Publications, Patents & Public Work

Research Direction

Currently building a research corpus towards a PhD by prior publication

Publications

A Dataset for Evaluating ASR on Specialised Vocabulary

LREC-COLING 2026

Research on specialised vocabulary recognition in enterprise ASR using open-vocabulary keyword spotting and synthetic data

accepted2026

Patents

Note: Patents represent IP protection, not peer-reviewed scientific publications. These innovations translate enterprise language-processing and speech AI gaps into protected technical methods.

Semantic Domain Assignment Referencing Governance Domains and Term Databases

US Patent 12518105-B2

Automated assignment of semantic-domain labels to terminology / translation data

granted2024

System and Method Performing Terminology Disambiguation

US Patent 12386820-B2

Context and metadata-based terminology disambiguation

granted2024

Detection of Abbreviation and Mapping to Full Original Term

US Patent 12067370-B2

Detecting and resolving non-standard enterprise abbreviations

granted2024

Adaptive Fidelity Pipeline for Minimizing Hallucinations and Skipped Content in STT Systems

STT preprocessing / verification logic for reducing hallucinations and skipped content

pending2026

Enhancing Pronunciation Accuracy and Naturalness in Japanese TTS without Extensive Fine-tuning

Pipeline-based Japanese TTS pronunciation optimisation for terminology-heavy scenarios

under review2026

Planned Public Artifacts

Technical notes, datasets, templates, and tools in preparation to share evaluation methodology, representativeness thinking, and speech AI quality knowledge publicly.

Evaluation Architecture for Enterprise Speech AI

Technical note on designing evaluation systems that connect metrics, benchmarks, and deployment decisions

in preparation

Representativeness Before Metrics

Technical note on why benchmark representativeness matters more than isolated metric scores

in preparation

Speech Evaluation Dataset Card Template

GitHub repository with templates for documenting speech evaluation datasets

in preparation

When Benchmarks Saturate: A Case Study in Enterprise Speech AI

Technical article on evaluation methodology and production validity

in preparation

Academic Projects

Dialogue Management: Improving Task-Oriented Conversational Agents with Deep Learning

Bachelor thesis on task-oriented dialogue systems using deep learning approaches

completed2018

NLG from Structured Inputs for Image Description Generation

Academic research project on natural language generation

completed2018

LexRank-based Text Summarization with Semantic Similarity Enhancements

Academic project on extractive text summarization with graph-based methods

completed2017

CHERTOY: Word Sense Induction for Web Search Result Clustering

Academic NLP project on computational linguistics

completed2017

Academic Profiles