Rajarshi1001/IITK-SemEval-2024-Task-1

IITK at SemEval Task 1: Semantic Textual Relatedness for African and Asian Languages IITK at SemEval 2024 Task 1: Semantic Textual Relatedness for African and Asian Languages View Paper | Raise an Issue | Submit Pull Request ## Introduction The semantic relatedness of two language units has long been considered fundamental to understanding meaning (Halliday and Hassan 1976, Miller and Charles 1991), and automatically determining relatedness has many applications such as evaluating sentence representation methods, question answering, and summarization (Abdalla et al. 2023). Two sentences are considered semantically similar when they have a paraphrasal or entailment relation. On the other hand, relatedness is a much broader concept that accounts for all the commonalities between two sentences: whether they are on the same topic, express the same view, originate from the same time period, one elaborates on (or follows from) the other, etc. Consider the two sentences: | Pair | Sentence 1 | Sentence 2 | |------|------------------------------------------|----------------------------------------------------------| | 1 | There was a lemon tree next to the house | The boy liked playing football next under the lemon tree | | 2 | There was a lemon tree next to the house | The boy liked playing football | As for __Relatedness__, sentence pair 1 seems to be more related than pair 2. ## Objective The main aim of this problem statement is to design a system for efficient calculation of relatedness score between two sentences in a pair. The relatedness should be measured in terms of a score in the range of `0` (representing completely **unrelated**) and `1` (representing completely **related**). The entire problem statement has been divided into three stages: - **Track A - Supervised** Developing a relatedness calculation system by training on labelled data provided in 14 different languages to ensure diversity. This involves training a language …