal setting.
... As the popularity of pre-trained language models (PLMs), some recent work, such as Yu et al. [41] and Rodriguez et al. [32], make preliminary attempts to finetune PLMs for humor recognition and generation. Thanks to the powerful understanding and generation capabilities that PLMs have learned from massive amounts of data, they significantly reduce human cost and enable the recognition (or generation) on more types of humorous expressions. ...
Humor understanding is an important and challenging research in natural language processing. As the popularity of pre-trained language models (PLMs), some recent work makes preliminary attempts to adopt PLMs for humor recognition and generation. However, these simple attempts do not substantially answer the question: {\em whether PLMs are capable of humor understanding?} This paper is the first work that systematically investigates the humor understanding ability of PLMs. For this purpose, a comprehensive framework with three evaluation steps and four evaluation tasks is designed. We also construct a comprehensive Chinese humor dataset, which can fully meet all the data requirements of the proposed evaluation framework. Our empirical study on the Chinese humor dataset yields some valuable observations, which are of great guiding value for future optimization of PLMs in humor understanding and generation.
... There are two main research lines in the field, namely humour generation and humour recognition. Early works in computational humour predominantly focused on humour generation, i.e., using natural language processing tools to imitate human humour and automatically generate humorous texts such as jokes, puns, etc. (Binsted & Ritchie, 1994;Hong & Ong, 2009;Yu et al., 2018). Humour recognition deals with the detection of humour in different linguistic units (Meaney et al., 2021;Mihalcea & Pulman, 2007;Taylor & Mazlack, 2004;Yang et al., 2015). ...
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.
... Specifically, in heterographic pun generation, systems generate puns using a pair of homophones involving a pun word and an alternative word (He et al., 2019;Yu et al., 2020;Mittal et al., 2022). Alternatively, in homographic pun generation, systems generate puns that must support both given senses of a single polysemous word (Yu et al., 2018;Luo et al., 2019;. Despite the great progress that has been made under such experimental settings, real-world applications for pun generation (e.g., in dialogue systems or creative slogan Table 1: Two examples each of heterographic puns and homographic puns in the SemEval 2017 Task 7 dataset. ...
... Phonetically constrained language generation is a primary subarea of computational creativity in natural language generation (NLG), primarily encompassing lyric and poetry generation (Tian and Peng, 2022;Wöckener et al., 2021;Xue et al., 2021;Zhang et al., 2020a;Agarwal and Kann, 2020), as well as pun generation (Sun et al., 2022;He et al., 2019;Yu et al., 2018), and continues to prove challenging for myriad reasons. Primarily, such works require the inclusion of phonetic factors such as metre and rhyme, which involves careful consideration of candidate vocabulary on the syllable level, leading to a reduced pool of allowable vocabulary once these constraints are in place. ...
... The lexicon-based approach uses linguistic rules and a database of words that have been labeled by researchers to identify the overall text subjectivity or sense of humor. This approach is simpler to implement, but not as precise as the deep learning approach [22]. The deep learning approach, specifically Convolutional Neural Networks (CNN), is able to understand the complexities of humor by generating an abstract representation of what it has learned, rather than just analyzing the text at a surface level. ...
Artificial intelligence (AI) technologies have been utilized in the education industry for enhancing student’s performance by generating spontaneous, timely, and personalized query response. One such technology is a dialogue system which is capable of generating humorous and empathetic responses for enhancing students’ learning outcomes. There is, however, limited research on the combination of humor, empathy, and culture in education. Thus, this paper proposes a dialogue system that is based on Wasserstein’s Generative Adversarial Network (WGAN) for generating responses with humor, empathy, and cultural sensitivity. The dialogue system has the ability to generate responses that take into account both coarse-grained emotions at the conversation level and fine-grained emotions at the token level, allowing for a nuanced understanding of a student’s emotional state. It can utilize external knowledge and prior context to enhance the ability of AI dialogue systems to comprehend emotions in a multimodal context. It can also analyze large corpora of text and other data, providing valuable insights into cultural context, semantic properties, and language variations. The dialogue system is a promising AI technology that can improve learning outcomes in various academic fields by generating responses with humor, empathy, and cultural sensitivity. In our study, the dialogue system achieved an accuracy rate of 94.12%, 93.83% and 92.60% in humor, empathy and culture models, respectively.
