FIRE 2023

The proliferation of the Internet in South Asia had led to a rapidly growing mass of available information in Indian languages in early 2000s. The challenge to retrieve information from online articles written in several Indian languages was significant. A well rounded solution was required that could address the diversity of the character encodings, morphology, spelling variations of these languages to support efficient query translation and web search.

A cross lingual evaluation forum to conduct text retrieval research using large collections had been started in the TREC (Text REtrieval Conference) in 1992 but the conference primarily targeted European languages. Such a TREC-like evaluation forum was missing to encourage Information Retrieval (IR) research for Indian languages.

The Forum of Information Retrieval (FIRE) was conceptualized and started by the Information Retrieval Research Group of Indian Statistical Institute, Kolkata in 2008 to provide a platform to meet the new challenges in multilingual information access and retrieval, mostly for the languages in the Indian Subcontinent.

Over the past 15 years, FIRE has offered diverse tracks that addressed sentiment analysis, event detection, fake news analysis, and anaphora in English and Indian languages. Targeting a unique connection between industry and academia, FIRE hosts tracks from Microsoft, Yahoo, Rediff, TCS and others, which cater to specific industry-related retrieval challenges.

Several institutes like Dhirubhai Ambani Institute of Information and Communication Technology, Gandhinagar; Indian Institute of Science, Bangalore; Indian Institute of Technology, Bombay; Indian Institute of Technology, Kharagpur joined the journey and collaborated with FIRE to encourage IR research in India.

The connection that FIRE has built over the years between industry and academia has provided enormous learning opportunities for numerous researchers to have hands-on experience with real-time scalable information retrieval challenges and their solutions. This conference has provided a platform to build a global, well-connected, committed and supportive community and provide accessibility to understand and discuss the research ideas with distinguished researchers in person.

FIRE constantly has thrived to increase communication and collaboration between the industry-borne challenges of information retrieval and the possible set of prototypes developed in academia for Indian languages, primarily. This has led to the formation of large sized datasets that are being significantly downloaded and used by researchers from across the world and scalable solutions and substantial improvements in retrieval technologies.

The trends and additions of various new tracks for different new challenges are a testimonial to this fact.

In its 15th milestone year, FIRE brings in a renewed industry track and a doctoral colloquium with the aim of engaging industry and academia bilaterally with ideas, mentorship and delivery. The doctoral consortium aims to provide a platform for PhD students to share their ideas with distinguished researchers across academia and industry that can help them shape their PhD problem into a scalable product.