Ehsan Kamalloo
My name is Ehsan (pronounced /eh☀️/ or /ih☀️/, or even /e☀️/). I am a Research Scientist in the Conversational Assistant team at ServiceNow Resaerch.
I am broadly interested in the intersection of natural language processing and information retrieval. My goal is to facilitate information access over massive unstructured materials via:
- Building scalable and robust retrieval systems that are suitable for deployment “in the wild”,
- Data-centric strategies to systematically improve data quality and to overcome bias in data.
Prior to this, I was a Post-doctoral Fellow at the David R. Cheriton School of Computer Science, University of Waterloo, hosted by Jimmy Lin. I completed my PhD at the University of Alberta, advised by Davood Rafiei. I was also fortunate enough to collaborate with Charles Clarke and Siva Reddy. I also interned as a researcher at Huawei Noah’s Ark Lab in Montreal and worked with Mehdi Rezagholizadeh. Before PhD, I was working as a software engineer.
news
| May 2024 |  New paper: “LLMs Can Patch Up Missing Relevance Judgments in Evaluation” [arXiv] |
|---|---|
| Mar 2024 |  3 papers at SIGIR 2024: 2 resource (BEIR [Paper] and Touché 2020 error analysis) and 1 demo (QA evaluation using LLMs [Paper]). |
| Jan 2024 | Joined ServiceNow Research as a Research Scientist in the Conversational Assistant team. |
| Dec 2023 | New paper: “NoMIRACL: Knowing When You Don’t Know for Robust Multilingual Retrieval-Augmented Generation” [arXiv] |
| Jul 2023 | HAGRID  , a new dataset for generative search with citations, released. [Paper] [Dataset] |
| May 2023 | MIRACL 🌍🙌🌏 got accepted to TACL. |
| May 2023 | Our paper “Evaluating Embedding APIs for Information Retrieval” got accepted at ACL 2023 Industry Track. [Paper] |
| May 2023 | Our paper “Evaluating Open-Domain Question Answering in the Era of Large Language Models” got accepted at ACL 2023 (oral). [Paper] [Code] |