... Phonetically constrained language generation is a primary subarea of computational creativity in natural language generation (NLG), primarily encompassing lyric and poetry generation (Tian and Peng, 2022;Wöckener et al., 2021;Xue et al., 2021;Zhang et al., 2020a;Agarwal and Kann, 2020), as well as pun generation (Sun et al., 2022;He et al., 2019;Yu et al., 2018), and continues to prove challenging for myriad reasons. Primarily, such works require the inclusion of phonetic factors such as metre and rhyme, which involves careful consideration of candidate vocabulary on the syllable level, leading to a reduced pool of allowable vocabulary once these constraints are in place. ...
Previous work in phonetically-grounded language generation has mainly focused on domains such as lyrics and poetry. In this paper, we present work on the generation of tongue twisters - a form of language that is required to be phonetically conditioned to maximise sound overlap, whilst maintaining semantic consistency with an input topic, and still being grammatically correct. We present \textbf{TwistList}, a large annotated dataset of tongue twisters, consisting of 2.1K+ human-authored examples. We additionally present several benchmark systems (referred to as TwisterMisters) for the proposed task of tongue twister generation, including models that both do and do not require training on in-domain data. We present the results of automatic and human evaluation to demonstrate the performance of existing mainstream pre-trained models in this task with limited (or no) task specific training and data, and no explicit phonetic knowledge. We find that the task of tongue twister generation is challenging for models under these conditions, yet some models are still capable of generating acceptable examples of this language type.
... A similar conclusion was drawn by Amin and Burghardt (2020). Focusing on the broader task of humor generation, they analyzed possible evaluation approaches: human ratings on a Likert scale for humorousness, human ratings on a Likert scale for the likeness that a humorous text was written by a human-the soft Turing test as in Yu, Tan, and Wan (2018)-and "humorous frequency" as the proportion of funny instances out of a set of generated texts. All of them failed to present a criterion to evaluate humor in text objectively. ...
Humans are naturally endowed with the ability to write in a particular style. They can, for instance, rephrase a formal letter in an informal way, convey a literal message with the use of figures of speech or edit a novel by mimicking the style of some well-known authors. Automating this form of creativity constitutes the goal of style transfer. As a natural language generation task, style transfer aims at rewriting existing texts, and specifically, it creates paraphrases that exhibit some desired stylistic attributes. From a practical perspective, it envisions beneficial applications, like chatbots that modulate their communicative style to appear empathetic, or systems that automatically simplify technical articles for a non-expert audience.
Several style-aware paraphrasing methods have attempted to tackle style transfer. A handful of surveys give a methodological overview of the field, but they do not support researchers to focus on specific styles. With this paper, we aim at providing a comprehensive discussion of the styles that have received attention in the transfer task. We organize them in a hierarchy, highlighting the challenges for the definition of each of them and pointing out gaps in the current research landscape. The hierarchy comprises two main groups. One encompasses styles that people modulate arbitrarily, along the lines of registers and genres. The other group corresponds to unintentionally expressed styles, due to an author’s personal characteristics. Hence, our review shows how these groups relate to one another and where specific styles, including some that have not yet been explored, belong in the hierarchy. Moreover, we summarize the methods employed for different stylistic families, hinting researchers towards those that would be the most fitting for future research.
... Figurative Language Generation As a figure of speech, hyperbole generation is related to the general task of figurative language generation. Previous studies have tackled the generation of metaphor (Yu and Wan, 2019;Stowe et al., 2020;Chakrabarty et al., 2021;Stowe et al., 2021), simile (Chakrabarty et al., 2020bZhang et al., 2021), idiom (Zhou et al., 2021), pun (Yu et al., 2018;Luo et al., 2019b;He et al., 2019;sarcasm (Chakrabarty et al., 2020a). HypoGen (Tian et al., 2021) is a concurrent work with ours on hyperbole generation. ...
... One system [38] used lexical constraints to generate adult humour by substituting one word in a pre-existing text; another study [20] trained a system to automatically extract humorous templates which were then used for pun generation. Some current efforts to tackle this difficult problem more generally using neural approaches have been hindered by the lack of a sizable pun corpus [44]. ...
... The results however, were not satisfactory. Another such attempt was made by using a conditioned neural language model that was trained using general text corpus which was then passed through a decoding algorithm (Yu et al., 2018). Though these neural network techniques gave creative outputs, they lacked humor in comparison to template-based approaches which showed fairly good humor despite lacking creativity. ...
... Yu and Wan (2019) generate metaphor unsupervisedly by extracting the metaphorically-used verbs; propose a metaphor generation method with symbolism and discriminative decoding; Stowe et al. (2021) study diverse metaphor generation using conceptual mapping. Given a pair of homophones, Yu et al. (2018) train a conditional neural language model with an decoding algorithm for pun generation; He et al. (2019) tackle the same task with a local-global surprisal principle and a retrieve-and-edit pipeline; Luo et al. (2019) on the other hand propose an adversarial pun generative network. ...
