Open-access network science: Investigating phonological similarity networks based on the SUBTLEX-US lexicon

Area of Science:

• Psycholinguistics
• Network Science
• Computational Linguistics

Background:

• Network science tools are vital for psycholinguistics.
• Limited open-access datasets hinder research on English phonological networks.
• Understanding word relationships through networks is crucial.

Purpose of the Study:

• To construct and release open-access phonological similarity networks for English.
• To analyze network properties like small-worldness and robustness.
• To provide resources for psycholinguistic network research.

Main Methods:

• Created phonological similarity networks based on the SUBTLEX-US English corpus.
• Defined network nodes as lemmas or word forms, differing by one phoneme.
• Utilized networkX package for analysis and data structure.

Main Results:

• Networks exhibit small-world properties, broad degree distributions, and robustness.
• Results are consistent across different network sizes and word representations.
• Validated networks show comparable contrasts to prior studies.

Conclusions:

• The released phonological networks are valuable resources for psycholinguistics.
• These networks demonstrate key network science characteristics.
• Open-access data facilitates further exploration of English word structure.