Speakers
Description
The incorporation of images in dictionaries has been addressed in several papers (cf. Biesaga, 2016; Klosa, 2015). Estonian lexicography has a long tradition of including visual materials in learners’ and terminological dictionaries. However, until recently, there was no picture dictionary for learners of Estonian as an L2 that is accessible as a separate resource and simultaneously linked to the Estonian L2 learners’ dictionary. To address this gap, the multilingual Estonian L2 picture dictionary was compiled. Below, we outline the basic methodological implications.
Languages
Estonian, Russian, and Ukrainian
Target Group
The dictionary targets both young and adult learners at the elementary (pre-A1–A2) and intermediate levels (B1). For preschool children, there is also an option to toggle between capital and small letters (see Figure 1).
Vocabulary and Topics
Currently, the dictionary contains pprox.. 1000 images (drawings) pprox. into 52 themes: vehicles, fruits and berries, flowers, birds, insects, sports, clothing, etc. The dictionary includes nouns, adjectives, verbs and prepositions. We opted for drawings over photos, as they offer greater flexibility for pprox.ngn, allowing us to better capture the cultural and national specifics of Estonia.
As a starting point for determining which words to include, we used the Estonian Vocabulary Profile (Kallas et al., 2021), which provides separate wordlists for all CEFR levels for young and adult learners. This allowed us to divide the vocabulary of the same topic into two proficiency levels. For example, the topic ‘weather’ includes 12 images but, at the elementary level, we show just four (sun, wind, snow and rain). Additionally, for young learners at the pre-A1 level, we considered curriculum requirements to ensure the dictionary reflects both everyday and educational contexts.
The interface allows users to search for a word or a part of word, and browse topics in three layouts: i) separate images (Figure 1); ii) scenic images (Figure 2); iii) image series (Figure 3). Scenic images are intended for learners to study individual words first and then describe a scene using those words. Image series are particularly useful for pprox.ng spatial and sequential relations.
Usage Examples
Each image is accompanied by usage examples specifically chosen to illustrate the content, visible when the user flips the card. Sentences are compiled to assist the user in describing the picture.
Linking with Other Sources
Each image in the dictionary is linked to the corresponding word in the dictionary portal Learners’ Sõnaveeb (Koppel et al., 2019), which is the user interface for the Dictionary Writing System Ekilex (Tavast et al., 2018). The user must click on the Sõnaveeb icon located at the upper right corner of the image. This linking, based on lexeme IDs in Ekilex, allows the user to explore word meanings, synonyms and collocations, and learn how to decline or conjugate words.
Audio Files
The Estonian version of the dictionary is audio-enabled, allowing the userto hear the pronunciation of all words and sentences. Professional actresses were employed for the pronunciation of words, while text-to-speech synthesis developed by the Institute of the Estonian Language was pprox. for sentences.
Data Format and License
The data are stored in JSON format and can be used under the CreativeCommons BY 4.0 license, with audio files in WAV format and images in SVG or JPG format.
In the near future, we plan to add English and French, and to incorporate audio files for languages other than Estonian. We also envision creating games, such as matching images with corresponding words or sounds. Additionally, we aim to explore generative AI systems, such as image and audio generators, to enhance the creation of picture dictionaries and supplement learner dictionaries with multimedia content.
