SEUPD@CLEF: Team hextech on Argument Retrieval for Comparative Questions. The importance of adjectives in documents quality evaluation

This report explains our approach to solve the Task 2 challenge about Argument Retrieval for Comparative Questions proposed by the third Touché lab on argument retrieval at CLEF 2022. Given a comparative topic, the task is to retrieve relevant argumentative passages from a collection of documents for either compared object or for both. Our approach follows the Information Retrieval pipeline that is: the parsing of the document collection, the indexing of the parsed documents by using an analyzer with commons filters, and the query matching using a retrieval model. In addition, we implemented an index field that catches the overall quality of each passage for our specific task. From an analysis of our results, the implementation of the quality field slightly improves the ranking of the retrieved documents.