... Focusing on the broader task of humor generation, Amin and Burghardt (2020) analyzed possible evaluation approaches. Human ratings on a Likertscale for humorousness, human ratings on a Likert-scale for the likeness that a humorous text was written by a human -the soft Turing test as in Yu et al. (2018) -and "humorous frequency" as the proportion of funny instances out of a set of generated texts: all of them failed to present a criterion to objectively evaluate humor in text. ...
Humans are naturally endowed with the ability to write in a particular style. They can, for instance, rephrase a formal letter in an informal way, convey a literal message with the use of figures of speech, edit a novel mimicking the style of some well-known authors. Automating this form of creativity constitutes the goal of style transfer. As a natural language generation task, style transfer aims at re-writing existing texts, and specifically, it creates paraphrases that exhibit some desired stylistic attributes. From a practical perspective, it envisions beneficial applications, like chat-bots that modulate their communicative style to appear empathetic, or systems that automatically simplify technical articles for a non-expert audience. Style transfer has been dedicated several style-aware paraphrasing methods. A handful of surveys give a methodological overview of the field, but they do not support researchers to focus on specific styles. With this paper, we aim at providing a comprehensive discussion of the styles that have received attention in the transfer task. We organize them into a hierarchy, highlighting the challenges for the definition of each of them, and pointing out gaps in the current research landscape. The hierarchy comprises two main groups. One encompasses styles that people modulate arbitrarily, along the lines of registers and genres. The other group corresponds to unintentionally expressed styles, due to an author's personal characteristics. Hence, our review shows how the groups relate to one another, and where specific styles, including some that have never been explored, belong in the hierarchy. Moreover, we summarize the methods employed for different stylistic families, hinting researchers towards those that would be the most fitting for future research.
... Pun generation with a neural model language model is one of the most recent efforts on humor generation (Yu, Tan, and Wan 2018). Their approach consists of training a conditional language model an using a beam search to find sentences that can support two polysemous meanings for a given word. ...
Automated news generation has become a major interest for new agencies in the past. Oftentimes headlines for such automatically generated news articles are unimaginative as they have been generated with ready-made templates. We present a computationally creative approach for headline generation that can generate humorous versions of existing headlines. We evaluate our system with human judges and compare the results to human authored humorous titles. The headlines produced by the system are considered funny 36\% of the time by human evaluators.
... Figurative Language Generation As a figure of speech, hyperbole generation is related to general figurative language generation. Previous studies have tackled the generation of metaphor Stowe, Ribeiro, and Gurevych 2020;Chakrabarty et al. 2021;Stowe et al. 2021), simile (Chakrabarty, Muresan, and Peng 2020;Zhang et al. 2021), idiom (Zhou et al. 2021), pun (Yu, Tan, and Wan 2018;Luo et al. 2019b;He, Peng, and Liang 2019;Yu, Zang, and Wan 2020), sarcasm (Chakrabarty et al. 2020b), and irony (Zhu, Yu, and Wan 2019). Several figurative language datasets are proposed alongside these works. ...
Despite being a common figure of speech, hyperbole is under-researched with only a few studies addressing its identification task. In this paper, we introduce a new task of hyperbole generation to transfer a literal sentence into its hyperbolic paraphrase. To tackle the lack of available hyperbolic sentences, we construct HYPO-XL, the first large-scale hyperbole corpus containing 17,862 hyperbolic sentences in a non-trivial way. Based on our corpus, we propose an unsupervised method for hyperbole generation with no need for parallel literal-hyperbole pairs. During training, we fine-tune BART to infill masked hyperbolic spans of sentences from HYPO-XL. During inference, we mask part of an input literal sentence and over-generate multiple possible hyperbolic versions. Then a BERT-based ranker selects the best candidate by hyperbolicity and paraphrase quality. Human evaluation results show that our model is capable of generating hyperbolic paraphrase sentences and outperforms several baseline systems.
... Yu and Wan (2019) generate metaphor unsupervisedly by extracting the metaphorically-used verbs; propose a metaphor generation method with symbolism and discriminative decoding; Stowe et al. (2021) study diverse metaphor generation using conceptual mapping. Given a pair of homophones, Yu et al. (2018) train a conditional neural language model with an decoding algorithm for pun generation; He et al. (2019) tackle the same task with a local-global surprisal principle and a retrieve-and-edit pipeline; Luo et al. (2019) on the other hand propose an adversarial pun generative network. ...
