Utility-University Collaboration Publication Data

This dataset is a collection of metadata describing the authors, their organizational affiliations, and locations associated with academic publications that result from collaborations between academic researchers and electric utilities. It is queried from the Scopus database by searching for publications where at least one author is affiliated with one of the 20 largest U.S. electric utilities.

We used this data set to better understand the nature of and factors in utility-university collaboration formation. In addition to understanding the role geography/proximity plays, we also conducted limited network analysis to identify high frequency collaborators at both the author and organizational scale. We identified some time series trends such as increasing numbers of publications and increasing distances between collaborators over time, but we did not determine their significance by controlling for external factors like funding, regulation, and technological changes. Future work could use the included classifications for each publication to understand the changing mix of research topics over time.

The interviews we conducted for the accompanying research suggest that several types of collaborations are not represented in the publication dataset, including unsuccessful collaborations, many types of student-driven practicum-style work, and for-hire work that may assist in regulatory filings, internal documents, or other non-academic publications.

We include four separate versions of this dataset at different stages of its refinement to better enable any reproductions, expansions or refinements of the dataset.

The first file (Initial Publication Queries By Utility.zip) is our raw output from the Scopus queries.

The second file (Author-Parsed Publication Queries By Utilities.zip) is the parsed output of the queries, where each author and affiliation are separated.

The third file (Publication Dataset with Duplicates and Erroneous Entries.csv) combines all utilities into a single file and includes many manual corrections to parsed or missing information, as well as some additional fields to classify data and identify duplicates and records erroneously included.

The fourth file (Final Utility-University Publication Dataset.csv) then removes some of those additional fields as well as all duplicates and erroneously included records. This was the file we used for our final analyses.

本数据集为元数据（metadata）合集，收录了学术研究者与电力公用事业企业开展合作所产出的学术出版物的关联元数据，内容涵盖作者、所属机构及所在地点信息。该数据集通过检索Scopus数据库获取，筛选条件为至少有一位作者隶属于美国20家规模最大的电力公用事业企业之一。
我们依托本数据集，旨在更深入地剖析电力企业与高校合作的形成本质及其影响因素。除探究地理区位/空间邻近性所起到的作用外，我们还开展了有限的网络分析，以识别作者层面与机构层面的高频合作对象。我们观测到若干时间序列趋势：随时间推移，出版物数量持续增加，合作双方的空间距离不断拉大，但并未通过控制资助、监管政策、技术变革等外部变量，验证这些趋势的统计学显著性。后续研究可借助每条出版物附带的分类标签，解析研究主题构成随时间的演变规律。
我们为配套研究开展的访谈结果显示，本出版物数据集未覆盖若干类别的合作场景，其中包括未成功的合作、多种学生主导的实践类工作，以及可用于监管申报、内部文档编制或其他非学术出版物的付费委托工作。
为更好地支持该数据集的复现、扩展与进一步细化工作，我们提供了四个处于不同细化阶段的独立数据集版本。
第一个文件（"Initial Publication Queries By Utility.zip"）为我们通过Scopus检索得到的原始输出结果。
第二个文件（"Author-Parsed Publication Queries By Utilities.zip"）为经解析后的检索结果，其中每位作者及其所属机构均已完成分离。
第三个文件（"Publication Dataset with Duplicates and Erroneous Entries.csv"）将所有电力企业的数据整合为单个文件，包含对解析后信息或缺失信息的多处人工修正，同时增设了若干用于数据分类、重复项识别以及错误记录标记的额外字段。
第四个文件（"Final Utility-University Publication Dataset.csv"）则移除了部分额外字段，并清理了所有重复项与错误收录的记录，该文件亦是我们开展最终分析时所使用的数据集。