Expert Artificial Intelligence Terminology Landscape (based on English)
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
This research investigates the contemporary terminology landscape of Artificial Intelligence (AI), focusing on the specialized English vocabulary employed by experts. Given AI’s dynamic nature and profound impact, understanding the characteristics of its expert lexicon is vital for clear communication, effective knowledge transfer, and navigating the complexities of the domain. The study aims to identify, characterize, and analyze the key linguistic features of modern expert AI terminology using an empirical, corpus-based methodology. A specialized, synchronic corpus comprising 2,865 titles and abstracts (approx. 544,000 words) from the 2024 AAAI Conference on Artificial Intelligence proceedings was compiled for this purpose. Analysis was conducted employing corpus linguistic tools, primarily within the Sketch Engine environment, alongside initial processing by the Gemini API. The specialized corpus was compared against a large general English reference corpus (English Trends 2014-today) to assess domain specificity. The findings highlight a terminology heavily concentrated on core methodological themes such as model architectures, learning paradigms, optimization techniques, and data engineering. Structurally, the lexicon is dominated by nominal forms, exhibiting a high prevalence of multi-word terms (MWTs), particularly following N+N and Adj+N patterns, and extensive use of acronyms (e.g., LLM, GNN, RL). Keyword analysis confirmed a high degree of domain specificity, resulting from both technical neologisms unique to AI and the significant semantic specialization of common English words (e.g., model, attention, learning, training, bias, hallucination). Semantic specialization narrows general meanings to precise computational or algorithmic concepts, while metaphorical extension (e.g., mapping cognitive or psychological concepts like learning, attention, or hallucination onto computational processes) serves as a crucial mechanism for term creation and conceptualization.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it