AI in Language Research
International Conference

The AI in Language Research International Conference (AILRIC) is an international gathering of researchers, practitioners, and scholars working at the intersection of artificial intelligence and human language.  It brings together expertise from linguistics, natural language processing (NLP), speech technology, and allied disciplines to advance the analysis, processing, and generation of human languages.

Organized by the De La Salle University College of Computer Studies – Advanced Research Institute for Informatics, Computing, and Networking (AdRIC), through its Center for Language Technologies (CeLT) in partnership with the University of San Carlos Department of Computer, Information Sciences, and Mathematics and University of Cebu, AILRIC 2026 serves as a dynamic forum for presenting cutting-edge research, fostering interdisciplinary collaboration, and strengthening research networks across institutions and regions.  Past events organized by the center have featured a diverse range of NLP topics and have been enriched by the participation of distinguished international invited speakers.

In the rapidly evolving landscape of AI - marked by the rise of large language models, multimodal systems, and data-driven approaches - AILRIC 2026 aims to critically engage with both technological advances and their linguistic, cultural, and societal implications.

AILRIC 2026 aims to:

1. Strengthen research ecosystems by establishing sustainable mechanisms for advancing AI-driven language research within academic and research institutions; 

2. Promote capacity building through exposure to current trends, methodologies, and tools in NLP, AI, and language technologies; 

3. Support research development by providing mentorship, collaboration opportunities, and technical guidance for ongoing and emerging projects; 

4. Foster interdisciplinary dialogue among linguists, computer scientists, and domain experts to address complex language-related challenges; and 

5. Encourage responsible and inclusive AI by highlighting issues of ethics, linguistic diversity, and low-resource language support.

AILRIC 2026 provides a platform for discussing both foundational and emerging challenges in AI-driven language research, including the integration of human language technologies for documentation, preservation, and large-scale processing of languages - particularly in multilingual and low-resource contexts.

Topics:

The Living Manga: Giving Manga a Heartbeat Through Empathic Computing and Music

Norshuhani Zamin, Universiti Tenaga Nasional (UNITEN), Malaysia

Abstract
Traditional manga relies heavily on static visual cues like line art, panel layout, and text to evoke reader emotion. However, these static formats lack the dynamic sensory layer needed to match a reader's shifting psychological state in real time. The research introduces "The Living Manga," a novel empathic computing framework that dynamically syncs background music with manga scenes based on real-time emotional tracking. With Machine Learning, our system maps the semantic of the page scene with  specific emotional states. It then triggers algorithmic music tracks tailored to the exact tone, pacing, and mood of the current manga panel.

Preliminary user studies demonstrate a significant increase in story immersion, deeper emotional resonance, and higher reader engagement compared to traditional silent reading. By bridging the gap between digital art and sensory technology, this research provides a foundational blueprint for the future of interactive storytelling and emotionally aware media.

Speaker Bionote
Dr. Norshuhani Zamin is a senior academic and international higher education professional at Universiti Tenaga Nasional (UNITEN), Malaysia, affiliated with the College of Computing and Informatics (CCI) and involved in international student mobility and global engagement through the International Office. She also serves as a Visiting Professor at the University of San Carlos, Philippines. With more than 25 years of academic and higher education experience, she has held academic positions at institutions including De La Salle University, Philippines; Saudi Electronic University, Saudi Arabia; Universiti Teknologi PETRONAS (UTP); Universiti Malaysia of Computer Science and Engineering (UNIMY); and Universiti Sains Islam Malaysia (USIM). She obtained her PhD in Information Technology from Universiti Teknologi PETRONAS in 2014.

Her research expertise encompasses Artificial Intelligence, Machine Learning, Natural Language Processing, IoT, empathic computing, and applied computing, with research contributions spanning Malay text processing, assistive communication technologies, educational technology, AI-based agriculture, and machine learning applications. She has published scientific articles, contributed to international collaborations, secured research funding, and received recognition through research and innovation activities at national and international levels.

