I’m glad to announce that the second unnatural language processing contest will be held at
as a theme session, which is the sequel of Baidu.jp unnatural language processing contest:
successfully held this summer. The “call for papers” is as follows. We seek for your participation and paper submission regarding unnatural language phenomena seen on blogs, mini-blogs, SNSs, and so on.
This time, it is not exactly a contest, but a theme session. I hope we can have deeper academic discussion regarding processing of unnatural language. The proposers of this theme session are: Mr Aramaki @ARAMAKI, Mr. Hashimoto @taiichi84, Mr. Murakami @kmura, Mr. Mizuno @mizuno_takaaki and me @mhagiwara. We also invited two advisers: Prof. Sekine (Rakuten & NYU) and Mr. Takesako (at Cybozu Lab) @takesako.
We are looking forward to the submission from any engineers and researchers who love “unnatural language”!
Theme session 2: UnNatural Language Processing: Processing out-of-box linguistic expressions
As the development of the Web and resulting increase of primary information source such as blogs, mini-blogs, SNSs, and so on, the processing of “unnatural language” such as informal expressions, variations, spelling errors etc. is more and more important. We are also seeing troublesome language phenomena such as irregular proper nouns, emoticons, unknown words in real data. It is one of the important issues for practical systems how to deal with such kind of language phenomena which cannot be captured by conventional “text-book” natural language processing.
In this session, we invite papers and presentations of research result, systems and demos regarding “UnNatural Language Processing; UNLP”. Also, we’ll have a chance to deepen our understanding through discussion.
The topics include, but not limited to:
- Recognition & classification of irregular proper nouns, onomatopoeia, and novel words.
- Extraction and normalization of spelling variations and abbreviations.
- Analysis and extraction of emoticons and ascii arts.
- Spam / fraud detection
- Spelling correction, proofreading, readability evaluation, obfuscation
- Processing of blog, mini-blog, SNSs, BBS, etc.
- Processing and generation of proverbs, idioms, anagram, riddles, jokes
- Demos, systems etc. of these topics