A hyperbole is an intentional and creative exaggeration not to be taken literally. Despite its ubiquity in daily life, the computational explorations of hyperboles are scarce. In this paper, we tackle the under-explored and challenging task: sentence-level hyperbole generation. We start with a representative syntactic pattern for intensification and systematically study the semantic (commonsense and counterfactual) relationships between each component in such hyperboles. Next, we leverage the COMeT and reverse COMeT models to do commonsense and counterfactual inference. We then generate multiple hyperbole candidates based on our findings from the pattern, and train neural classifiers to rank and select high-quality hyperboles. Automatic and human evaluations show that our generation method is able to generate hyperboles creatively with high success rate and intensity scores.
... Pun generation with a neural model language model is one of the most recent efforts on humor generation (Yu, Tan, and Wan 2018). Their approach consists of training a conditional language model an using a beam search to find sentences that can support two polysemous meanings for a given word. ...
Automated news generation has become a major interest for new agencies in the past. Oftentimes headlines for such automatically generated news articles are unimaginative as they have been generated with ready-made templates. We present a computationally creative approach for headline generation that can generate humorous versions of existing headlines. We evaluate our system with human judges and compare the results to human authored humorous titles. The headlines produced by the system are considered funny 36% of the time by human evaluators.
Creativity is the ability to produce novel, useful, and surprising ideas, and has been widely studied as a crucial aspect of human cognition. Machine creativity on the other hand has been a long-standing challenge. With the rise of advanced generative AI, there has been renewed interest and debate regarding AI's creative capabilities. Therefore, it is imperative to revisit the state of creativity in AI and identify key progresses and remaining challenges. In this work, we survey leading works studying the creative capabilities of AI systems, focusing on creative problem-solving, linguistic, artistic, and scientific creativity. Our review suggests that while the latest AI models are largely capable of producing linguistically and artistically creative outputs such as poems, images, and musical pieces, they struggle with tasks that require creative problem-solving, abstract thinking and compositionality and their generations suffer from a lack of diversity, originality, long-range incoherence and hallucinations. We also discuss key questions concerning copyright and authorship issues with generative models. Furthermore, we highlight the need for a comprehensive evaluation of creativity that is process-driven and considers several dimensions of creativity. Finally, we propose future research directions to improve the creativity of AI outputs, drawing inspiration from cognitive science and psychology.
Figurative language generation (FLG) is the task of reformulating a given text to include a desired figure of speech, such as a hyperbole, a simile, and several others, while still being faithful to the original context. This is a fundamental, yet challenging task in Natural Language Processing (NLP), which has recently received increased attention due to the promising performance brought by pre-trained language models. Our survey provides a systematic overview of the development of FLG, mostly in English, starting with the description of some common figures of speech, their corresponding generation tasks and datasets. We then focus on various modelling approaches and assessment strategies, leading us to discussing some challenges in this field, and suggesting some potential directions for future research. To the best of our knowledge, this is the first survey that summarizes the progress of FLG including the most recent development in NLP. We also organize corresponding resources, e.g., paper lists and datasets, and make them accessible in an open repository. We hope this survey can help researchers in NLP and related fields to easily track the academic frontier, providing them with a landscape and a roadmap of this area.
Language learning relies on language sense. To write a good essay, the author not only needs skills but also original accumulation. In this paper, we approach from two perspectives of essay scoring and essay writing, utilizing beautiful sentence evaluation and generation to assist in essay scoring and writing. We first established a beautiful sentence dataset and trained a beautiful sentence evaluation model using the dataset, and then applied the evaluation model to automatic Chinese essay scoring. The experimental results demonstrated the effectiveness of beautiful sentence features in the essay scoring tasks. Moreover, we trained a beautiful sentence generator using the pre-trained model GPT-2 to improve writing ability of essay writers. The experimental results demonstrated the effectiveness of the beautiful sentence generator in essay writing assistance.KeywordsBeautiful sentence generationBeautiful sentence evaluationAutomatic Chinese essay scoring
The rapid advancement in machine learning has led to a surge in automatic data generation, making it increasingly challenging to differentiate between naturally or human-generated data and machine-generated data. Despite these advancements, the generation of creative data remains a challenge. This paper aims to investigate and comprehend the essence of creativity, both in general and within the context of natural language generation. We review various approaches to creative writing devices and tasks, with a specific focus on the generation of poetry. We aim to shed light on the challenges and opportunities in the field of creative data generation.
