Retrieving words from their "meanings"

Durgar El-Kahlout, İlknur (2003) Retrieving words from their "meanings". [Thesis]

[thumbnail of durgarkahloutilknur.pdf] PDF

Download (773kB)


The human brain is the best memory that can record and keep a huge number of information for a long time. Words, their meanings, domains, relationships between different words, and the grammars of languages are well organized in the linguistic component of brain. While speaking or writing, we can generally express our thoughts and feelings by words without thinking for a long time what the correct words can be. But, sometimes things do not go like clockwork even for human brain. In our daily life, we can often forget or not remember a word that we use frequently and exactly know its meaning. While writing a document, talking with friends, or solving a puzzle, we can not remember which word to say or to write. When we face this problem, it will be of no use to attempt searching in a traditional dictionary to find the word that we can not remember. In such cases, there is a need for resources that can locate the word from its meaning. This thesis presents the design and the implementation of a Meaning to Word dictionary (MTW), that locates a set of Turkish words, which most closely matches the correct/appropriate one based on a definition entered by the user. The approach of extracting words from " meaning" s is based on checking the similarity between the user's definition and an entry of the Turkish dictionary without considering any semantics or grammatical information. MTW can be used in various application areas such as computer-assisted language learning, finding the correct words for the definition questions in solving crossword puzzles, and searching the one word representations or synonyms of a multi-word definitions in a reverse dictionary. Results on unseen data indicate that in 72% of the real users queries and 90% of different dictionaries queries, our system returns the correct answer in the first 50 results, respectively.
Item Type: Thesis
Subjects: Q Science > QA Mathematics > QA076 Computer software
Divisions: Faculty of Engineering and Natural Sciences > Academic programs > Computer Science & Eng.
Faculty of Engineering and Natural Sciences
Depositing User: IC-Cataloging
Date Deposited: 17 Apr 2008 16:19
Last Modified: 26 Apr 2022 09:42

Actions (login required)

View Item
View Item