Detecting negative words
Download as PPTX, PDF1 like380 views
This document discusses sentiment analysis and datasets for detecting negative words in Persian text. It describes several existing English sentiment datasets and a refined Persian polarity corpus. It also mentions a corpus of exceptions extracted from a Flexicon database and the need to refine this exceptions list. The document outlines how the negative word detection algorithm works and proposes areas for further development, including creating a database of positive affixed words and using statistical approaches to increase the algorithm's accuracy.
![Datasets
• Refined Persian Polarity corpus
• Dehdarbehbahani, I., Shakery, A., & Faili, H. (2014). Semi-supervised word
polarity identification in resource-lean languages. Neural Networks, 58, 50-
59.
• Corpus of exceptions
• [working dataset]](https://image.slidesharecdn.com/presentation1-170511094522/85/Detecting-negative-words-5-320.jpg)
Detecting negative words
- 1. Detecting Negative Words Adel Rahimi Adel.Rahimi@mehr.sharif.edu RezaTakhshid Reza.takhshid95@student.sharfi.edu Sharif University ofTechnology Spring 2017
- 2. Sentiment datasets for other languages • AFINN by Finn Årup Nielsen AFINN is a list of English words rated for valence with an integer between minus five (negative) and plus five (positive).The words have been manually labeled by Finn Årup Nielsen in 2009-2011.The file is tab-separated.
- 3. Sentiment datasets for other languages • Opinion Lexicon by Hu and Liu A list of English positive and negative opinion words or sentiment words (around 6800 words)
- 4. Sentiment datasets for other languages • NRCWord-EmotionAssociation Lexicon by Saif Mohammad and PeterTurney The NRC Emotion Lexicon is a list of English words and their associations with eight basic emotions (anger, fear, anticipation, trust, surprise, sadness, joy, and disgust) and two sentiments (negative and positive).The annotations were manually done by crowdsourcing.
- 5. Datasets • Refined Persian Polarity corpus • Dehdarbehbahani, I., Shakery, A., & Faili, H. (2014). Semi-supervised word polarity identification in resource-lean languages. Neural Networks, 58, 50- 59. • Corpus of exceptions • [working dataset]
- 6. Datasets • Corpus of exceptions was extracted from Flexicon database • ناشتا • نارنگی • نارج • بیدمشک • الروبی
- 7. Datasets • The exceptions list also needs to be refined • اجتماعی غیر • اصولی غیر • ناشنوایی • پادگان
- 8. Datasets • List of validly affixed words
- 9. How does the algorithm work? INPUT Not Negative Not Negative Negatives list Exception list Negative Affix searching Negative
- 10. Further development • Creating a database of Affixed but positive words • پروا بی • ضدآب • Using Elasticsearch as database for making the process faster • Using a corpus instead of FLexicon • Using statistical approaches for increasing accuracy
- 11. Further Research • The datasets are still not reliable enough. Need to be worked on so that the accuracy of the algorithm will be higher.