The rapid advancement in machine learning has led to a surge in automatic data generation, making it increasingly challenging to differentiate between naturally or human-generated data and machine-generated data. Despite these advancements, the generation of creative data remains a challenge. This paper aims to investigate and comprehend the essence of creativity, both in general and within the context of natural language generation. We review various approaches to creative writing devices and tasks, with a specific focus on the generation of poetry. We aim to shed light on the challenges and opportunities in the field of creative data generation.
AI assistants are gradually becoming embedded in our lives, utilized for everyday tasks like shopping or music. In addition to the everyday utilization of AI assistants, many users engage them with playful shopping requests, gauging their ability to understand – or simply seeking amusement. However, these requests are often not being responded to in the same playful manner, causing dissatisfaction and even trust issues.In this work, we focus on equipping AI assistants with the ability to respond in a playful manner to irrational shopping requests. We first evaluate several neural generation models, which lead to unsuitable results – showing that this task is non-trivial. We devise a simple, yet effective, solution, that utilizes a knowledge graph to generate template-based responses grounded with commonsense. While the commonsense-aware solution is slightly less diverse than the generative models, it provides better responses to playful requests. This emphasizes the gap in commonsense exhibited by neural language models.
In this study, we address the problem of generating funny headlines for news articles. Funny headlines are beneficial even for serious news stories – they attract and entertain the reader. Automatically generated funny headlines can serve as prompts for news editors. More generally, humor generation can be applied to other domains, e.g. conversational systems. Like previous approaches, our methods are based on lexical substitutions. We consider two techniques for generating substitute words: one based on BERT and another based on collocation strength and semantic distance. At the final stage, a humor classifier chooses the funniest variant from the generated pool. An in-house evaluation of 200 generated headlines showed that the BERT-based model produces the funniest and in most cases grammatically correct output.
Crosstalk is a traditional Chinese theatrical performance art. It is commonly performed by two performers in the form of a dialogue. With the typical features of dialogues, crosstalks are also designed to be hilarious for the purpose of amusing the audience. In this study, we introduce CrossDial, the first open-source dataset containing most classic Chinese crosstalks crawled from the Web. Moreover, we define two new tasks, provide two benchmarks, and investigate the ability of current dialogue generation models in the field of crosstalk generation. The experiment results and case studies demonstrate that crosstalk generation is challenging for straightforward methods and remains an interesting topic for future works.
While humour and wordplay are among the most intensively studied problems in the field of translation studies, they have been almost completely ignored in machine translation. This is partly because most AI-based translation tools require a quality and quantity of training data (e.g., parallel corpora) that has historically been lacking for humour and wordplay. The goal of the JOKER@CLEF 2022 workshop was to bring together translators and computer scientists to work on an evaluation framework for wordplay, including data and metric development, and to foster work on automatic methods for wordplay translation. To this end, we defined three pilot tasks: (1) classify and explain instances of wordplay, (2) translate single terms containing wordplay, and (3) translate entire phrases containing wordplay (punning jokes). This paper describes and discusses each of these pilot tasks, as well as the participating systems and their results.KeywordsMachine translationHumourWordplayPunsNeologismsParallel corporaEvaluation metricsCreative language analysis
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 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}.
Open-domain chatbots engage with users in natural conversations to socialize and establish bonds. However, designing and developing an effective open-domain chatbot is challenging. It is unclear what qualities of a chatbot most correspond to users’ expectations and preferences. Even though existing work has considered a wide range of aspects, some key components are still missing. For example, the role of chatbots’ ability to communicate with humans at the emotional level remains an open subject of study. Furthermore, these trait qualities are likely to cover several dimensions. It is crucial to understand how the different qualities relate and interact with each other and what the core aspects would be. For this purpose, we first designed an exploratory user study aimed at gaining a basic understanding of the desired qualities of chatbots with a special focus on their emotional intelligence. Using the findings from the first study, we constructed a model of the desired traits by carefully selecting a set of features. With the help of a large-scale survey and structural equation modeling, we further validated the model using data collected from the survey. The final outcome is called the PEACE model (Politeness, Entertainment, Attentive Curiosity, and Empathy). By analyzing the dependencies between the different PEACE constructs, we shed light on the importance of and interplay between the chatbots’ qualities and the effect of users’ attitudes and concerns on their expectations of the technology. Not only PEACE defines the key ingredients of the social qualities of a chatbot, it also helped us derive a set of design implications useful for the development of socially adequate and emotionally aware open-domain chatbots.
https://www.joker-project.com/clef-2022/EN/project The goal of the JOKER workshop is to bring together translators and computer scientists to work on an evaluation framework for wordplay, including data and metric development, and to foster work on automatic methods for wordplay translation. Tasks We invite you to submit both automatic and manual runs! Manual intervention should be reported. Task 1: Сlassify and explain instances of wordplay. Task 2: Translate single words containing wordplay. Task 3: Translate entire phrases containing wordplay. Unshared task: We welcome any other type of submission that uses our data as an open task.
