Nurassyl Askar

—

Unclaimed · AI-structured profileORCID OpenAlex

KGQA researcher focused on traceable graph navigation and fidelity-first evaluation of small language models

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-15) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output—

This axis would measure publication volume and consistency, but the available data is still too thin to score: OpenAlex shows an h-index of 1 and an i10-index of 0, with only six recent three-year works and two Evolve-reviewed papers.

Impact—

This axis would capture citation-based influence such as citation counts and h-index trajectory. The record is too young to score: an h-index of 1 and i10-index of 0 mean no reliable impact signal exists yet.

Focus0/100

Focus scores 0/100, the lowest axis, and the raw metrics explain why: the field distribution splits exactly 50/50 between knowledge graphs and KGQA/LLM reasoning, while OpenAlex topics span millimeter-wave propagation and modeling (4), advanced wireless communication technologies (4), topic modeling (2), advanced MIMO systems optimization (2), and semantic web and ontologies (2). The research identity is still forming, with an NLP thread running alongside an earlier wireless communication strand.

Collaboration32/100

Collaboration scores 32/100, a modest level consistent with the raw footprint: only 4 distinct co-authors appear in the recent three-year window. The recurring circle is small and stable — Eduin E. Hernandez, Luis F. Garcia, Sergio A. Diaz, and Stefano Rini. On both Evolve-reviewed 2026 papers the listed position is middle author and not corresponding, indicating a contributing role inside a compact team rather than independent or senior leadership.

Activity57/100

Activity scores 57/100, reflecting a moderate but remarkably even pace: OpenAlex countsByYear records 2 works in each of 2024, 2025, and 2026, totaling 6 works across the recent three-year window. A steady two-per-year cadence suggests consistent research engagement, though the absolute volume remains small and typical of an early-stage record; this measure does not weigh citations.

Reproducibility78/100

Reproducibility is the strongest axis at 78/100, computed as a mean over 2 Evolve-assessed papers. THESEUS in the Graph is rated high reproducibility readiness, since gold navigation paths and fidelity metrics support replication, while The Path Matters is rated medium. Evolve verdicts describe controlled, sound designs, tempered by small benchmarks, missing significance tests, and the exclusion of the closest prior baselines.

Research summary

Nurassyl Askar (ORCID 0009-0000-1865-3059) is an early-stage researcher whose visible record, name-resolved from OpenAlex, spans six works over the last three years — two in each of 2024, 2025, and 2026 — with an h-index of 1 and an i10-index of 0. The profile's center of gravity is knowledge graph question answering: the field distribution splits evenly (50/50) between knowledge graphs and KGQA/LLM reasoning, and the two Evolve-reviewed 2026 arXiv papers define a distinctive agenda. THESEUS in the Graph reframes KGQA as question-conditioned, traceable multi-hop graph navigation with gold paths and fidelity metrics, while The Path Matters argues that small language models should be judged on navigation fidelity — via Path Edit Distance — beyond final-answer accuracy alone. Reproducibility is the strongest dimension at 78/100, averaged over two assessed papers (one rated high, one medium readiness). Collaboration remains modest at 32/100, with 4 distinct co-authors in the recent window, recurring with Eduin E. Hernandez, Luis F. Garcia, Sergio A. Diaz, and Stefano Rini; on both catalog papers the position is middle author and not corresponding, suggesting a contributing role in a compact team. The broader OpenAlex topic mix — millimeter-wave propagation and modeling, advanced wireless technologies, MIMO optimization, topic modeling, and semantic web — indicates prior or parallel work outside NLP, and the 0/100 focus score reflects exactly that breadth.

Collaboration pattern

Collaboration is modest (32/100): OpenAlex records only 4 distinct co-authors in the recent three-year window, forming a small recurring circle of Eduin E. Hernandez, Luis F. Garcia, Sergio A. Diaz, and Stefano Rini. On both Evolve-reviewed 2026 papers the position is middle author and not corresponding, so the current pattern is that of a contributing member of a compact, stable team rather than a group lead.

Research keywords

knowledge graph question answeringmulti-hop reasoninggraph navigationpath fidelityPath Edit Distancesmall language modelstraceabilityreinforcement learning agentsLLM reasoningknowledge graphssemantic web and ontologiesmillimeter-wave propagation

Highlights

THESEUS: KGQA as question-conditioned graph navigation with gold paths
The 2026 arXiv paper reframes knowledge graph question answering as traceable multi-hop graph navigation, supplying gold paths and fidelity metrics alongside reinforcement learning agents; Evolve calls the benchmark solid but notes the closest prior baselines were excluded. Reproducibility readiness: high.
Evaluating small language models beyond answer accuracy
The companion 2026 study shows that ranking SLMs by KG navigation fidelity can diverge from ranking by answer accuracy, introducing Path Edit Distance; Evolve judges the controlled design sound but flags small benchmarks and the absence of significance tests. Reproducibility readiness: medium.
A fidelity-first evaluation agenda taking shape
Across both 2026 papers one thread is consistent: reasoning quality in KGQA should be judged by how models traverse the graph, not only by final answers. The stance rests on a perfectly even two-works-per-year record for 2024-2026 and on the profile's strongest score, reproducibility at 78/100.

Timeline

2024First visible OpenAlex activity: 2 works recorded, marking the start of the publication record.
2025Output holds at 2 works, keeping the three-year cadence perfectly even.
2026THESEUS in the Graph appears on arXiv, framing KGQA as traceable multi-hop graph navigation; Evolve rates its reproducibility readiness high.
2026The Path Matters follows, using Path Edit Distance to evaluate small language models beyond answer accuracy; reproducibility readiness medium.

Papers in the Evolve catalog

Frequent co-authors

Eduin E. Hernandez×2 Luis F. Garcia×2 Sergio A. Diaz×2 Stefano Rini×2

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

Bibliometrics are drawn from OpenAlex only; institution and country are not recorded in the source profile.
Output and impact axes are unscored: with an h-index of 1 and an i10-index of 0, the citation record is too young for reliable estimates.
Only 2 papers have been Evolve-reviewed; the OpenAlex topic mix (millimeter-wave, wireless, MIMO) suggests additional work not represented in the reviewed catalog.
Identity is name-resolved from public bibliographic data (OpenAlex) and linked to ORCID 0009-0000-1865-3059; verify before reuse.