Aims and Scope

As the Web rapidly evolves, people are becoming increasingly enthusiastic about interacting, sharing, and collaborating through social networks, online communities, blogs, wikis, therefore, it is critical to correctly interpret sentiments and opinions expressed or reported about social events, political movements, company strategies, marketing campaigns, and any other form or online interaction. Models driven by sub-symbolic Artificial Intelligence such as machine learning algorithms and vector representations have achieved state-of-the-art results for Sentiment Analysis tasks. These models get very accurate results but with a limited understanding of patterns and features used to correctly classify into sentiment categories. Thus, these models lack transparency, traceability, and explainability on how the decisions are taken. This limits Artificial Intelligence methods to produce human-comprehensible solutions, reducing the trust towards the machine generated results. Within this scenario, semantic technologies with explicit semantics can be leveraged to explain why a resource has been classified in a specific sentiment category, inducing trustworthiness and avoiding biases.

Papers deadline

March 1st, 2021

March 14th, 2021


March 31st, 2021

April 7th, 2021


April 15th, 2021

Zip with final papers

April 15th, 2021

Workshop Topics

We are interested in novel contributions to explain Sentiment Analysis results through the use and development of methodologies based on semantics, focused but not limited to the following areas. We seek to receive papers that clearly state and contextualize how the proposed contribution is integrated in the real-world scenario and supports all the stakeholders in using and applying Sentiment Analysis.

  • Methodologies
    • Methods to build ontologies for Sentiment Analysis.
    • Methods to build knowledge graphs for Sentiment Analysis.
    • Methods to build lexicons for Sentiment Analysis.
    • Explainable Machine Learning models for Sentiment Analysis.
    • Bias detection through semantics for Sentiment Analysis.
    • Semantics for toxicity and hate speech detection.
    • Application of existing semantic technologies multimedia data.
    • Multilingual Sentiment Analysis methodologies.
  • Representation Methods
    • Ontologies for Sentiment Analysis.
    • Knowledge graphs for Sentiment Analysis.
    • Combination of existing resources (e.g., SenticNet) with embedding representations.
    • Novel Word Embeddings for Sentiment Analysis.
    • Knowledge Graph Embeddings for Sentiment Analysis.
    • Reviews about existing limitations.
  • Case Study Exploration
    • Educational environments.
    • Healthcare systems.
    • Scholarly discussions (e.g., peer review process discussions, mailing lists, etc.).
    • Mental health systems.
    • News platforms.
    • Social networks.


The submissions must be in English and adhere to the CEUR-WS one-column template. The papers should be submitted as PDF files to OpenReview. The review process will be single-blind. Please be aware that at least one author per paper must be registered and attend the workshop to present the work.

We will consider three different submission types:

  • Full Papers (10-12 pages) should be clearly placed with respect to the state of the art and state the contribution of the proposal in the domain of application, even if presenting preliminary results. In particular, research papers should describe the methodology in detail, experiments should be repeatable, and a comparison with the existing approaches in the literature is encouraged.
  • Short Papers (5-9 pages) should describe significant novel work in progress. Compared to full papers, their contribution may be narrower in scope, be applied to a narrower set of application domains, or have weaker empirical support than that expected for a full paper. Submissions likely to generate discussions in new and emerging areas of sentiment explainability are encouraged.
  • Position Papers (2-4 pages) should introduce new point of views in the workshop topics or summarize the experience of a group in the field.

Submissions should not exceed the indicated number of pages, including any diagrams and references.

Each submission will be reviewed by three independent reviewers on the basis of relevance for the workshop, novelty/originality, significance, technical quality and correctness, quality and clarity of presentation, quality of references and reproducibility.

The accepted papers and the material generated during the meeting will be available on the workshop website. The workshop proceedings will be sent for inclusion in a CEUR-WS volume and consequently indexed on Google Scholar, DBLP, and Scopus. The best paper may be included in the supplementary proceedings of ESWC 2021, which will appear in the Springer LNCS series.


To participate, you need to register to the ESWC conference . Please note that the Early Registration is until May 28th.


X-SENTIMENT workshop will take place online on June 7th, 2021, starting at 9:00 (UTC+2, CEST TIME)..

Timing Content
09:00 09:15 Workshop Opening - Welcome
09:15 10:00 Keynote on Ensemble Application of Symbolic and Subsymbolic AI for Sentiment Analysis

Speaker: Prof. Dr. Erik Cambria, School of Computer Science and Engineering

Short Bio: Prof. Dr. Erik Cambria is the Founder of SenticNet, a Singapore-based company offering B2B sentiment analysis services, and an Associate Professor at NTU, where he also holds the appointment of Provost Chair in Computer Science and Engineering. Prior to joining NTU, he worked at Microsoft Research Asia (Beijing) and HP Labs India (Bangalore) and earned his PhD through a joint programme between the University of Stirling and MIT Media Lab. His research focuses on the ensemble application of symbolic and subsymbolic AI to natural language processing tasks such as sentiment analysis, dialogue systems, and financial forecasting. Erik is recipient of many awards, e.g., the 2019 IEEE Outstanding Early Career Award, he was listed among the 2018 AI's 10 to Watch, and was featured in Forbes as one of the 5 People Building Our AI Future. He is Associate Editor of several top AI journals, e.g., INFFUS, IEEE CIM, and KBS, Special Content Editor of FGCS, Department Editor of IEEE Intelligent Systems, and is involved in many international conferences as program chair and invited speaker.

Abstract: Deep learning has unlocked new paths towards the emulation of the peculiarly-human capability of learning from examples. While this kind of bottom-up learning works well for tasks such as image classification or object detection, it is not as effective when it comes to natural language processing. Communication is much more than learning a sequence of letters and words: it requires a basic understanding of the world and social norms, cultural awareness, commonsense knowledge, etc.; all things that we mostly learn in a top-down manner. In SenticNet, we integrate top-down and bottom-up learning via an ensemble of symbolic and subsymbolic AI tools, which we apply to the interesting problem of polarity detection from text.

10:00 10:15
Keynote Talk Q&A
10:15 10:35
Coffee Break
10:35 11:50
Paper Presentations (20 mins + 5 mins Q&A)
  • 10:35 - 11:00
    Explaining Sentiment from Lexicon
    Authors: Sergio Consoli, Luca Barbaglia, and Sebastiano Manzan
  • 11:00 - 11:25
    Extractive Summarization for Explainable Sentiment Analysis using Transformers
    Luca Bacco, Andrea Cimino, Felice Dell'Orletta, Mario Meron
  • 11:25 - 11:50
    Deep Learning Enhanced with Graph Knowledge for Sentiment Analysis
    Fernando Andres Lovera, Yudith Cardinale, Davide Buscaldi, Thierry Charnois, Masun Nabhan Homsi.
11:50 12:10
General Discussion
12:10 12:20
Closing Remarks


Workshop Chairs

Program Committee


For general enquiries on the workshop, please send an email to: