Sentiment analysis and opinion mining are text analysis methods that identify and extract people’s opinions, attitudes and sentiments within a text (Zhang and Liu 2017, 1153). Sentiment analysis refers to the method of determining the sentiment of a particular setence (or potentially other grammatical construction) by assigning it a ternary ("positive", "neutral", "negative") or scalar value to the sentence. Opinion mining is a text extraction method targeting those parts of a text in which a person's attitude, opinion, or sentiment is expressed. In terms of applicability, sentiment analysis and opinion mining are widely used in domains like social media or customer reviews, where the user's voice is expressed.
The CLARIN infrastructure provides access to 5 tools dedicated to sentiment analysis or opinion mining, available for download or online access. 4 of the tools can be used for sentiment analysis/opinion mining within a single language (Polish, Greek, and Finnish), while 1 tool has a broad multilingual scope and can be used for opinion mining in 10 languages: English, French, Italian, German, Swedish, Norwegian, Danish, Finnish, Spanish, and Dutch.
For comments, changes of the existing content or inclusion of new tools, send us an resource-families [at] clarin.eu (email).
Tools for Sentiment Analysis and Opinion Mining in the CLARIN Infrastructure
| Tool | Language | Description |
|---|---|---|
|
Etuma Customer Feedback Analysis Functionality: sentiment analysis, opinion mining |
English, French, Italian, German, Swedish, Norwegian, Danish, Finnish, Spanish, Dutch |
The tool is used for ranking customers' feedback in order of intensity and detects the sentiment of ongoing discussions in order to determine whether the overall response to a product or campaign is positive or negative. CLARIN centre: FIN-CLARIN |
|
Functionality: sentiment analysis |
Finnish |
This tool relies on three resources:
The neural network is trained to predict the rating associated with product reviews, and the prediction it gives to the input text is converted to a sentiment. Availability: online |
|
OptaHopper: phrase-level sentiment with opinion targets Functionality: sentiment analysis, opinion mining |
Polish |
This is a phrase- and sentence-level sentiment analysis tool based on TreeLSTM (Sheng Tai et al. 2015) integrated with opinion mining. Any sentiment dictionary may be used as an input feature, including lemma-level and plWordNet emo. In the case of plWordNet emo, provided integration with the WSD module. The Opinion Finder OPFI (Wawer 2016) can be used for opinion target extraction. Availability: download |
|
Functionality: sentiment analysis, opinion mining |
Greek |
The sentiment analysis tool is a text classification and sentiment extraction tool based on n-gram graph text representations. It may be paired with various machine learning algorithms for the generation of the language model. It can be accessed by a URL endpoint as a service. It has been used as is, or as part of bigger pipelines in many research tasks. It is also embedded in gov.insight as an annotator producer for sentiment classification. This tool also makes use of an Opinion Mining process. Availability: online |
|
TreeHopper (TreeLSTM): wydźwięk na poziomie zdań i fraz Functionality: sentiment analysis |
Polish |
This is a TreeLSTM-based (Sheng Tai et al. 2015) dependency tree sentiment labeller, implemented in PyTorch and optimized for morphologically rich languages with a relatively loose word order (such as Polish). Availability: download |
Publications
[Kiomourtzis et al. 2014] George Kiomourtzis, George Giannakopoulos, Georgios Petasis, Pythagoras Karampiperis, and Vangelis Karkaletsis. 2014. NOMAD: Linguistic Resources and Tools Aimed at Policy Formulation and Validation. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC'14), edited by Nicoletta Calzolari et al., 26–31. European Language Resources Association.
[Korbak and Żak 2017] Tomasz Korbak, and Paulina Żak. 2017. Fine-tuning Tree-LSTM for phrase-level sentiment classification on a Polish dependency treebank. Submission to PolEval task 2.
[Sheng Tai et al. 2015] Kai Sheng Tai, Richard Socher, and Christopher D. Manning. 2015. Improved Semantic Representations From Tree-Structured Long Short-Term Memory Networks. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), edited by Chengqing Zong and Michael Strube, 1556–1566. Association for Computational Linguistics.
[Wawer 2016] Aleksander Wawer. 2016. OPFI: A Tool for Opinion Finding in Polish. In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16), edited by Nicoletta Calzolari et al., 2906–2909. European Language Resources Association.
[Zhang and Liu 2017] Zhang, Lei, and Bing Liu. 2017. Sentiment Analysis and Opinion Mining. In Encyclopedia of Machine Learning and Data Mining, edited by Claude Sammut and Geoffrey I. Webb, 1152–1161.