The increase in development of machines that have capability to understand the complicated behavior to solve the human brain involvement problems, the auto text generation application also gets the wide attention. The Language modeling or text generation is a task of next character or word prediction in a sequence with analysis of input data. The ATG enable the machines to write and provide the help to reduce the human brain effort. The ATG is also useful for understanding and analysis of languages and provide the techniques that enable the machines to exchange information in natural languages. At the large scale the text data are created everywhere (whatsApp, facebook, and tweets etc.) and freely online available therefore an effective system is needed for automation of text generation process and analysis of the text data for extracting meaningful information from it so in this work, a case study is presented on how develop a text generation model using hybrid recurrent neural network for English language. The explore model find the dependencies between characters and the conditional probabilities of character in sequences from the available input text data and generate the wholly new sequences of characters like human beings writing (correct in meaning, spelling and sentence structure). A comprehensive comparison between these models, namely, LSTM, deep LSTM, GRU and HRNN is also presented. Previously the RNN models are used for text predictions or auto text generation but these models created the problem of vanishing gradient (short memory) when process long text, therefore the GRU and LSTM models were created for solving this problem. The text generated by GRU and LSTM have many spellings error, incorrect sentence structure, therefore, filling this gap the HRNN model is explore. The HRNN model is the combination of LSTM, GRU and a dense layer. The experiments performed on Penn Treebank, Shakespeare, and Nietzsche datasets. The perplexity of HRNN model is 3.27, the bit per character is 1.18 and average word prediction accuracy is 0.63. As compare with baseline work and previous models (LSTM, deep LSTM and GRU), our model (HRNN) perplexity and bit per character are less. The texts generated by HRNN have fewer spelling errors and sentence structure mistakes. A closer analysis of explored models’ performance and efficiency is described with the help of graph plots and generated texts by taking some input strings. These graphs explain the performance for each model.
Recent text generation research has increasingly focused on open-ended domains such as story and poetry generation. Because models built for such tasks are difficult to evaluate automatically, most researchers in the space justify their modeling choices by collecting crowdsourced human judgments of text quality (e.g., Likert scores of coherence or grammaticality) from Amazon Mechanical Turk (AMT). In this paper, we first conduct a survey of 45 open-ended text generation papers and find that the vast majority of them fail to report crucial details about their AMT tasks, hindering reproducibility. We then run a series of story evaluation experiments with both AMT workers and English teachers and discover that even with strict qualification filters, AMT workers (unlike teachers) fail to distinguish between model-generated text and human-generated references. We show that AMT worker judgments improve when they are shown model-generated output alongside human-generated references, which enables the workers to better calibrate their ratings. Finally, interviews with the English teachers provide deeper insights into the challenges of the evaluation process, particularly when rating model-generated text.
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 article describes a model of automatic interpretation of English puns, based on Roget's Thesaurus, and its implementation, PunFields. In a pun, the algorithm discovers two groups of words that belong to two main semantic fields. The fields become a semantic vector based on which an SVM classifier learns to recognize puns. A rule-based model is then applied for recognition of intentionally ambiguous (target) words and their definitions. In SemEval Task 7 PunFields shows a considerably good result in pun classification, but requires improvement in searching for the target word and its definition.
Recent language models, especially those based on recurrent neural networks
(RNNs), make it possible to generate natural language from a learned
probability. Language generation has wide applications including machine
translation, summarization, question answering, conversation systems, etc.
Existing methods typically learn a joint probability of words conditioned on
additional information, which is (either statically or dynamically) fed to
RNN's hidden layer. In many applications, we are likely to impose hard
constraints on the generated texts, i.e., a particular word must appear in the
sentence. Unfortunately, existing methods could not solve this problem. In this
paper, we propose a backbone language model (backbone LM) for constrained
language generation. Provided a specific word, our model generates previous
words and future words simultaneously. In this way, the given word could appear
at any position in the sentence. Experimental results show that the generated
texts are coherent and fluent.
