Keywords

Multilingual barriers, Multilingual extraction ontologies, Cross-language communication, Linguistic grounding, Tech-transfer opportunities

Abstract

In our global society, multilingual barriers sometimes prohibit and often discourage international travelers from receiving desired goods and services. In a multilingual setting, easing linguistic-based tasks such as finding a restaurant, learning how to use local mass transit, and making simple business transactions would help remove barriers and encourage international travelers to seek and enjoy a wider variety goods and services. We propose multilingual extraction ontologies accessible via mobile devices as an aid to resolving these issues. Multilingual ontologies can communicate in multiple languages with the wireless local web as well as the international traveler and can thus break through language barriers and allow access to local markets and services. As envisioned, ontologies provide a conceptual framework for a narrow domain of interest. Grounding narrow-domain ontologies linguistically enables them to map relevant utterances and text to meaningful concepts in the ontology. Linguistically grounding ontologies in multiple languages enables cross-language communication. Technically, we can gauge the success of linguistically grounded ontologies by measuring precision and recall of extracted concepts, and we can gauge the success of automated crosslinguistic-mapping construction by measuring the speed of creation and the accuracy of generated lexical resources. From a business perspective, we can measure success in at least two ways. First, tech-transfer success ultimately generates measurable royalty revenue and a user base that demonstrates whether the technology is successful at a consumer level. Second, we can see how well a business plan built on our core technology performs in student competitions hosted at BYU and other universities. We envision an interdisciplinary mentored research environment as an ideal way to resolve the myriad of technical problems involved in building the envisioned collection of multilingual extraction ontologies. While a computer science student leverages new algorithms and data structures to construct linguistically grounded domain ontologies and a linguistics student finds cost-effective ways to mesh multiple languages with respect to various ontologies, an entrepreneurial information-systems student can investigate tech-transfer opportunities and resolve technical issues between a research prototype and an industrial-strength, deployable system. In addition to being interdisciplinary the PIs have expert foreign language knowledge. Each BYU PI speaks one or more foreign languages (German, Spanish, French, and Japanese). Further, the involvement of Professor Tijerino, who will be on sabbatical leave at BYU from Kwansei Gakuin University in Japan, along with his two native Chinese graduate students provide the necessary international constituents that comprise the mentoring research team.

Original Publication Citation

David W. Embley, Stephen W. Liddle, Deryle W. Lonsdale, and Yuri Tijerino (2012). Multilingual Extraction Ontologies. ORCA Mentoring Environment Grant (MEG) Final Report, Journal of Undergraduate Research, BYU.

Document Type

Report

Publication Date

2012

Publisher

Brigham Young University

Language

English

College

Humanities

Department

Linguistics and English Language

University Standing at Time of Publication

Associate Professor

Download

Included in

Linguistics Commons

Share

COinS