Portugal

High-quality human-created artifacts are distinguished by cohesion and semantic richness that can be difficult for computational systems to emulate effectively. Certain classes of artifacts feature relationships between their constituent elements (e.g. words in a story) that naturally form a hierarchical structure that underpins the artifact&#39;s meaning. We formalize a tractable method for programmatically extracting such artifacts&#39; structures framed in hierarchical Bayesian program learning (HBPL) and present HIEROS, a creative system that uses linguistic and semantic structures---collectively called formats---extracted from human-written six-word stories to guide the creation of novel stories. We describe how HIEROS infers formats from exemplar stories, how those formats inform its selection of words when writing novel stories, and how the system evaluates its stories to search for high-quality output. We present and evaluate stories HIEROS has written and discuss how our format model enhances the system&#39;s creativity.

ICCC&#39;20

Creating Six-word Stories via Inferred Linguistic and Semantic Formats

story formats

hbpl

creative system

six-word stories

2021 Annual Conference of the North American Chapter of the Association for Computational Linguistics.

**Whova App** 
Stay in touch with your fellow conference attendees via the [Whova App](https://whova.com/portal/webapp/nacon_202106/)

**Conference Structure**
https://2021.naacl.org/blog/conference-structure/

**Walkthrough video of how to NAACL 2021** 

Please take a moment to view this video explaining how to navigate the platform, attend sessions network with other attendees. 


<figure class="video_container">
  <iframe src="https://screencast-o-matic.com/watch/crhwbGVh3vx?v=6&ff=1&title=0&controls=1" width=640  height=350  frameborder="0" allowfullscreen="true"> </iframe>
</figure>

<br>
<br>

NAACL 2021

High-quality human-created artifacts are distinguished by cohesion and semantic richness that can be difficult for computational systems to emulate effectively. Certain classes of artifacts feature relationships between their constituent elements (e.g. words in a story) that naturally form a hierarchical structure that underpins the artifact's meaning. We formalize a tractable method for programmatically extracting such artifacts' structures framed in hierarchical Bayesian program learning (HBPL) and present HIEROS, a creative system that uses linguistic and semantic structures---collectively called formats---extracted from human-written six-word stories to guide the creation of novel stories. We describe how HIEROS infers formats from exemplar stories, how those formats inform its selection of words when writing novel stories, and how the system evaluates its stories to search for high-quality output. We present and evaluate stories HIEROS has written and discuss how our format model enhances the system's creativity.

technical paper

Computational creativity is the art, science, philosophy and engineering of computational systems which, by taking on particular responsibilities, exhibit behaviours that unbiased observers would deem to be creative. As a field of research, this area is thriving, with progress in formalising what it means for software to be creative, along with many exciting and valuable applications of creative software in the sciences, the arts, literature, gaming and elsewhere.

The ICCC conference series organized by the Association for Computational Creativity since 2010 is the only scientific conference that focuses on computational creativity alone and also covers all aspects of it.


ICCC'20

The ICCC conference series organized by the Association for Computational Creativity since 2010 is the only scientific conference that focuses on computational creativity alone and also covers all aspects of it.
Computational creativity is the art, science, philosophy and engineering of computational systems which, by taking on particular responsibilities, exhibit behaviors that unbiased observers would deem to be creative. As a field of research, this area is thriving, with progress in formalizing what it means for software to be creative, along with many exciting and valuable applications of creative software in the sciences, the arts, literature, gaming and elsewhere.



We present a methodology to understand and name, from a computational creativity perspective, the agency of algorithm and artists in  interactive processes. It consists in first analysing precisely the algorithm creation and then the back and forth creative loop during the interaction. We focus on a particular inteactive setting and in which a painter and various neural style transfer algorithms interact on a real canvas. We gather a set of paired painting-pictures images and analyse the style transfer  predictivity. Then, in the creative interaction , the algorithm outputs were perceived as a source of inspiration for human painters, portraying the machine as a \textit{computational catalyst}.

Interactive neural style transfer with artists

Modular creative systems employ specialized modules that take on challenging responsibilities. Such modules can be expensive to develop and are often imperfect, leading to uncertainty in whether a system's poor results are due to module imperfection or larger systemic issues. Computational creativity research would benefit from a method for validating the design of creative systems before expending development resources implementing them. We present ideal-module prototyping (IMP), a design validation paradigm that serves these needs by using modules that delegate responsibilities to humans as an ideal against which to compare system expectations and performance. We relate an experiment we conducted in applying IMP to an existing creative system to demonstrate the insights that can be gained from this method. We argue for the widespread adoption of modular creative system design validated by ideal-module prototyping.

Humans in the Black Box: A New Paradigm for Evaluating the Design of Creative Systems

Not all knowledge is created equal. A hierarchical architecture is a method to classify knowledge for use in the field of human cognition and computational creativity. This paper introduces an Insight-Knowledge Object (IKO) model as a framework for Artificial Creative Intelligence (ACI), a step forward in the pursuit of replicating general human intelligence with computing machinery. To achieve ACI, it is hypothesized that a fundamental rethinking of the architecture of human cognition and knowledge processing is required. One possible novel architecture could be the IKO model. The authors include a description of on-going work at Carnegie Mellon University that applies the IKO model in practice with an artificial music improvisation embodiment.

