This paper presents and evaluates a new educational tool and a model for learning theory and practice of game development, by involving geographically separated and culturally different industrial and academic organizations. We propose a TRIAD education model, which is as an extension to the conventional project-based learning, by uniting industry staff, faculty members, and students in an online discussion space. This model is a two-phase cyclic model where in Phase-1) industry members and academic staff discuss the progress of PBL transparently, showing the cross-cultural gaps and advising how to solve them, and in Phase-2) students make progress on the projects. Additionally, the TRIAD model adopts asynchronous text messaging as a core communication method, which is suitable for a multi-timezone situation, as well as exchanges of self-reviews from students and feedback from instructors. The online space not only facilitates interns in overcoming the hurdles represented by the gaps between the industry and academia, but it also visualizes the students' implicit educational progress by periodically analyzing all stored data by applying quantitative text analysis methods. This paper shows the result of an empirical study by utilizing this TRIAD model and analytics tool conducted at a large-scale Japanese game company with over 2,000 employees focusing on two students from the division of game development at a university in Sweden. We applied a quantitative text analysis tool for this model during an on-going internship program to clarify how students change their thoughts and behaviors by acquiring cross-cultural professionalism in the game industry. In addition, the result indicates that this TRIAD model is resilient, even under the COVID-19 epidemic situation due to its text-based approach.

A novel method to induce wide-coverage Combinatory Categorial Grammar (CCG) resources for Japanese is proposed in this article. For some languages including English, the availability of large annotated corpora and the development of data-based induction of lexicalized grammar have enabled deep parsing, i.e., parsing based on lexicalized grammars. However, deep parsing for Japanese has not been widely studied. This is mainly because most Japanese syntactic resources are represented in chunk-based dependency structures, while previous methods for inducing grammars are dependent on tree corpora. To translate syntactic information presented in chunk-based dependencies to phrase structures as accurately as possible, integration of annotation from multiple dependency-based corpora is proposed. Our method first integrates dependency structures and predicate-argument information and converts them into phrase structure trees. The trees are then transformed into CCG derivations in a similar way to previously proposed methods. The quality of the conversion is empirically evaluated in terms of the coverage of the obtained CCG lexicon and the accuracy of the parsing with the grammar. While the transforming process used in this study is specialized for Japanese, the framework of our method would be applicable to other languages for which dependency-based analysis has been regarded as more appropriate than phrase structure-based analysis due to morphosyntactic features.