Researcher:
Kuru, Onur

Profile Picture
ORCID

Job Title

Master Student

First Name

Onur

Last Name

Kuru

Name

Name Variants

Kuru, Onur

Email Address

Birth Date

Filters

20162

Advanced Search

Filter by

Settings

Researcher Of Publications: search.filters.isAuthorOfPublication.fca3e651-c41d-4822-9379-48d3ffc8a2eb

Search Results

Now showing 1 - 2 of 2
  • Placeholder
    PublicationMetadata only
    AI-KU at SemEval-2016 task 11: word embeddings and substring features for complex word identification
    (Association for Computational Linguistics (ACL), 2016) N/A; N/A; Kuru, Onur; Master Student; Koç Üniversitesi İş Bankası Yapay Zeka Uygulama ve Araştırma Merkezi (KUIS AI)/ Koç University İş Bank Artificial Intelligence Center (KUIS AI); Graduate School of Sciences and Engineering; N/A
    We investigate the usage of word embeddings, namely Glove and SCODE, along with substring features on Complex Word Identification task. We introduce two systems: the first system utilizes the word embeddings of the target word and its substrings as features while the other considers the context information by using the embeddings of the surrounding words as well. Although the proposed representations perform below the average with nonlinear models, we show that word embeddings with substring features is an effective representation choice when employed with linear classifiers.
  • Placeholder
    PublicationMetadata only
    CharNER: character-level named entity recognition
    (Association for Computational Linguistics (ACL), 2016) N/A; N/A; N/A; Department of Computer Engineering; Kuru, Onur; Can, Ozan Arkan; Yüret, Deniz; Master Student; PhD Student; Faculty Member; Department of Computer Engineering; Graduate School of Sciences and Engineering; Graduate School of Sciences and Engineering; College of Engineering; N/A; N/A; 179996
    We describe and evaluate a character-level tagger for language-independent Named Entity Recognition (NER). Instead of words, a sentence is represented as a sequence of characters. The model consists of stacked bidirectional LSTMs which inputs characters and outputs tag probabilities for each character. These probabilities are then converted to consistent word level named entity tags using a Viterbi decoder. We are able to achieve close to state-of-the-art NER performance in seven languages with the same basic model using only labeled NER data and no hand-engineered features or other external resources like syntactic taggers or Gazetteers.