• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Student
Title
Supervisor
Faculty
Educational Programme
Final Grade
Year of Graduation
Ekaterina Mizerova
Automated Image Analysis as a Method of Selecting Visual Stimuli in Lexical Typology Research
10
2019
Typological studies demand the creation of a significant amount of questionnaires of very different types. Since this process is complex, time-consuming and not always objective, automatic methods are very relevant in this area. New methods and tools for automatic analysis of texts and images are being developed, which, among other things, can be effectively used to automate the process of a questionnaire construction. This study is devoted to the development of such methods for lexical typology.

Although some practical algorithms facilitating the task of creating a context-based questionnaire for typological studies of lexicon have already been proposed, automatic image processing has never been used in this area. However, when applied to lexicology, this concept can also be useful.

In my research I set two main tasks.

Firstly, I intend to develop a tool that would help us to understand which of the typologically relevant situations can be identified on the basis of automatic image processing. As an example, the verbs of falling well studied in lexical typology were taken. To accomplish the task I will create my own collection of images, using the images from Google.com, using the requests for the verb "fall" in different languages. Then I will group all the images into clusters. Next, I will compare my results to those obtained for context questionnaires in previous studies. It will provide us with deeper understanding of the semantic characteristics of the domain of falling.

Secondly, I contemplate to improve the results of previous research in this area by adding image vectors to existing text vectors. To do this, I will modify the algorithm from previous studies based on vector representations of Russian phrases with the verbs of falling and include images obtained at the previous stage of this work. This will allow us to draw conclusions about whether this algorithm can improve the quality of lexico-typological questionnaires created automatically.

Key-words: Questionnaire, Lexical typology; Natural language processing; Semantic Role labelling, Distributional semantic models

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses