Text-to-image generation systems are increasingly capable of producing professional-looking infographics from natural language prompts. While these systems offer substantial efficiency gains for data communication, their outputs risk being adopted uncritically by users who lack the data visualization literacy needed to identify misleading elements — a competency that the systems themselves do not possess.

Due to the growing number of available games, game reviews provide a vital basis for decision-making when buying a game. Apart from that, these reviews could serve as a rich design resource and inspiration for game designers. In this paper, we analyze game reviews formulated by players via a large language model and ask GPT-3, an artificial intelligence, game design-related questions based on the Player Experience Inventory (PXI) questionnaire. The obtained results are evaluated in expert interviews, where we asked four indie game designers about the quality of the game review analyses. The experts were generally impressed by the AI-generated texts but also provided some recommendations regarding the game review analyses’ quality, structure, and setup.

In this work we investigate the possibilities of integrating a Deep Learning language model for a Natural Language Interface (NLI) of an information visualisation software. For this purpose, we have developed a prototype web application that uses the deep learning model OpenAI Codex from the GPT3 family to create visualisations from text input. For comparison, we created a second prototype with a classical NLP approach based on NL4DV toolkit (with subtasks like part-of-speech (POS) tagging, entity recognition, and dependency parsing) and an almost identical interface. The two variants were subjected to a study with test persons, and the advantages and disadvantages of the two approaches and the suitability for the most common visualisation types were investigated. The Deep Learning approach offers greater expressiveness for describing the graphics, but also the danger of not always being entirely comprehensible. The participants were able to use it to create more complex visualisations, but also sometimes had problems finding the right text input to solve the tasks. In our preliminary usability study, the Deep Learning prototype performed slightly better than the comparison prototype and achieved a useful usability score.

We provide a method for visualizing the information associated with the clusters used for topic modeling of Instagram Message Feeds. For this purpose, a series of interactive dashboards are used to determine the right number of clusters and a suitable interpretation of each cluster. These extend previous approaches for regular text documents and focus on including specific information in Instagram feeds such as hashtags and linking structure.