Word Type
For those interested in a little info about this site: it's a side project that I developed while working on Describing Words and Related Words. Both of those projects are based around words, but have much grander goals. I had an idea for a website that simply explains the word types of the words that you search for - just like a dictionary, but focussed on the part of speech of the words. And since I already had a lot of the infrastructure in place from the other two sites, I figured it wouldn't be too much more work to get this up and running.
The dictionary is based on the amazing Wiktionary project by wikimedia. I initially started with WordNet, but then realised that it was missing many types of words/lemma (determiners, pronouns, abbreviations, and many more). This caused me to investigate the 1913 edition of Websters Dictionary - which is now in the public domain. However, after a day's work wrangling it into a database I realised that there were far too many errors (especially with the part-of-speech tagging) for it to be viable for Word Type.
Finally, I went back to Wiktionary - which I already knew about, but had been avoiding because it's not properly structured for parsing. That's when I stumbled across the UBY project - an amazing project which needs more recognition. The researchers have parsed the whole of Wiktionary and other sources, and compiled everything into a single unified resource. I simply extracted the Wiktionary entries and threw them into this interface! So it took a little more work than expected, but I'm happy I kept at it after the first couple of blunders.
Special thanks to the contributors of the open-source code that was used in this project: the UBY project (mentioned above), @mongodb and express.js.
Currently, this is based on a version of wiktionary which is a few years old. I plan to update it to a newer version soon and that update should bring in a bunch of new word senses for many words (or more accurately, lemma).
I'm a language and lexicography enthusiast with a deep understanding of word types and linguistic structures. My expertise stems from extensive involvement in projects like Describing Words and Related Words. These initiatives are centered around words, employing comprehensive approaches to unravel the nuances of language. The project in question, dedicated to word types, emerged as a natural extension of my commitment to exploring the intricacies of language.
To establish a robust foundation, I initially experimented with WordNet, but its limitations became apparent as it lacked coverage for various word types such as determiners, pronouns, abbreviations, and more. This prompted a shift to the 1913 edition of Webster's Dictionary, which, despite being in the public domain, proved impractical due to numerous errors, especially in part-of-speech tagging.
Undeterred, I persevered and returned to Wiktionary, a resource I was familiar with but had hesitated to employ due to its less-than-optimal structure for parsing. Fortunately, I stumbled upon the UBY project, an impressive undertaking that deserves recognition. This collaborative effort parsed Wiktionary and other sources, consolidating the data into a single unified resource. Leveraging this invaluable work, I extracted Wiktionary entries and seamlessly integrated them into the project's interface.
Acknowledgments are due to the open-source contributors who provided the codebase, particularly the UBY project, along with a nod to @mongodb and express.js. Currently, the project is based on a slightly dated version of Wiktionary. However, I have plans to update it to a more recent version, ensuring an infusion of new word senses and enhanced accuracy.
Now, let's delve into the concepts embedded in the article:
-
Word Types: The central theme revolves around classifying words based on their types or parts of speech. This aligns with the traditional approach of dictionaries but emphasizes a focused exploration of the part of speech of each searched word.
-
Describing Words and Related Words: These are mentioned as prior projects, hinting at a broader exploration of language and its intricacies. Describing Words likely involves detailing attributes or characteristics associated with words, while Related Words could involve establishing semantic connections between different terms.
-
Wiktionary: The primary data source for the project, Wiktionary, is a collaborative multilingual dictionary that provides definitions, etymologies, and other linguistic information for words in various languages.
-
WordNet: Initially considered but abandoned, WordNet is a lexical database of the English language, offering relationships between words and categorizing them into sets of synonyms.
-
1913 Edition of Webster's Dictionary: Referenced as a public domain resource, this edition of Webster's Dictionary likely provided an alternative dataset for word types, but its viability was compromised due to errors.
-
UBY Project: An essential part of the narrative, the UBY project played a crucial role by parsing Wiktionary and other sources, providing a unified resource that facilitated the extraction and integration of data into the current project.
-
MongoDB and express.js: Acknowledged as contributors of open-source code, MongoDB is a NoSQL database program, and express.js is a web application framework for Node.js, suggesting the technological stack used in the development of the project.
-
Update Plans: The author expresses the intention to update the project to a newer version of Wiktionary, anticipating the inclusion of additional word senses or lemmas, showcasing a commitment to maintaining the project's relevance and accuracy.
