User:M137K/sandbox

Introduction

Connected Papers is a Web-based visual exploration tool which allows users to find, organise and share papers about their research interests. With Connected Papers, One can quickly find similar and related papers by giving one “seed paper” (relevant paper of the user’s choice) and can obtain a visual graph of all papers. The obtained visual graph is a graph connecting the papers based on a well-defined similarity metric based on co-citations and bibliographic coupling. One can also create “multi-origin graphs” by adding other seed papers one at a time to refine the graph.

Graphs in Connected Papers

To generate a graph, the user has to enter one of the following: a search string, a paper title, a DOI, or the URL of an ArXiv, a Semantic scholar, or a PubMed paper. When a query is entered, Connected Papers compiles a list of the most pertinent papers based on co-citation. The most desirable document among them will have to be chosen by the user; this paper is referred to as the origin paper or seed paper. Connected Papers analyses approximately ~50,000 publications to generate one graph and chooses a few dozen with the strongest association with the seed paper. In the graph, The papers are arranged in the graph according to their similarity. The similarity is defined by a similarity metric based on the concepts of Co-citation and Bibliographic Coupling. This statistic assumes that two papers with similar citation and reference counts are more likely to address comparable topics. After finding relevant papers, The algorithm creates a Force Directed Graph to organise the papers in a manner in which similar papers are visually clustered and less similar papers are kept away from each other. Each node represents a related publication. When a node is selected, It highlights the shortest path from the selected node to the origin paper in similarity space. The larger size of the node represents a paper which has higher number of citations and vice versa. The colour of the node represents the year of publication in a way that darker node represents more recent works.

Usefulness

The Literature Review is of great importance in research. Finding and organising different papers pertinent to a person's area of interest in research can often take a lot of time. It can be tedious to go through the reference list or look up relevant keywords in databases and textual search engines in order to locate papers that are similar. Various tools are available to speed up and simplify the literature review process. The Connected Papers is one such tool which can help a researcher in his/her literature review by picking similar papers and arranging them visually.  In addition, Connected Papers offers the following features: Prior Work and Derivative Work. The Prior Work feature makes it easier to locate essential studies that significantly impacted later research generations. Whereas The Derivative Work feature helps to find systematic reviews and meta-analyses in the respective field. There is also an option to download the bibliographic references as a file in BibTeX format. The downloaded file can be read by importing it into any reference management software.

The database of Connected Papers is connected to the Semantic Scholar Paper Corpus (Licensed under ODC-BY). Thus, Its effectiveness relies upon the comprehensiveness and diversity of the publications within the Semantic Scholar Paper Corpus. The tool's output might not give a comprehensive or impartial picture if some fields, subfields, or regions are underrepresented in the corpus. Being a relatively new platform still in an ongoing development phase, Connected Papers might not yet offer all the features and functionalities that certain researchers need. This may limit its usefulness to specific research works of researchers with specialized requirements.

History

Founded by Alex Tarnavsky Eitan, Eddie Smolyansky, Itay Knaan Harpaz, and Sahar Perets, Connected Papers began as a side project during the weekends. Its founders, with the developer Ofer Mustigman, made it public in 2020. The aim was to figure out how to handle the difficulties involved in conducting a thorough literature review. The basic idea to achieve this aim was to simplify the complexities of extensive literature exploration by selecting a relevant paper as the origin paper, and then Connected Papers discovers other papers connected to the same theme.