Yaqeen
Arabic research produces — it does not accumulate.
Yaqeen is a semantic-representation layer for peer-reviewed Arabic knowledge: it makes what is written in Arabic machine-perceptible at the level of meaning, so search knows what came before, and knowledge builds on knowledge.
What Yaqeen is
Representation, not matching
Yaqeen turns Arabic text from a string of characters into a mathematical representation that carries its meaning. That transformation is the foundation of everything that follows: once meaning is measurable, comparison, retrieval, and classification become possible at the level of sense rather than wording.
The architecture rests on three pillars, each built on what precedes it:
Semantic representation
Processing grounded in the morphological and semantic properties of Arabic, producing for every text a representation of its meaning — not its words — so two texts that agree in meaning converge even when their wording differs entirely.
Indexing
Building a semantic index of peer-reviewed Arabic scholarly output that grows by addition without rebuilding, and is not constrained by the size of the archive.
Retrieval and comparison
Meaning-based retrieval: present a text or a question, and receive what approximates it in the index, ranked by proximity, with the loci of affinity and their sources identified.
What it enables
The semantic architecture is a foundation, and foundations are built once and invested across directions. Four directions:
- 01
Discovering what came before and what follows
Enabling the researcher and the peer reviewer to reach what has been written on the question in Arabic — not by matching keywords, but by approximating the research question itself.
- 02
Verifying originality
Detecting semantic proximity between a submitted work and what preceded it, including transfer after paraphrasing and synonym substitution. A report that specifies locations, sources, and degrees — supporting the editorial board’s decision without replacing it.
- 03
Supporting peer review
Nominating reviewers on the basis of semantic proximity between the submitted paper and their published output, rather than relying on declared specialisation alone.
- 04
Maps of knowledge production
Seeing the structure of an institution’s research production as it is: points of concentration, points of absence, and lines of affinity among its units and tracks.
Where the architecture sits
An Arabic knowledge index is an asset you own, not a service you rent
The ability to perceive Arabic knowledge computationally is not a service bought by subscription; it is infrastructure. And knowledge infrastructure is kept where knowledge is produced.
Yaqeen is built to be deployed inside the institution or within its sovereign scope — not hosted on its behalf. Research remains in its owner’s custody, the index remains one of its assets, and the capability over it remains un-rented.
Who it is for
- Universities and research centres
- Scholarly publishers and peer-reviewed journals
- Scientific societies and academies
- Bodies that govern scientific research
- Research-integrity units
Start from your archive
Work usually begins with the institution’s own archive: a semantic index is built for its peer-reviewed output, and the first results appear on its own knowledge. The scope then expands according to purpose.