About ICTIR
-----------
The ACM SIGIR International Conference on the Theory of Information Retrieval (ICTIR) is the premier conference of theoretical information retrieval (IR). The conference welcomes papers on core IR and any paper on connections between IR and its neighboring disciplines. We welcome experimental and industrial papers that validate approaches from the lens of practical applicability. Such papers should, apart from validating the practical applicability of an existing approach, provide a clear message to the community as to which aspects need further (theoretical) investigation, based on the experimental findings. ICTIR also encourages papers attempting to define new tasks or apply fundamentally different ideas to information retrieval tasks.


Page Length
-----------
To facilitate exchange of ideas on the theory of IR, this year we allow submissions of any length between 2 and 9 pages, plus unrestricted space for references. Page length should be commensurate with contribution size. We do not differentiate between short papers and long papers during submissions. We may allocate more presentation time to longer papers during the conference.


Submission Guidelines
---------------------
All submissions must be in English, in PDF format, and in ACM two-column format. Suitable LaTeX and Word templates for ICTIR are available from the ACM website. The algorithms, resources, and methods used within a paper should be described as completely as possible. Authors may upload supplementary material, including detailed descriptions, test datasets, or code; however, this information is not part of the submission paper, and will be read only at the discretion of the reviewers. Papers must be submitted electronically via the conference submission system. Authors should carefully go through ACM’s authorship policy before submitting a paper.


Reviewing Process
-----------------
The review process for all submissions is double-blind, and authors must take all reasonable steps to conceal their identity. For this reason, ICTIR follows the ACM SIGIR policy of strongly discouraging uploading papers onto arXiv.org until after final notification.

Accepted papers must be relevant to the theme of ICTIR. Results will be judged on the degree to which they have advanced the theoretical understanding of IR or how much they have offered generalizable observations in IR. Evaluation expectations for submissions are less demanding than in other IR conferences, but some degree of evaluation is still expected and valued unless it is a purely theoretical work (e.g., a proof of an important theorem). The relevant literature should be properly cited and reviewed.

Submissions will be rejected without review if they are over the specified page limit, if they violate the formatting guidelines, if they are not anonymous, or if they are incomplete.


Dual Submission
---------------
All submissions must describe work that is not previously published, not accepted for publication elsewhere, and not currently under review elsewhere.


Topics
------
IR theory and practice
- New retrieval problems
- Retrieval algorithm behavior analysis
- Recommendation algorithm behavior analysis
- Understanding IR models
- Quantum IR or other non-conventional retrieval models and ranking
- New evaluation measures, mathematical analysis of measures
- New evaluation paradigms

User aspects of IR
- Interactive and dynamic search
- Information seeking
- Modelling and simulation of information interaction
- Contextual and personalized search
- Search interfaces
- Understanding context information
- Understanding retrieval results
- User modeling for ranking and evaluation

IR system architectures and scalability
- Understanding IR system architecture
- Efficient content representation
- Compression for IR
- String processing for IR
- IR system scalability

Content representation and processing
- Natural language processing for IR
- Multimodal IR
- Multilingual IR
- Linguistic analysis
- Text representation
- Topic models

Machine learning for IR
- Deep learning
- Kernel methods
- Markov decision processes
- Online learning
- Optimization
- Hyper-parameter and model selection
- Reinforcement learning

Applications of IR
- Information extraction
- Information fusion
- Legal IR
- Patent search
- Search applied to the Internet of Things
- Semantic search
- Sentiment analysis
- Social good and collaborative search
- Voice and image search
- Structured, semi-structured data and databases
- Task-based and exploratory search
- Web, digital libraries and social media search