NLP researcher at MBZUAI spanning low-resource language processing, author profiling, and LLM bias evaluation
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
32
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-16) 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—
Not yet scored: this snapshot lacks a verified total publication count and a venue or publication-type breakdown. The axis would measure career output volume and how it is distributed across venues, publication types, and years.
Impact—
Not yet scored: per-paper citation counts and venue-level influence are not aggregated here, so no composite impact score exists, although an h-index of 32 and i10-index of 89 are recorded. This axis would measure citation-based research influence.
Focus0/100
Scores 0 because the field distribution splits evenly, 0.5 to NLP (computational linguistics) and 0.5 to NLP/LLM evaluation, rather than concentrating in one labeled field. The underlying topic counts are nonetheless coherent: topic modeling (138) and NLP techniques (128) lead, followed by sentiment analysis and opinion mining (48), hate speech and cyberbullying detection (35), and authorship attribution and profiling (32).
Collaboration41/100
Scores 41, computed from 7 distinct coauthors in the current snapshot. Recurring collaborators include Toqeer Ehsan, An Vo, Vy Tuong Dang, Khai-Nguyen Nguyen, Emilio Villa-Cueva, Anh Totti Nguyen, and Daeyoung Kim. In the two 2026 catalog papers Solorio holds last and middle author positions, with neither paper listing them as corresponding author.
Activity100/100
Scores 100, the strongest axis here. The record shows 73 works in the last three years, with yearly counts of 37 in 2024, 21 in 2025, and 15 so far in 2026, against an indexed history reaching back to 1998 and a prior peak of 25 in 2021. Recent volume clearly exceeds the career baseline.
Reproducibility78/100
Scores 78, the mean across 2 Evolve-rated papers. The language-model bias paper is rated high: a well-designed, reproducible bias depth score, though threshold-sensitive with no uncertainty reporting. The NER annotation-correction framework is rated medium: reported gains of +1.4 to 4.0 F1 lack significance testing and ablations.
Research summary
Thamar Solorio is an NLP researcher at the Mohamed bin Zayed University of Artificial Intelligence whose OpenAlex-indexed record spans 1998 to 2026, with an h-index of 32 and an i10-index of 89. The profile is name-resolved from public bibliographic data. Activity is the standout radar axis, scored 100: 73 works in the last three years, peaking at 37 indexed works in 2024. Topic counts concentrate on topic modeling (138) and NLP techniques (128), with sustained lines in sentiment analysis and opinion mining (48), hate speech and cyberbullying detection (35), and authorship attribution and profiling (32). The field distribution splits evenly, 0.5 to NLP (computational linguistics) and 0.5 to NLP/LLM evaluation, which yields a focus score of 0 despite the topical coherence. Two 2026 Evolve-reviewed papers illustrate current directions: a scalable self-training framework that repairs NER annotation in Urdu, Shahmukhi, and Sindhi datasets (+1.4-4.0 F1, last author position), and a bias depth score separating 'deep' from 'shallow' opinion biases in language models (middle author). Reproducibility scores 78 across 2 rated papers, one high and one medium pending significance testing and ablations. Collaboration scores 41 on 7 distinct coauthors in the snapshot.
Collaboration pattern
The collaboration score of 41 rests on 7 distinct coauthors in the snapshot. The frequent-collaborator list, Toqeer Ehsan, An Vo, Vy Tuong Dang, Khai-Nguyen Nguyen, Emilio Villa-Cueva, Anh Totti Nguyen, and Daeyoung Kim, is compact, and both reviewed 2026 papers are multi-author efforts in which Solorio appears in last and middle positions without corresponding authorship. The data points to a small, stable working cluster rather than a wide ad-hoc network.
Research keywords
Highlights
Self-training repairs NER labels in low-resource languages
A bias depth score for language model opinions
A broad, long-running NLP agenda
OpenAlex topic counts show 138 works in topic modeling and 128 in NLP techniques, plus 48 in sentiment analysis and opinion mining, 35 in hate speech and cyberbullying detection, and 32 in authorship attribution and profiling, indicating sustained programs across several NLP subfields.
Timeline
- 1998Earliest year in the OpenAlex indexed record, with 1 work counted.
- 201314 indexed works, part of the steady output through the 2000s and 2010s.
- 202125 indexed works, the highest yearly count before 2024.
- 202437 indexed works, the peak year in the record.
- 2026Two papers enter the Evolve catalog: a scalable NER annotation-correction framework for low-resource languages and a bias depth score for language model opinions.
Papers in the Evolve catalog
Frequent co-authors
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 (h-index, yearly counts, topics) come from OpenAlex records with identity resolved from public bibliographic data; coverage may be incomplete or include name variants.
- Output and impact radar dimensions are unscored because no verified total publication count or per-paper citation aggregation was available in this snapshot.
- Qualitative assessment draws on only two Evolve-reviewed papers, both from 2026 on arXiv; earlier career work is not individually reviewed here.
- The source record lists the country as GB while the stated affiliation is MBZUAI; the inconsistency is reported as provided.