**Introduction** Computational Cultural Understanding Korean (LDC2026T11) was developed by the Linguistic Data Consortium (LDC) and consists of Korean source data, manual annotations, assessment results, and associated supplemental resources used for development and evaluation in the DARPA Computational Cultural Understanding (CCU) program. The CCU program focused on developing technologies to improve cross-cultural understanding and communication with the goal to create systems capable of analyzing and interpreting sociocultural norms, recognizing emotion across cultural contexts, and detecting communicative changes that affect interpersonal understanding. Those systems in turn would support development of a dialogue assistance application designed to enhance operator situational awareness and interactional effectiveness in cross-cultural settings. Research was organized into three technical areas (TAs): TA1 Sociocultural Analysis, TA2 Cross-Cultural Dialogue Assistance, and TA3 Data Creation for Development and Evaluation. Systems were evaluated on valence, arousal and emotion category recognition; sociocultural norm detection; and the detection of potentially impactful changes in the outcome of the interaction. LDC was a TA3 performer and responsible for collecting, curating, annotating, assessing and distributing multimodal data to support TA1 research and system evaluation across six language-culture pairs: Mandarin Chinese, Spanish, Russian, Japanese, Korean, and Turkish. **Data** This corpus contains over 6500 documents and URLs with corresponding annotations and constitutes the complete set of Korean CCU TA1 development and evaluation data, annotated metadata for each source document, manual transcripts for development data, and automatic apeech recognition (ASR) output for the evaluation data. Source data consists of naturally occurring conversations involving two or more Korean speakers in text, audio, or video (provided as URLs) that represent a wide variety of speaker demographics, social roles, settings, genres, topics, emotions, and sociocultural norms. Documents were manually labeled for the presence of emotion categories, valence, and arousal. A subset of files was also annotated for defined sociocultural norms, including norm adherence or violation, and for the presence of changepoints (points in a conversation where a change occurs with the potential to significantly impact conversation outcome). Source data is presented as .psm, .ltf, and .flac files. Annotations, assessment, and ASR ouput are presented as tab separated files (.tab). All text data is UTF-8 encoded. Also included is a software tool to recreate original source data from the processed XML material. **Sponsorship** This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Contract No. HR0011-22-C-0033. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of DARPA. **Updates** No updates at this time.

Creator(s)

Dana Delgado

Summer Ploegman

Song Chen

Christopher Caruso

Brian Gainor

Ann Bies

Kira Griffitt

Stephanie Strassel

Distributor(s)

Linguistic Data Consortium

Right Holder(s)

Linguistic Data Consortium