In this paper, we propose a novel neural network model called RNN Encoder--Decoder that consists of two recurrent neural networks (RNN). One RNN encodes a sequence of symbols into a fixed-length vector representation, and the other decodes the representation into another sequence of symbols. The encoder and decoder of the proposed model are jointly trained to maximize the conditional probability of a target sequence given a source sequence. The performance of a statistical machine translation system is empirically found to improve by using the conditional probabilities of phrase pairs computed by the RNN Encoder--Decoder as an additional feature in the existing log-linear model. Qualitatively, we show that the proposed model learns a semantically and syntactically meaningful representation of linguistic phrases.
In this paper, a technique for the generation of advertis- ing headlines based on affect is described. The system takes as input the target concept to evoke and possibly a sentiment to communicate. A fundamental dimen- sion of advertising relies on the affective connotation of words. We have used it as an important component of our algorithms for the creative variation of familiar expressions such as proverbs and famous movie titles. Recognition of affective connotation is based on a se- mantic similarity acquired in an unsupervised way from large corpora and it allows the system selecting words that are slanted towards a desired emotion, or producing strong ironic effects.
This paper will propose an objective measure based on the information theoretic notion of mutual information, for estimating word association norms from computer readable corpora. (The standard method of obtaining word association norms, testing a few thousand subjects on a few hundred words, is both costly and unreliable.) The proposed measure, the association ratio, estimates word association norms directly from computer readable corpora, making it possible to estimate norms for tens of thousands of words
A goal of statistical language modeling is to learn the joint probability function of sequences of words. This is intrinsically difficult because of the curse of dimensionality: we propose to fight it with its own weapons.
We have developed a formal model of certain types of riddles, and implemented it in a computer program, jape, which generates simple punning riddles. In order to test the model, we evaluated the behaviour of the program, by having 120 children aged eight to eleven years old rate jape--generated texts, human--generated texts, and non--joke texts for `jokiness' and funniness. This confirmed that jape's output texts are indeed jokes, and that there is no significant difference in funniness or jokiness between jape's most comprehensible texts and published human--generated jokes. Now at Sony Computer Science Laboratory Inc., Takanawa Muse Building., 3-14-13 Higashi-Gotanda Shinagawa-ku, Tokyo 141, Japan. 1 1 Introduction There is at present no general theory of humour which is fully detailed and rigorous. Even in the slightly more restricted area of verbal humour --- humour which is transmitted through language --- there is no agreed or well developed model of the underlying m...
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...
Deep Neural Networks (DNNs) are powerful models that have achieved excellent performance on difficult learning tasks. Although DNNs work well whenever large labeled training sets are available, they cannot be used to map sequences to sequences. In this paper, we present a general end-to-end approach to sequence learning that makes minimal assumptions on the sequence structure. Our method uses a multilayered Long Short-Term Memory (LSTM) to map the input sequence to a vector of a fixed dimensionality, and then another deep LSTM to decode the target sequence from the vector. Our main result is that on an English to French translation task from the WMT-14 dataset, the translations produced by the LSTM achieve a BLEU score of 34.7 on the entire test set, where the LSTM's BLEU score was penalized on out-of-vocabulary words. Additionally, the LSTM did not have difficulty on long sentences. For comparison, a strong phrase-based SMT system achieves a BLEU score of 33.3 on the same dataset. When we used the LSTM to rerank the 1000 hypotheses produced by the aforementioned SMT system, its BLEU score increases to 36.5, which beats the previous state of the art. The LSTM also learned sensible phrase and sentence representations that are sensitive to word order and are relatively invariant to the active and the passive voice. Finally, we found that reversing the order of the words in all source sentences (but not target sentences) improved the LSTM's performance markedly, because doing so introduced many short term dependencies between the source and the target sentence which made the optimization problem easier.
This tutorial introduces a new and powerful set of techniques variously called "neural machine translation" or "neural sequence-to-sequence models". These techniques have been used in a number of tasks regarding the handling of human language, and can be a powerful tool in the toolbox of anyone who wants to model sequential data of some sort. The tutorial assumes that the reader knows the basics of math and programming, but does not assume any particular experience with neural networks or natural language processing. It attempts to explain the intuition behind the various methods covered, then delves into them with enough mathematical detail to understand them concretely, and culiminates with a suggestion for an implementation exercise, where readers can test that they understood the content in practice.
In this work we explore recent advances in Recurrent Neural Networks for large scale Language Modeling, a task central to language understanding. We extend current models to deal with two key challenges present in this task: corpora and vocabulary sizes, and complex, long term structure of language. We perform an exhaustive study on techniques such as character Convolutional Neural Networks or Long-Short Term Memory, on the One Billion Word Benchmark. Our best single model significantly improves state-of-the-art perplexity from 51.3 down to 30.0 (whilst reducing the number of parameters by a factor of 20), while an ensemble of models sets a new record by improving perplexity from 41.0 down to 23.7. We also release these models for the NLP and ML community to study and improve upon.
