Guest Talk by Ayu Purwarianti
Date/Time: 10.05.2023, 12:00
In her talk, Ayu will present several branches of NLP research for the Indonesian language conducted at her group in the Bandung Institute of Technology (ITB), namely:
1. Data collection on Indonesian national and regional languages, along with the NLP tools and pre-trained language model built from the collected data. NLP research requires qualified NLP data unavailable for Indonesian national and regional languages. Thus, we collected these data via in-house data labelling team and crowd-sourcing. We use the collected data to build Indonesian pre-trained language models and various NLP tools.
2. Social media analytics for the Indonesian language, including sentiment analysis and fake news classification. The sentiment analysis includes document-level and aspect-based sentiment analysis for Indonesian. We compared the extraction, classification and generative approach for the aspect-based sentiment analysis. Further, we compare the social media message and fake news database on the fake news classification.
3. Regulatory technology for Indonesian legal rules. We crawled Indonesian legal regulation documents, extracted important information, and built a connection map between Indonesian legal regulations to make it easier for users to understand the relationship between rules.
4. Question Answering for the Indonesian language. We tried two approaches in the Indonesian question-answering system: a) machine learning approach in the answer finder on the retrieved relevant passages; b) transforming questions into SPARQL against knowledge graph constructed from Indonesian text (domain: cosmetics, text sources: Indonesian e-commerce portal)
Ayu Purwarianti was graduated from PhD program at Toyohashi University of Technology in December 2007 with dissertation title of “Cross Lingual Question Answering System (Indonesian Monolingual QA, Indonesian-English CLQA, Indonesian-Japanese CLQA)”. The dissertation was in Natural Language Processing or also known as Computational Linguistics which is part of Artificial Intelligence knowledge domain. Since then, she has worked as a lecturer at ITB (Bandung Institute of Technology), the alma mater where she got her bachelor’s and master’s degree. She has got several research funds from ITB, DIKTI, LPDP, and ASEA UNINET. In 2017, she received an award from ASEA UNINET on the collaboration research project with TU Wien and University of Indonesia. Her research works are mostly on Indonesian natural language processing. According to scopus, she already published 155 papers, with h-index of 13 and 848 citations. Other than teaching and doing research, she was also assigned as the head of Informatics/Computer Science bachelor’s degree at ITB on 2012-2015 which eventually took her to an award of 2nd best national head of program in 2014. Then on 2016-2020, she was assigned as the Head of Doctoral Degree in Electrical Engineering and Informatics, ITB. Her other activity is in Indonesian Association for Computational Linguistics where she was elected as the chair for 2016-2018; and she was also the chair of IEEE Education chapter of Indonesian section for 2017-2019. She has joined IABEE (Indonesian Accreditation Board of Engineering Education) since 2015 until now. She also founded a start up named Prosa.ai since 2018. She is now the Chair of Artificial Intelligence Center at ITB since August 2019.