Tilburg University (Exact location to follow)
Proceedings will be available here one week before the workshop
|09:00–11:00 Session 1|
|09:10–10:10||Content determination strategies for narrative: real-life reporting vs. fiction (Invited Talk)
|10:10–10:25||Seeking the Ideal Narrative Model for Computer-Generated Narratives
Mariana Ferreira and Hugo Gonçalo Oliveira.
|10:25–10:40||Discourse Embellishment Using a Deep Encoder-Decoder Network
Leonid Berov and Kai Standvoss.
|10:40–11:00||Exploring Lexical-Semantic Knowledge in the Generation of Novel Riddles in Portuguese
Hugo Gonçalo Oliveira and Ricardo Rodrigues.
|11:30–12:30 Session 2|
|11:30–11:50||Content Determination for Chess as a Source for Suspenseful NarrativesRichard Doust and Pablo Gervas.|
|11:50–12:00||Generating Stories Using Role-playing Games and Simulated Human-like Conversations
Alan Tapscott, Carlos León and Pablo Gervás.
|12:00–12:30||A further discussion around topics raised by the papers.|
Call for Papers
CC-NLG 2018 is the third edition of a series of workshops that aim to bring together researchers dealing with text generation from a computational creativity (CC) perspective, and researchers in natural language generation (NLG) with an interest in creative aspects, while also discussing the crossovers in CC and NLG, as well as looking to dialogue systems and their creative possibilities, for example in social networks.
Discussions at CC-NLG will cover the distinct approaches of CC and NLG brought about by their respective focuses; research in computational creativity has tended to deal less with technical shifts, directed instead at cognition, aesthetics, and novelty; whilst NLG research has tended to focus on the technical and theoretical challenges of topics like readability. However, in recent years this distinction has become far less defined. NLG research deals actively with concepts of style, variation, poetics, and narrative, whilst creative researchers are developing robust implementations. This change can be seen in dialogue systems, where the usability of an interface relies on it handling out-of-domain or spontaneous user input. For example, a virtual teacher should be able to process ‘Why can’t pigs fly?’, and a shopping assistant should be able to handle ‘What cool movies are out?’, while a humorous Twitterbot should be able to digest today’s news. These are just a few examples in the broad domain of CC + NLG.
Relevant Fields of Research
Research contributions are solicited on work that takes creativity into a general context, showing how the facets of creative language also enrich useful language. The workshop will still focus on core issues in computational creativity, like narrative, poetics, and humour, yet look to its communicative side. This will be of interest to researchers from computational creativity itself, as well as those looking to apply creative methods within other text generation systems.
Topics of interest include, but are not limited to:
- Poetry Generation
- Story Generation
- Generation of Metaphor, Figurative and Rhetorical Language
- Generation of Verbal Humor
- Personality and Emotion in NLG
- Creative Data-to-Text Models
- Interactive Language Generation
- Character-based Generation
- Style Generation
- Digital Literature
- Submissions due: September 7, 2018
- Notification of acceptance: October 7, 2018
- Camera-ready papers due: October 17, 2018
- Workshop: November 5, 2018
- INLG Conference: November 5-8, 2018
They should follow the INLG submission format and length limits. We are happy to accept both long papers, which will be given more time to present, and short papers . Check instructions at the INLG Call for Papers page. Dual submission with INLG 2018 is permitted, provided that authors notify us on submission, and withdraw from one venue after acceptance on the other.
- Long papers are most appropriate for presenting substantial research results and must not exceed 8 pages of content, with up to two additional pages for references.
- Short papers are more appropriate for presenting an ongoing research effort and must not exceed 4 pages of content, with up to one extra page for references.
Papers should be submitted electronically through the CC-NLG EasyChair submissions page.
As in previous editions, the organisers of CC-NLG will apply to have the accepted papers included as INLG Workshop Proceedings on the ACL Anthology.
The workshop will feature a keynote talk by Pablo Gervás:
“Content determination strategies for narrative: real-life reporting vs. fiction”
The crucial role of narrative in human communication has two different modes. Narrative is often used to report events extracted from our experience of real life that we want other people to know about. At other times, narrative is used to convey entirely fictional set of events that have been put together for the purpose of entertaining or exercising our sensibility. Although the surface forms of the discourses created in each of these modes are fundamentally similar, the content determination strategies employed during composition present significant differences. Whereas in one case the process is mainly one of selection, in the other there is likely to be a large percentage of invention. Yet both the selection and the invention are likely to be guided or informed by considerations of narrative structure and successful narrative outcomes. The talk will describe recent work on computational modelling of these two narrative modes and present insights arising from their joint consideration into a single framework capable of switching from one to the other (and considering hybrid modes of operation) much as humans are seen to do.