Arash Hajikhani

VTT Technical Research Centre of Finland · Finland

Unclaimed · AI-structured profileORCID OpenAlex

VTT Finland researcher applying NLP, LLMs and text mining to innovation studies, IP analytics and policy monitoring

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-09-28) 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—

No output score has been computed yet; this axis would measure total publication volume and its distribution over time. The snapshot does record 25 works in the last three years and per-year counts from 2014 to 2026, but a comparable standardized output score is not available.

Impact—

No impact score is available yet; this axis would measure citation-based influence such as field-normalized citation rates against peer benchmarks. The snapshot lists an h-index of 10 and an i10-index of 12, but the normalized indicators required for scoring are not provided.

Focus100/100

Scores 100/100, reflecting a highly concentrated research identity. Topic counts cluster tightly — innovation and knowledge management (8), intellectual property and patents (6), digital marketing and social media (5), computational and text analysis methods (5) and topic modeling (4) — all served by one applied text-mining toolkit. The reviewed 2026 work narrows further into NLP, with the sole recorded field being natural language processing (vision-language models).

Collaboration44/100

Scores 44/100. The OpenAlex snapshot records 8 distinct coauthors, with recurring partners including Toqeer Ehsan, Carolyn Cole, Matthias Deschryvere, Nico Penttilä, Richard Schmidt and Victoria Palacin. Both 2026 catalog papers are team-authored without a corresponding-author role (last position on the policy paper, middle on the shared-task paper), consistent with senior-contributor participation in compact teams rather than large-consortium leadership.

Activity95/100

Scores 95/100, the strongest signal in this profile. The snapshot counts 25 works over the last three years, and per-year counts climb from 2 (2023) to 6 (2024), 7 (2025) and 12 (2026) — the highest annual count in a record reaching back to 2014 (3 works), exceeding the earlier peak of 8 in 2021.

Reproducibility100/100

Scores 100/100, with an important base-size caveat: the mean of 100 is computed from only 1 reproducibility-assessed paper. That paper — the SHROOM-Visions multi-judge committee study — is rated high readiness. The other reviewed 2026 paper (LLM-based OECD STIP survey pre-filling) is rated low readiness, with Evolve noting circular self-evaluation and lenient metrics.

Research summary

Arash Hajikhani is a researcher at VTT Technical Research Centre of Finland whose work sits at the junction of computational text analysis and innovation studies. The OpenAlex snapshot, identity-resolved against ORCID 0000-0003-2032-9180, shows a publication record spanning 2014 to 2026, with 25 works in the last three years and citation indicators of h-index 10 and i10-index 12. Topic counts concentrate on innovation and knowledge management (8), intellectual property and patents (6), digital marketing and social media (5), computational and text analysis methods (5) and topic modeling (4) — a coherent applied text-mining agenda. Output is accelerating: 2 works in 2023, 6 in 2024, 7 in 2025 and 12 in 2026, the highest annual count in the record. The two 2026 arXiv papers reviewed by Evolve mark a pivot toward large language model evaluation. One converts OECD STIP policy documents into structured survey responses with GPT-4o, reporting 84-95% label overlap, though circular self-evaluation and lenient metrics weaken the claim. The other detects hallucinated spans with a five-model vision-language committee, winning 3 of 4 SHROOM-Visions languages and rated high reproducibility readiness. Radar dimensions score activity 95, focus 100 and reproducibility 100, with collaboration at 44 across 8 distinct coauthors.

Collaboration pattern

The OpenAlex snapshot records 8 distinct coauthors and a collaboration score of 44/100. Recurring partners include Toqeer Ehsan, Carolyn Cole, Matthias Deschryvere, Nico Penttilä, Richard Schmidt and Victoria Palacin. Both 2026 catalog papers are team-authored without a corresponding-author role — last position on the policy paper, middle on the shared-task paper — consistent with senior-contributor participation in compact teams rather than large-consortium leadership.

Research keywords

large language modelsvision-language modelshallucination detectionensemble methodspolicy monitoringsurvey automationinformation extractiontopic modelingcomputational text analysisinnovation and knowledge managementintellectual property and patentsdigital marketing and social media

Highlights

Multi-judge VLM committee wins 3 of 4 SHROOM-Visions languages
At the SHROOM-Visions shared task, a five-model vision-language committee detected hallucinated spans through majority voting, winning 3 of 4 languages. Evolve rates the aggregation design strong and reproducibility readiness high, but notes the vote threshold was tuned on the test split and no significance tests were reported.
GPT-4o pre-fills OECD STIP policy surveys with 84-95% label overlap
The paper converts OECD STIP policy documents into structured survey responses using long-context in-context learning, reporting 84-95% label overlap with reference answers. Evolve cautions that circular self-evaluation and lenient metrics weaken the headline claim, and rates reproducibility readiness low.
Sustained applied text-mining agenda across innovation studies
Topic distribution across the record clusters on innovation and knowledge management (8), intellectual property and patents (6), digital marketing and social media (5), computational and text analysis methods (5) and topic modeling (4), with 25 works recorded in the last three years.

Timeline

2014Earliest year in the OpenAlex publication record, with 3 works recorded
2021Pre-2026 peak year, with 8 works recorded
2023Lowest recent output at 2 works, before the current ramp-up
2024Recovery year with 6 recorded works; the recent three-year window totals 25 works
2026Highest recorded annual output (12 works); both Evolve-reviewed arXiv papers appear — the SHROOM-Visions multi-judge committee study and the LLM policy-monitoring evaluation

Papers in the Evolve catalog

Frequent co-authors

Toqeer Ehsan×2 Carolyn Cole×1 Matthias Deschryvere×1 Toqeer Ehsan×1 Arash Hajikhani×1 Nico Penttilä×1 Richard Schmidt×1 Victoria Palacin×1

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.

Bibliometric indicators come from a single OpenAlex snapshot (identity: openalex) tied to ORCID 0000-0003-2032-9180; citation counts and venue details were not independently verified.
Output and impact radar axes are unscored; the only citation indicators available are h-index 10 and i10-index 12.
The reproducibility score of 100 is based on just 1 assessed paper (reproPapers = 1); the second catalog paper is rated low reproducibility readiness.
The frequent-coauthor list contained a self-entry (Arash Hajikhani) and a duplicate (Toqeer Ehsan); the self-entry was excluded and duplicates collapsed for collaboration analysis.