Grammar or Crammer? The Role of Morphology in Distinguishing Orthographically Similar but Semantically Unrelated Words

Gökhan Ercan, Olcay Taner Yıldız
IEEE Access, vol. 13, pp. 64412–64458, 2025

Abstract

We show that n-gram-based distributional models fail to distinguish unrelated words due to the noise in semantic spaces. This issue remains hidden in conventional benchmarks but becomes more pronounced when orthographic similarity is high. To highlight this problem, we introduce OSimUnr, a dataset of nearly one million English and Turkish word-pairs that are orthographically similar but semantically unrelated (e.g., grammar – crammer). These pairs are generated through a graph-based WordNet approach and morphological resources. We define two evaluation tasks — unrelatedness identification and relatedness classification — to test semantic models. Our experiments reveal that FastText, with default n-gram segmentation, performs poorly (below 5% accuracy) in identifying unrelated words. However, morphological segmentation overcomes this issue, boosting accuracy to 68% (English) and 71% (Turkish) without compromising performance on standard benchmarks (RareWords, MTurk771, MEN, AnlamVer). Furthermore, our results suggest that even state-of-the-art LLMs, including Llama 3.3 and GPT-4o-mini, may exhibit noise in their semantic spaces, particularly in highly synthetic languages such as Turkish. To ensure dataset quality, we leverage WordNet, MorphoLex, and NLTK, covering fully derivational morphology supporting atomic roots (e.g., "-co_here+ance+y" for "coherency"), with 405 affixes in Turkish and 467 in English.

Key findings

  • FastText with default character n-gram segmentation identifies orthographically similar but semantically unrelated words with below 5% accuracy — a failure that conventional word-similarity benchmarks hide entirely.
  • Replacing n-gram segmentation with morphological segmentation raises that accuracy to 68% in English and 71% in Turkish, with no loss on standard benchmarks (RareWords, MTurk771, MEN, AnlamVer).
  • State-of-the-art large language models, including Llama 3.3 and GPT-4o-mini, show the same kind of noise in their semantic spaces, and it is more pronounced in highly synthetic languages such as Turkish.
  • OSimUnr: a dataset of nearly one million English and Turkish word-pairs that are orthographically similar yet semantically unrelated (e.g. grammar – crammer), built from a graph-based WordNet approach plus morphological resources.
  • Two evaluation tasks are defined for probing semantic models: unrelatedness identification and relatedness classification.
  • Derivational morphology coverage supporting atomic roots (e.g. "-co_here+ance+y" for "coherency"), spanning 405 affixes in Turkish and 467 in English, drawn from WordNet, MorphoLex and NLTK.

Download PDF  DOI: 10.1109/ACCESS.2025.3557086


Gökhan Ercan See it on X
A pleasure to teach Agentic AI this term at Istanbul Technical University.

We covered 17 agentic patterns: Agentic RAG, ReAct, Memory, Guardrails, multi-agent, Graph-of-Thoughts. Best part: seeing 40 student projects move from "use an LLM" to "design an agent."
Jun 25, 2026
Gökhan Ercan See it on X
My latest paper is out in IEEE Access:

📄Grammar or Crammer? The Role of Morphology in Distinguishing Orthographically Similar but Semantically Unrelated Words

A 47-page study on morphology and semantic modeling in English and Turkish. t.co/gf663FIG5A
Jun 11, 2025
Gökhan Ercan See it on X
“As a leader, you’re the force multiplier for your organization. Force multiplying is a vector quantity, not a scalar quantity. It’s a vector quantity because the direction is very important. It’s not just the magnitude. It’s the direction, too.” t.co/HePyFKJ84S
Feb 27, 2025
Gökhan Ercan See it on X
Let’s come up with a new name for #Agile, create our own backlogs, and start fresh. Keep it secret from the business folks. t.co/yAulbTeMW2
Nov 12, 2024
Gökhan Ercan See it on X
"...I don't think that anybody really thinks that AlphaGo is happy when it wins the game..."
"...We can use a knife to cook and to kill.."
Yuval Noah Harari (Sapiens) versus Yann Le Cun (Meta) on artificial intelligence t.co/IlgRKDWaNc via @LePoint
May 16, 2023
Gökhan Ercan See it on X
Even experts cannot predict social changes (e.g., prejudice, violence, individualism, political polarization). They're no better than naive people or basic statistical models. They're more aware of the uncertainty and complexity, though.
t.co/EV95EXZ1gw
Mar 25, 2023