The Problem The academic quality assessment system is reaching a breaking point. Current metrics — citation counts, h-index, journal impact factors, and quartile rankings (Q1–Q4) — measure the social influence and network position of a piece of research rather than its intellectual value. This crisis, formally articulated in 2015 by Hicks, Wouters, Waltman, de Rijcke, and Rafols in the Leiden Manifesto published in Nature, has deepened in the age of artificial intelligence. The fact that high-volume paper production can now be accomplished in minutes has rendered the system's foundational "publish" metric meaningless. The social sciences and humanities have been the most severely affected by this collapse: narrow citation networks, long publication cycles, the invisibility of multilingual scholarship, and the systematic penalisation of interdisciplinary work prevent quality scholarship in these fields from being recognised.
The Solution ARISE (Academic Research Intelligence & Scholarly Evaluation) is a three-stage assessment system that does not abolish the classical peer review process but redesigns it to meet the demands of the age. The process works as follows: in the first stage, an AI model trained on the full academic corpus independently evaluates the submitted work according to five predefined quality dimensions and produces a report. In the second stage, a human reviewer using the same dimensions reviews the work blind to the AI's evaluation and produces their own report. In the third stage, both reports — with author identity withheld — are forwarded to an independent third-party expert who synthesises the two evaluations and determines the final ARISE score.
Five-Dimensional Scoring Rather than reducing the assessment to a single number, the system scores each work along five dimensions: Originality, Intellectual Rigour, Interdisciplinary Impact, Openness and Reproducibility, and Humanistic Dimension. This multidimensional structure both increases the transparency of evaluation and ensures that works from different disciplines are assessed in ways appropriate to their nature.
What Makes It Different Rather than inventing a new concept, ARISE restructures peer review without severing it from its historical roots. Its difference from fully automated AI evaluation tools is that human judgement remains at the centre of the system; its difference from traditional peer review lies in improved consistency, speed, transparency, and the elimination of blindness toward interdisciplinary work. The system evaluates the intellectual content of a work directly rather than relying on proxy indicators such as citation counts or journal prestige.
