Changes between Initial Version and Version 1 of en/AdvancedNlpCourse2020/RelationExtraction


Ignore:
Timestamp:
Aug 31, 2021, 2:11:54 PM (3 months ago)
Author:
Ales Horak
Comment:

copied from private/AdvancedNlpCourse/RelationExtraction

Legend:

Unmodified
Added
Removed
Modified
  • en/AdvancedNlpCourse2020/RelationExtraction

    v1 v1  
     1= Automatic relation extraction =
     2
     3[[https://is.muni.cz/auth/predmet/fi/ia161|IA161]] [[en/AdvancedNlpCourse|Advanced NLP Course]], Course Guarantee: Aleš Horák
     4
     5Prepared by: Adam Rambousek
     6
     7== State of the Art ==
     8
     9=== References ===
     10
     11 1. Lefever, Els, Marjan Van de Kauter, and Véronique Hoste. "Evaluation of Automatic Hypernym Extraction from Technical Corpora in English and Dutch." Proceedings of the 9th International Conference on Language Resources and Evaluation (LREC 2014). 2014.
     12 1. Wang, Tong, and Graeme Hirst. "Exploring patterns in dictionary definitions for synonym extraction." Natural Language Engineering 18.03 (2012): 313-342.
     13 1. Schropp, Gwendolijn, Els Lefever, and Véronique Hoste. "A Combined Pattern-based and Distributional Approach for Automatic Hypernym Detection in Dutch." RANLP. 2013.
     14 1. Grefenstette, Gregory. "INRIASAC: Simple Hypernym Extraction Methods." arXiv preprint arXiv:1502.01271 (2015).
     15 1. Shen, Yatian, and Xuan-Jing Huang. "Attention-based convolutional neural network for semantic relation extraction." Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers. 2016.
     16 1. Li, Qing, et al. "A comprehensive exploration of semantic relation extraction via pre-trained CNNs." Knowledge-Based Systems (2020): 105488.
     17
     18== Practical Session ==
     19
     20Enhance hypernym detection to provide better results.
     21
     22 * Download [[htdocs:bigdata/ia161-hyper.zip|prepared scripts and data]]:
     23 {{{
     24wget https://nlp.fi.muni.cz/trac/research/chrome/site/bigdata/ia161-hyper.zip
     25}}}
     26 * `pip install majka`
     27 * Unzip, `cd ia161-hyper` and run {{{./hyper.py}}}
     28 * The script reads file {{{vstup.txt}}} (each line is word|definition) and outputs hypernym for each word.
     29 * Default approach is naive: ''first noun in definition is hypernym''
     30 * majka gives ''noun'' to some ''adjectives'', deal with this to improve results
     31 * Update the {{{find_hyper()}}} function in `hyper.py` to provide better results.
     32 * Upload updated script plus the output.
     33 * Gold standard to evaluate your result: [[raw-attachment:gold.txt|gold.txt]]