I'm a PhD student in Information Retrieval at the Center for Intelligent Information Retrieval at University of Massachusetts Amherst, supervised by Hamed Zamani.
I’m currently doing research in conversational search. I'm also the founder of Naget, an upcoming low-fee everything-platform that among other features will introduce personal instruction-based search.
Previously, I was at Ro5 where I worked on the automated construction of a biomedical knowledge graph for drug discovery, a research assistant at the National University of Singapore where I did research supervised by Wynne Hsu and Mong Li Lee in automated fact-checking and detection of misinformation in social media, at Plural AI where I worked on entity extraction from webpages and semantic search for finance, and at Donaco where I worked on contextual recommendation of charity campaigns.
My goal is to help overcome the limitations of current machine learning models (efficiency, robustness, limited training resources), and improve multimodal content organization & discovery. My recent work is in conversational search, representation learning and ranking for information retrieval, contextual recommendation systems and knowledge base construction from semi-structured text.
Research Interests: Information Retrieval, NLP, Machine Learning, Recommendation Systems, Knowledge Graphs
Previously, I was at Ro5 where I worked on the automated construction of a biomedical knowledge graph for drug discovery, a research assistant at the National University of Singapore where I did research supervised by Wynne Hsu and Mong Li Lee in automated fact-checking and detection of misinformation in social media, at Plural AI where I worked on entity extraction from webpages and semantic search for finance, and at Donaco where I worked on contextual recommendation of charity campaigns.
My goal is to help overcome the limitations of current machine learning models (efficiency, robustness, limited training resources), and improve multimodal content organization & discovery. My recent work is in conversational search, representation learning and ranking for information retrieval, contextual recommendation systems and knowledge base construction from semi-structured text.
Research Interests: Information Retrieval, NLP, Machine Learning, Recommendation Systems, Knowledge Graphs
Recent News
- [28/09/23] I presented a poster on end-to-end TOD at the Task Focused IR in the Era of Generative AI workshop
- [15/02/23] I'm leading our team at UMass in Alexa Prize Taskbot Challenge 2
- [24/06/22] I was awarded a SIGIR travel grant to attend ICTIR 2022 in Madrid!
- [10/06/22] Our paper on query facet extraction & generation was accepted to ICTIR 2022
- [31/05/22] I will be a Research Intern in NLP at Adobe Research for Summer 2022
- [23/07/21] I got a scholarship for my PhD studies from Bodossaki Foundation!
- [18/03/21] Our demo paper on evidence retrieval for fact-checking was accepted to NAACL 2021
- [12/02/21] In September 2021, I will join the Center for Intelligent Information Retrieval at University of Massachusetts Amherst, the top department in Information Retrieval, for a MSc/PhD in Computer Science
- [03/09/20] Our long paper on latent retrieval for fact-checking and QA was accepted to IEEE ICTAI 2020
- [31/08/20] NUS President Professor Tan Eng Chye launched our LetsCheck platform
Recent Publications
-
MarunaBot V2: Towards End-to-End Multi-Modal Task-Oriented Dialogue Systems
Chris Samarinas, Pracha Promthaw, Rohan Lekhwani, Sheshera Mysore, Sung Ming Huang, Atharva Nijasure, Hansi Zeng, Hamed Zamani
Alexa Prize Taskbot Challenge 2022 -
Revisiting Open Domain Query Facet Extraction and Generation
Chris Samarinas, Arkin Dharawat, Hamed Zamani
ICTIR 2022 -
MarunaBot: An Extensible Retrieval-Focused Framework for Task-Oriented Dialogues
Amit Hattimare, Chris Samarinas, Arkin Dharawat, Yelman Khan, Yen-Chieh Lien, George Wei, Yulin Yang, Hamed Zamani
Alexa Prize Taskbot Challenge 2021 -
Improving Evidence Eetrieval for Automated Explainable Fact-checking
Chris Samarinas, Wynne Hsu, Mong Li Lee
NAACL 2021 -
Latent Retrieval for Large-Scale Fact-Checking and Question Answering with NLI training
Chris Samarinas, Wynne Hsu, Mong Li Lee
ICTAI 2020
Projects
- Faspect, A library for open domain query facet extraction and generation
- Quin, An easy to use framework for large-scale fact-checking and question answering
- DeCodR, Model implementations for Semantic Code & Repository Search
- ReLM, Interactive Interface for Web-based Recommendations using LLMs
- Laby, Maze generator, editor & solver in Java (old project)
- mRand, Random YouTube music player (old project)
Awards
- SIGIR Travel Grant, June 2022, for attentding ICTIR in Madrid
- Bodossaki Foundation PhD Scholarship, July 2021, 5-year scholarship for PhD studies in Information Retrieval
- Bioengineering Hackathon, 1st place, EMB IEEE Student Chapter, April 2018, 1st place for developing DropSense in 48 hours, a platform that monitors and connects elderly with their relatives and paid carers using Wear OS smartwatches
- Great Moment for Education Scholarship, Eurobank Ergasias, 2012, scholarship for achieving the highest grade at school at the national university admission exam
- National Mathematics Competition, First Prize, Hellenic Mathematical Society, 2009, awarded for competing in the final stage ‘Archimedes’, top 50 out of 20,000 participants
Recent Experience
Adobe Research · Research Intern June 2022 - Aug 2022
- Research in multi-step query reformulation for dense retrieval at the Real-time Experiences and Algorithms Lab (REAL)
Ro5 · Machine Learning Consultant Feb 2021 - Aug 2021
- Biomedical relation extraction from research papers
National University of Singapore · Research Assistant Feb 2020 - Apr 2021
- Research on latent evidence retrieval for large-scale fact-checking and question answering, published at ICTAI & NAACL
- Research on fake news detection in social media using social context
Plural AI · Research Engineer Oct 2018 - Jan 2020
- Developed a novel multi-stage method for extraction and saliency scoring of concepts from corporate websites
- Owned the whole ML pipeline, from data annotation and website crawling, to productionization and deployment
- Developed a smart tool for efficient curation of an industrial category graph
- Developed a scalable semantic search service of companies using a novel topic modeling technique
Donaco · Research Engineer Aug 2018 - Nov 2018
- Designed a system for contextual recommendation of charity campaigns in news articles
- Worked on the advertising platform and tried to utilize the real-time user behavior for better ad serving
Foshan.io · Software Consultant Jun 2018 - Jan 2019
- Responsible for the development of the new version of the historical HotBot metasearch engine
Service
- Reviewing: ACM Transactions on Information Systems (TOIS)
- Mentoring:
- 2022 Early Research Scholars Program (ERSP) at UMass
- PhD Applicant Support Program (PASP) at UMass
Misc
- My first attempts in astrophotography 🔭 with an 127mm Mak on EQ mount: Whirpool Galaxy, Moon
- Some of my old attempts in drone 🚁 photography in Greece on YouTube
Last update: 16/02/24