No One Builds Filipino AI Alone: Open Models, Benchmarks, and Data for Southeast Asia

Railey Montalan, AI Singapore

Abstract
Southeast Asia continues to lag in AI development, and the Philippines faces this gap acutely. Frontier models available to Filipino users and developers are costly and weakly aligned to local languages and culture, limiting the ability of individuals, companies, and public institutions to build and scale services. Most open resources, from web-scale corpora to open-weight models to widely-used evaluation suites, are optimized for other languages first, leaving the Philippines behind regional peers like Thailand, Indonesia, and Vietnam.

This talk presents the approach taken at AI Singapore, a national AI programme, to address this gap. It covers a stack of tools for open, regionally aligned AI: the SEA-LION family of efficient open-source multilingual and multimodal models, now agentic and deployable on modest hardware; ancillary embedding and safety models, including SEA-Guard; the SEA-HELM evaluation suite and leaderboard; and ATLAS, an open catalogue of underrepresented-language datasets. On SEA-HELM, SEA-aligned models show consistent gains across several SEA languages, including Filipino, over the base models they adapt.

General capability, however, is not the endpoint. Real-world applications still require stronger downstream models for machine translation, speech transcription and synthesis, and agentic workloads, where local linguistic and cultural expertise is decisive. We call for collaboration with researchers, linguists, and students to create novel Filipino training and alignment datasets, build more culturally-aware evaluations, and co-develop models and resources that serve the region's communities.

Speaker Bionote
Railey Montalan is an AI researcher and engineer at AI Singapore, where he evaluates the cultural reasoning and agentic capabilities of large language models in Filipino and other Southeast Asian languages. He also helps develop SEA-LION, AI Singapore's family of multilingual LLMs and ancillary models optimized for SEA. Before joining AI Singapore, he taught NLP electives at the Ateneo de Manila University and worked in business intelligence.

Select papers presented at AILRIC 2026 will be invited for submission to the Philippine Computing Journal, subject to the journal’s peer-review process.

The Student Research Workshop (SRW) at AILRIC 2026 provides a dedicated venue for undergraduate and graduate students to present work-in-progress, exploratory studies, and small-scale projects in the area of AI and language research.

The workshop aims to foster early-stage research development by offering students the opportunity to receive constructive feedback from peers and senior researchers, while engaging with the broader AILRIC community.

In natural language processing (NLP), it is common practice to leverage data and resources from linguistically similar languages to address data scarcity.  Such approaches, often seen in cross-lingual transfer and multilingual modeling, assume that similarity between languages can be operationalized and exploited.  However, for many Philippine languages, systematic and empirically grounded measures of similarity remain underexplored. This shared task is motivated by the need to better understand and quantify these relationships.

Accepted shared task papers will be presented through a poster session at AILRIC 2026.

Participants were asked to:

• Use publicly available data or construct their own dataset

• Compute pairwise similarity scores between selected languages

• Generate a similarity matrix representing these relationships

• Produce language clusters (e.g., dendrograms, embeddings, or other visualizations)

• Compare results with established linguistic classifications (e.g., Ethnologue or other scholarly resources) and provide a well-motivated interpretation of the observed structures

The task is intentionally open-ended to allow diverse methodologies, including:

• Lexical similarity (e.g., edit distance, cognate detection)

• Character- and token-level distributional similarity (e.g., n-grams)

• Embedding-based similarity (e.g., word/sentence embeddings)

• Phonological or orthographic similarity modeling

• Typological or feature-based comparison

• Multimodal or metadata-informed approaches (e.g., geography, speaker communities)

Payment can be made through bank deposit.

• Account No. (in PHP):  004588008272    

• Account Name:  DE LA SALLE UNIVERSITY, INC.    

• Bank Name:  BDO Unibank Inc

• Bank Address:  2422 Taft Avenue Malate Manila  1004  Philippines    

• Kindly specify at the Purpose/Remarks: AILRIC

The deposit slip should also be presented during onsite check-in.

The registration fee covers admission to all conference sessions, lunch, morning and afternoon snacks. Accommodation and airport transportation are not included in the registration fee.

Every conference paper requires at least one registration by the early bird deadline to be included in the program.  For papers accepted to the Student Research Workshop, the first author and presenter must be a student.

All participants must fill up the registration form available at: https://forms.gle/SAkW7zKH9m1RTjm49.

Oriental COCOSDA 2026 | O-COCOSDA 2026