Developing or analyzing algorithms that find "frequent episodes" within event streams (e.g., telecommunication logs or user behavior data).
Testing models on episode summarization or genre prediction using a corpus of thousands of audio files. Suggested Paper Structure All Episodes
Creating a dataset from a large-scale collection of podcast episodes for natural language processing. All Episodes
Using video-based research to identify patterns in social or organizational behavior. All Episodes
Detail how the episodes were selected and transcribed.
Transcribe all episodes of a recorded interaction to ensure inter-rater reliability (IRR) and build overt interpretations into a central narrative document. Linguistics and AI: Spoken Document Corpora
Explore the "evolution of the story" or data patterns over the course of the episodes.