[2Win5] Poster session 2

[2Win5-95] A Note on Similar Research Project Retrieval via Deep Metric Learning to Support Selection of Co-researchers

〇Naoki Saito¹, Yusuke Mizutani¹, Takahiro Ogawa¹, Miki Haseyama¹ (1.Hokkaido University)

Keywords:Similar Research Project Retrieval, Deep Metric Learning, Selection of Co-researchers

This paper presents a method for retrieving similar research projects to support co-researcher selection. By retrieving and presenting research projects relevant to solving a company's issues, the company can use this information to aid in selecting a suitable co-researcher. The proposed method utilizes deep metric learning to construct a feature space where the features derived from the summary texts of research projects associated with the same researcher are highly coherent and discriminative. Additionally, research projects similar to the query's summary text features are presented as retrieval results in the constructed feature space. By referring to the researchers involved in past projects retrieved through this method, companies are expected to select co-researchers to address their challenges efficiently. Experimental results show that the proposed method effectively retrieves similar research projects for co-researcher selection.