FIRE 2026

The 18th meeting of the Forum for Information Retrieval Evaluation 2026 will be held at the Indian Statistical Institute Kolkata, India. It will be an in-person conference. We are seeking submissions of high-quality and original papers. Submissions will be reviewed by experts on the basis of the originality of the work, the validity of the results, chosen methodology, writing quality and the overall contribution to the field of IR/NLP. Authors are also encouraged to submit work in progress and late-breaking research results.

A Best Paper Award and a Best Paper Runner-up Award will be presented during the conference.

Authors of selected accepted papers will be invited to submit extended versions of their work for review and possible publication in the Information Retrieval Research Journal (IRRJ) after the conference.

Topics of interest include, but are not limited to

1. Search and Ranking: Research on core algorithmic topics in IR:
• Queries and query analysis (e.g., Query understanding, query reformulation, query representation etc.)
• Retrieval models and ranking (e.g., Cross lingual IR with a particular focus on Indian languages, ranking algorithms, language models, retrieval algorithms, learning to rank etc.)
• Efficiency and scalability (e.g., distributed search, search engine architecture, indexing, crawling etc.)
• Supervised/Weakly supervised deep neural networks.
• Other domain-specific applications of IR.
2. Evaluation: Research on evaluation of IR systems:
• User centric evaluation (e.g., User experience, user engagement etc.)
• System centric evaluation (e.g., Evaluation metrics).
• Query Performance Prediction and its applications.
3. Generative Models for IR/NLP.
• Conversational and Interactive Search Systems.
• Simulated Data for Personalized IR.
• In-Context Learning or Retrieval Augmented Generation (RAG) for Search and NLP downstream tasks, such as Question Answering, Summarization.
• Domain-specific generation, such as Code generation, Argument generation, Workflow generation.
4. Explainability, Fairness and Trust of IR/Recsys models
• Explainable models for ranking, text classification/clustering, summarization etc.
• User studies for explainable AI (XAI) applied to IR/Recsys.
• Issues related to fairness and trustworthiness of IR/Recsys models.
5. Multimodal and Crossmodal IR/Recsys model
• Visual and Multimodal Question Answering
• Image Search/Recommendation
• Multimodal Document Summarization
6. New and emerging topics:
• IR for Agentic Systems
• Memory Retrieval

Important dates 

Note: All submission deadlines are 11:59 PM AoE Time Zone (Anywhere on Earth).

Submission Guidelines

Submissions must describe substantial, original, and unpublished work. Wherever appropriate, concrete evaluation and analysis must be included.

The FIRE conference track is divided into two sub-tracks: Regular Papers, and Resource and Demo Papers. Each has its own distinct scope, objective, and reviewing policy (described below). Authors should ensure that their submission is made to the appropriate sub-track. Requests to transfer submissions between sub-tracks after the paper submission deadline will not be considered. Submissions that are assigned to an incorrect sub-track may be desk-rejected.

If the paper being submitted is under review at any other venue, this must be explicitly declared during submission. If accepted at FIRE, the paper must be withdrawn from all other venues.

Authors should provide sufficient information to enable verification of their results, including source code (which must be anonymized for double-blind submissions), datasets (where permissible), and a clear description of the experimental setup. Authors are also encouraged to include a brief reproducibility statement.

Authors must comply with the ACM Policy regarding the use of generative AI tools. Any use of AI-generated content or AI-assisted writing should be disclosed in the paper.

Authors are strongly encouraged to provide “alt text” (alternative text) for floats (images, tables, etc.) to make their works more inclusive and accessible, especially to the differently-abled. Authors may consult Writing Figure Descriptions, SIG ACCESS guidelines on describing figures, and ACM Accessibility Recommendations for Publishing in Color for guidance.

Paper Template and Submission

The submitted papers must follow the ACM ICPS template available on https://authors.acm.org/proceedings/production-information/overleaf (2-column “sigconf” LaTeX template).

LaTeX users should use the following:
Anonymous submission:
\documentclass[sigconf,natbib=true,anonymous=true]{acmart}
Non-anonymous submission:
\documentclass[sigconf,natbib=true]{acmart}

Please consult the LaTeX documentation and ACM’s LaTeX best practices guide for further instructions, ignoring the single-column instructions. Do not use the “manuscript” option, otherwise the document will not be compiled in double-column, as required. To ensure compatibility with The ACM Publishing System (TAPS), please restrict the use of packages to the whitelist of approved LaTeX packages.

The only accepted format of submissions is PDF. Please note that it is the responsibility of the authors to ensure that the PDF submission has been uploaded successfully (we suggest that you download the uploaded PDF to verify that it was submitted correctly). ACM’s CCS concepts and keywords must be included in the paper. When submitting a paper, select Conference Track and choose the appropriate paper type (Regular Paper or Resource and Demo Paper) under the Subject Area. Papers which do not conform to the requirements may get rejected without review. Authors are invited to submit in any of the following tracks:

• Regular paper

We welcome high-impact original papers in the area of IR/NLP. Submissions will be evaluated on the basis of originality, technical quality, significance, relevance, and clarity. As in the previous year, we are not making any explicit distinction between long and short papers. However, longer papers will be allocated proportionally more presentation time.

Reviewing policy: Double-blind. 

Number of pages: Variable length up to a maximum of 9 pages of content (excluding references). 

• Resource and Demo paper

Papers submitted to this track should describe datasets, software tools, benchmarks, and other research artifacts that address a research problem and are expected to benefit the IR/NLP/AI community. Ideally, such resources should be made publicly available for reviewers to judge their quality and usefulness. The demo papers should contain a link to a working software system that demonstrates the application of existing or proposed research methods as a proof-of-concept.

Reviewing policy: Single-blind.

Number of pages: Variable length up to a maximum of 9 pages of content (excluding references).

Double-blind Reviewing Policy

For papers submitted to the Regular Paper track, authors must not include any identifying information anywhere in the manuscript, including author names, affiliations, acknowledgments, grant information, project URLs that reveal their identity, or citations to prior work that disclose the authors’ identity. Authors should refer to themselves in third person when citing their own work. Expressions like "In our earlier work..." or "We previously showed that..." must be avoided.

Presentation Requirements

If accepted, at least one author will have to register for the conference and present their work in-person.

Conference Track Co-ordinator

• Debasis Ganguly (University of Glasgow, UK)
• Debarshi Kumar Sanyal (Indian Association for the Cultivation of Science, Kolkata, India)
• Surupendu Prakash Gangopadhyay (Interactions LLC)

For queries related to conference please email us at [ clia@isical.ac.in ]
For latest updates subscribe the FIRE mailing List [ https://groups.google.com/forum/#!forum/fire-list ]