Click on Bake to Get Cookies: Guiding Word-Finding with Semantic Associations
Sonya Nikolova, Marilyn Tremaine, Perry R. Cook · 2010 · Proceedings of the 12th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2010) · doi:10.1145/1878803.1878832
Summary
This paper presents ViVA (Visual Vocabulary for Aphasia), a vocabulary navigation system that uses semantic associations to help people with aphasia find words more effectively. The authors address a core challenge in augmentative and alternative communication (AAC): navigating large vocabularies of thousands of entries to locate the right word for building communication. Traditional AAC tools like Lingraphica organise words in rigid hierarchical categories, requiring users to know which category a word belongs to before they can find it. ViVA takes a different approach by augmenting a standard category hierarchy with links between words that reflect human judgments of semantic relatedness — associations gathered through large-scale evocation studies where participants listed words that came to mind given a prompt word. The system was built on a core vocabulary of 200 words, with each word represented by an icon showing a pictorial representation of the concept. When a user clicks on a word in ViVA, related words appear in a panel alongside the traditional category navigation, creating shortcuts through the vocabulary. The researchers designed an experimental task using 14 scenarios, each consisting of an image and a sentence with a missing word. Participants had to identify the missing word from multiple choices, then find two target words in the vocabulary.
Key findings
A within-subjects study with 20 people with aphasia compared word-finding using ViVA versus a hierarchical vocabulary (LG, modelled on Lingraphica). Results showed that participants took significantly shorter paths to find words with ViVA, with 60% of word pairs found by directly following semantic associations. The average path to the first target word was significantly shorter in ViVA than in LG (F(1, 011) = 10.68, p = 0.01). However, finding the first word to initiate a sentence remained problematic — there was no significant difference between the two systems for locating the first word in a sentence. Participants found ViVA less confusing to navigate and agreed that the associations helped them find words faster. Qualitative feedback revealed that some participants explored the vocabulary by clicking on icons without any evident plan, while others explicitly used the semantic browsing strategy. The study also highlighted that words in abstract categories such as modifiers were harder to locate in both systems, and that users had difficulty when the hierarchical organisation did not match their mental model of where a word should belong.
Relevance
This research demonstrates the practical value of organising vocabulary tools around semantic relationships rather than rigid categories, an insight applicable well beyond aphasia-specific AAC tools. The finding that semantic associations shorten navigation paths has implications for any interface where users must locate items in large collections — including accessibility glossaries, documentation systems, and assistive technology configuration. The study also highlights a persistent design challenge: helping users take the first step in a search when they cannot articulate what they are looking for. For accessibility practitioners, the paper underscores the importance of user-centred vocabulary organisation and the limitations of purely hierarchical information architecture for people with cognitive or language impairments.
Tags: aphasia · semantic networks · AAC · visual vocabularies · word-finding · assistive communication · adaptive user interfaces · lexical access