During our presentation, we will take on the role of the literary adept and talk with a chatbot, who we will treat as our master. We’ll ask him questions about how to write, present our works for his evaluation, and try to receive feedback. The Master will use phrases, sentences and paragraphs of texts, which until now have been used in literary discussions. The chatbot that we propose is based on texts from the history of Polish literature, foremost taking into account the exchange of views between literary critics and historians. It is said that one of the peculiarities of Polish mentality is strife, which especially in the digital age takes on monstrous proportions in the form of an uncontrolled wave of hate on the internet and verbal abuse that falls below the belt.
The starting point for our project is to recognize that the majority of existing chatbots are very nice. The available chatbots try to help, give advice and have an answer ready on how to proceed. They look good and behave impeccably. Our idea is radically different: we want to create a bot that is programmed to be unpleasant, to be a troll and hater. The first step will be to research this behavior in Polish literature and based on the literary haters and their hate, we will create a database of possible answers. We are going to use both classical texts, literary quarrels between the romantics and representatives of the Enlightenment, and the avant-gardists attacking tradition, and we will mix these with discussions on literary web portals, social media, statuses and comments. The chatbot will adjust its answer to the user’s questions by employing a simple word order analyzer and keywords. An archive of literary texts will be processed using a sampling method and Markov chains.
(Source: ELO 2017: Book of Abstracts and Catalogs)
