Identify research clusters

About We developed a four-phase TPSE-RCI framework to identify scientific research clusters: (1) Research Team Detection and Collaboration Network Construction, (2) Structural Representation Learning and Initial Cluster Identification, (3) Research Goal Extraction and Hierarchical Structuring, (4) Valid Research Cluster Identification. Citation article:A novel scientific research cluster identification method integrating topological potential and semantic embedding Explanations and Issues The release of this dataset is for research purposes only and should not be used for any inappropriate applications. Certain data included herein are derived from Clarivate™ (Web of Science™). © Clarivate 2025. All rights reserved. Dataset Details 1. WOS.json We constructed a structured dataset integrating bibliographic and author-level information from the Web of Science (WoS) database. Each record corresponds to a unique publication and includes identifiers, textual content, and author metadata. The key fields are summarized as follows: Field Name Type Description ID Integer Unique identifier for each record in the dataset paper_id String Internal identifier for each publication WOS_ID String Unique Web of Science identifier for the publication publication_year Integer Year in which the article was published article_title String Title of the publication abstract Text Abstract text describing the research content author_keywords String Author-provided keywords DOI String Digital Object Identifier for the publication author_id String Unique identifier assigned to each author author_full_name String Full name of the author as recorded in WoS ORCIDs String ORCID identifiers associated with the author Researcher_Ids String ResearcherID(s) identifiers associated with the author addresses String Author affiliation addresses 2. cluster_info.csv This dataset records the identified research clusters and their associated structural information, including participating teams, authors, and research goals. Each row represents a single research cluster and its corresponding attributes. The dataset contains the following fields: Field Name Type Description ID Integer Unique identifier of the record. cluster_id Text Unique identifier of the research cluster. valid Boolean (T/F) Indicates whether the cluster satisfies the defined validity criteria. team_list List of Text List of teams involved in the cluster. Each element represents a unique team identifier. author_list List of Text List of authors participating in the cluster. Authors may appear multiple times if they are associated with multiple teams. goal_list List of Text List of research goals associated with the cluster. Each goal represents a specific research direction within the cluster.