We describe a method for transforming beam search into a complete search algorithm that is guaranteed to find an optimal solution. Called beam-stack search, the algorithm uses a new data structure, called a beam stack, that makes it possible to integrate systematic backtracking with beam search. The resulting search algorithm is an anytime algo- rithm that finds a good, sub-optimal solution quickly, like beam search, and then backtracks and continues to find im- proved solutions until convergence to an optimal solution. We describe a memory-efficient implementation of beam-stack search, called divide-and-conquer beam-stack search, as well as an iterative-deepening version of the algorithm. The ap- proach is applied to domain-independent STRIPS planning, and computational results show its advantages.
Because meaningful sentences are composed of meaningful words, any system that hopes to process natural languages as people do must have information about words and their meanings. This information is traditionally provided through dictionaries, and machine-readable dictionaries are now widely available. But dictionary entries evolved for the convenience of human readers, not for machines. WordNet ¹ provides a more effective combination of traditional lexicographic information and modern computing. WordNet is an online lexical database designed for use under program control. English nouns, verbs, adjectives, and adverbs are organized into sets of synonyms, each representing a lexicalized concept. Semantic relations link the synonym sets [4].
An implemented model of punning riddles
- Kim Binsted
- Graeme Ritchie
Kim Binsted and Graeme Ritchie. 1994. An implemented model of punning riddles. In Proceedings of the 12th National Conference on
Artificial Intelligence, Seattle, WA, USA, July
31 -August 4, 1994, Volume 1.. pages 633-638.
http://www.aaai.org/Library/AAAI/1994/aaai94-096.php.
Automatically extracting word relationships as templates for pun generation
- Anthony Bryan
- Ethel Hong
- Ong
Bryan Anthony Hong and Ethel Ong. 2009. Automatically extracting word relationships as templates for
pun generation. In Proceedings of the Workshop
on Computational Approaches to Linguistic Creativity. Association for Computational Linguistics,
Stroudsburg, PA, USA, CALC '09, pages 24-31.
Second order co-occurrence PMI for determining the semantic similarity of words
- Aminul Islam
- Diana Inkpen
Aminul Islam and Diana Inkpen. 2006. Second order co-occurrence PMI for determining the semantic
similarity of words. In Proceedings of the Fifth International Conference on Language Resources and
Evaluation, LREC 2006, Genoa, Italy, May 22-28,
2006.. pages 1033-1038.
Computational modelling of linguistic humour: Tom swifties
- Greg Lessard
- Michael Levison
Greg Lessard and Michael Levison. 1992. Computational modelling of linguistic humour: Tom swifties.
In In ALLC/ACH Joint Annual Conference, Oxford.
pages 175-178.
Neural machine translation (seq2seq) tutorial
- Minh-Thang Luong
- Eugene Brevdo
- Rui Zhao
Minh-Thang Luong, Eugene Brevdo, and Rui Zhao.
2017. Neural machine translation (seq2seq) tutorial.
https://github.com/tensorflow/nmt.
Generation of idiom-based witticisms to aid second language learning
- Justin Mckay
Justin McKay. 2002. Generation of idiom-based witticisms to aid second language learning. In In Stock
et al.. pages 77-87.
Automatic disambiguation of english puns
- Tristan Miller
- Iryna Gurevych
Tristan Miller and Iryna Gurevych. 2015. Automatic
disambiguation of english puns. In Proceedings
of the 53rd Annual Meeting of the Association for
Computational Linguistics and the 7th International
Joint Conference on Natural Language Processing of the Asian Federation of Natural Language
Processing, ACL 2015, July 26-31, 2015, Beijing,
China, Volume 1: Long Papers. pages 719-729.
http://aclweb.org/anthology/P/P15/P15-1070.pdf.
A practical application of computational humour
- Graeme Ritchie
- Ruli Manurung
- Helen Pain
- Annalu Waller
- Rolf Black
- Dave O' Mara
Graeme Ritchie, Ruli Manurung, Helen Pain, Annalu
Waller, Rolf Black, and Dave O'Mara. 2007. A
practical application of computational humour. In
Proceedings of the 4th International Joint Conference on Computational Creativity. pages 91-98.
- Ziang Xie
Ziang Xie. 2017. Neural text generation: A practical
guide. arXiv preprint arXiv:1711.09534.