BIP! Ranker

BIP! Ranker is a software library that computes citation-based impact indicators on scholarly knowledge graphs and related bibliometric inputs. Most indicators are implemented in Apache Spark and operate on citation networks of research products (publications). The library also supports field-normalized publication metrics (FWCI and topic-based impact classes) and artefact-oriented indicators for research software and datasets.

Supported indicators, grouped by the impact aspect they capture:

Influence indicators (total impact; how established a research product is in general)

• Citation Count: The total number of citations of the product — the most widely used influence indicator.
• PageRank score: An influence indicator based on PageRank [1]. PageRank estimates the influence of each product from its centrality in the citation network. It alleviates some limitations of Citation Count (e.g., two products with the same number of citations can have very different PageRank scores if the influence of their citing products differs).

Popularity indicators (current impact; how popular a product is now)

• RAM score: A popularity indicator based on RAM [2]. It is essentially a Citation Count where recent citations are weighted more heavily, reducing bias against recently published products compared with methods like PageRank.
• AttRank score: A popularity indicator based on AttRank [3]. AttRank alleviates PageRank’s bias against recent work by using an attention-based mechanism, akin to a time-restricted form of preferential attachment, to capture preference for products that recently received attention.

Impulse indicators (initial momentum shortly after publication)

• Incubation Citation Count (3-year CC): A time-restricted Citation Count where only citations within a fixed window (e.g., 3 years) after each product’s publication date are counted.

Field-normalized publication indicators (impact relative to the same research topic/field)

• Field-Weighted Citation Impact (FWCI): Normalizes citation counts as the ratio of actual to expected citations within the same research concept (field), publication type, and year.
• 3-year FWCI (3y-FWCI): The same field-normalized measure applied to 3-year citation counts.

Artefact impact indicators (impact of research software and datasets)

• Indirect Citation Count (ICC): Sums citation counts of publications associated with an artifact (e.g., via linked DOIs), counting each underlying publication only once per artifact.
• In-text Mentions (IM): Counts how many distinct publications mention the artifact in text.
• Artefact Composite Impact Indicator (ACII): Experimental composite score that aggregates normalized ICC and IM through a configurable weighted scheme (default: equal weights).

More details about the aforementioned impact indicators, the way they are calculated and their interpretation can be found here and in the respective references (e.g., in [4]).

You can find more details and full documentation in our GitHub repository page: https://github.com/athenarc/Bip-Ranker

References:

1. R. Motwani L. Page, S. Brin and T. Winograd. 1999. The PageRank Citation Ranking: Bringing Order to the Web. Technical Report. Stanford InfoLab.
2. Rumi Ghosh, Tsung-Ting Kuo, Chun-Nan Hsu, Shou-De Lin, and Kristina Lerman. 2011. Time-Aware Ranking in Dynamic Citation Networks. In Data Mining Workshops (ICDMW). 373–380
3. I. Kanellos, T. Vergoulis, D. Sacharidis, T. Dalamagas, Y. Vassiliou: Ranking Papers by their Short-Term Scientific Impact. CoRR abs/2006.00951 (2020)
4. I. Kanellos, T. Vergoulis, D. Sacharidis, T. Dalamagas, Y. Vassiliou: Impact-Based Ranking of Scientific Publications: A Survey and Experimental Evaluation. TKDE 2019 (early access)