Artificial Creative Intelligence: Breaking the Imitation Barrier

Interactive fictions, or text-adventures, are games in which a player interacts with a world entirely through textual descriptions and text actions. Text-adventure games are typically structured as puzzles or quests wherein the player must execute certain actions in a certain order to succeed. In this paper, we consider the problem of procedurally generating a quest, defined as a series of actions required to progress towards a goal, in a text-adventure game. Quest generation in text environments is challenging because they must be semantically coherent. We present and evaluate two quest generation techniques: (1) a Markov model, and (2) a neural generative model. We specifically look at generating quests about cooking and train our models on recipe data. We evaluate our techniques with human participant studies looking at perceived creativity and coherence.

Toward Automated Quest Generation in Text-Adventure Games

"We present a system for real-time lyrical improvisation between a human and a robot in the style of hip hop. Our system takes vocal input from a human rapper, analyzes the semantic meaning, and generates a response that is rapped back by a robot over a musical groove. Previous work with real-time interactive music systems has largely focused on instrumental output, and vocal interactions with robots have been explored, but not in a musical context. Our generative system includes custom methods for censorship, voice, rhythm, rhyming and a novel deep learning pipeline based on phoneme embeddings. The rap performances are accompanied by synchronized robotic gestures and mouth movements. Key technical challenges that were overcome in the system are developing rhymes, performing with low-latency and dataset censorship.
We evaluated several aspects of the system through a survey of videos and sample text output. Analysis of comments showed that the overall perception of the system was positive. The model trained on our hip hop dataset was rated significantly higher than our metal dataset in coherence, rhyme quality, and enjoyment. Participants preferred outputs generated by a given input phrase over outputs generated from unknown keywords, indicating that the system successfully relates its output to its input. "

Shimon the Rapper: A Real-Time System for Human-Robot Interactive Rap Battles

The field of bisociative literature-based discovery aims at exploring scientific literature to reveal new discoveries based on yet uncovered relations between knowledge from different, relatively isolated fields of specialization. This paper outlines selected outlier-based literature mining approaches, which focus on finding outlier documents as means for finding unexpected links crossing different contexts. Selected approaches to bridging term detection through outlier document exploration are briefly outlined, together with the lessons learned from recent applications in medical and biological literature-based knowledge discovery. Finally, the paper addresses new prospects in bisociative literature-based discovery, emphasizing the use of advanced embeddings technology for cross-domain literature mining.

Bisociative Literature-based Discovery: Lessons Learned and New Prospects

This papers revisits the receptive theory in context of computational creativity. It presents a case study of a Paranoid Transformer — a fully autonomous text generation engine with raw output that could be read as the narrative of a mad digital persona without any additional human post-filtering. We describe technical details of the generative system, provide examples of output and discuss the impact of receptive theory, chance discovery and simulation of fringe mental state on the understanding of computational creativity.

Paranoid Transformer: Reading Narrative of Madnessas Computational Approach to Creativity

The ability to extract features from objects or concepts and to connect them in a meaningful way is believed to be crucial to creativity. This paper proposes a novel computational model for creative behaviour, by learning extractions and connections separately. Such separation enables adaptive feature and object connections, which means one object can be connected to different other objects within different contexts. This paper applies recent cognitive theories of computational creativity to two specific tasks: an image-to-image mapping and music- to-video mapping. In both cases deep neural networks are used to automate these two creative cognition tasks.

Being Creative: A Domain Mapping Network

We present Lemotif, an integrated natural language processing and image generation system that uses machine learning to (1) parse a text-based input journal entry describing the user's day for salient themes and emotions and (2) visualize the detected themes and emotions in creative and appealing image motifs. Synthesizing approaches from artificial intelligence and psychology, Lemotif acts as an affective visual journal, encouraging users to regularly write and reflect on their daily experiences through visual reinforcement. By making patterns in emotions and their sources more apparent, Lemotif aims to help users better understand their emotional lives, identify opportunities for action, and track the effectiveness of behavioral changes over time. We verify via human studies that prospective users prefer motifs generated by Lemotif over corresponding baselines, find the motifs representative of their journal entries, and think they would be more likely to journal regularly using a Lemotif-based app.

Lemotif: An Affective Visual Journal Using Deep Neural Networks

"This work is on the transformation of text relying mostly on a common analogy vector operation, computed in a distributional model, i.e., static word embeddings.
Given a theme, original song lyrics have their words replaced by new ones, related to the new theme as the original words are to the original theme.
As this is not enough for producing good lyrics, towards more coherent and singable text, constraints are gradually applied to the replacements.
Human opinions confirmed that more balanced lyrics are obtained when there is a one-to-one mapping between original words and their replacement; only content words are replaced; and the replacement has the same part-of-speech and rhythm as the original word."

Creating Six-word Stories via Inferred Linguistic and Semantic Formats

Downloads

Next from ICCC'20

Interactive neural style transfer with artists

Similar lecture

Multi-Adversarial Learning for Cross-Lingual Word Embeddings

Creating Six-word Stories via Inferred Linguistic and Semantic Formats

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from ICCC'20

Interactive neural style transfer with artists

Similar lecture

Multi-Adversarial Learning for Cross-Lingual Word Embeddings

Downloads