<!DOCTYPE html>
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-forward-container"><br>
[Apologies for cross-postings]<br>
<br>
Call for Papers<br>
<br>
First International Workshop on Extraction from Triplet
Text-Table-Knowledge Graph and associated Challenge<br>
<a class="moz-txt-link-freetext"
href="https://ecladatta.github.io/triplet2026/"
moz-do-not-send="true">https://ecladatta.github.io/triplet2026/</a><br>
<br>
in conjunction with the 23rd European Semantic Web Conference
(ESWC 2026)<br>
<a class="moz-txt-link-freetext"
href="https://2026.eswc-conferences.org/" moz-do-not-send="true">https://2026.eswc-conferences.org/</a>,
Dubrovnik, Croatia<br>
<br>
Important dates:<br>
- **Submission deadline (extended)**: 13 March, 2026 (11:59pm,
AoE)<br>
- **Notifications**: 31 March, 2026<br>
- **Challenge registration deadline**: 15 March, 2026<br>
- **Challenge results submission**: 10 April, 2026<br>
- **Camera-ready deadline**: 15 April, 2026 (11:59pm, AoE)<br>
- **Workshop**: Sunday 10 May OR Monday 11 May 2026<br>
<br>
Motivation:<br>
Understanding information spread across text and table is
essential for tasks such as question answering and fact checking.
Existing benchmarks primarily deal with semantic table
interpretation or reasoning over tables for question answering,
leaving a gap in evaluating models that integrate tabular and
textual information, perform joint information extraction across
modalities, or can automatically detect inconsistencies between
modalities.<br>
<br>
This workshop aims to provide a forum for exchanging ideas between
the NLP community working on open information extraction and the
vibrant Semantic Web community working on the core challenge of
matching tabular data to Knowledge Graphs, on populating knowledge
graphs using texts and on reasoning across text, tabular data and
knowledge graphs. The workshop also targets researchers focusing
on the intersection of learning over structured data and
information retrieval, for example, in retrieval augmented
generation (RAG) and question answering (QA) systems. Hence, the
goal of the workshop is to connect researchers and trigger
collaboration opportunities by bringing together views from the
Semantic Web, NLP, database, and IR disciplines.<br>
<br>
<br>
Scope:<br>
The topics of interest include but are not limited to:<br>
- Semantic Table Interpretation<br>
- Automated Tabular Data Understanding<br>
- Using Large Language Models (LLMs) for Information Extraction<br>
- Generative Models and LLMs for Structured Data<br>
- Knowledge Graph Construction and Completion with Tabular Data
and Texts<br>
- Analysis of Tabular Data on the Web (Web Tables)<br>
- Benchmarking and Evaluation Frameworks for Joint Text-Table Data
Analysis<br>
- Applications (e.g. data search, fact-checking,
Question-Answering, KG alignment)<br>
<br>
Submission Guidelines:<br>
We invite two types of submissions:<br>
1. Full research papers (12-15 pages) including references and
appendices<br>
2. Challenge papers (6-8 pages) including references and
appendices<br>
<br>
All submissions should be formatted in the CEUR layout format,
<a class="moz-txt-link-freetext"
href="https://www.overleaf.com/latex/templates/template-for-submissions-to-ceur-workshop-proceedings-ceur-ws-dot-org/wqyfdgftmcfw"
moz-do-not-send="true">https://www.overleaf.com/latex/templates/template-for-submissions-to-ceur-workshop-proceedings-ceur-ws-dot-org/wqyfdgftmcfw</a><br>
<br>
This workshop is double-blind and non-archival. Submissions are
managed through EasyChair at <a class="moz-txt-link-freetext"
href="https://easychair.org/conferences/?conf=triplet2026"
moz-do-not-send="true">https://easychair.org/conferences/?conf=triplet2026</a>.
All accepted papers will be presented as posters or as oral talks.<br>
<br>
**TRIPLET Challenge:**<br>
<br>
In recent years, the research community has shown increasing
interest in the joint understanding of text and tabular data,
often, for performing tasks such as question answering or fact
checking where evidences can be found in texts and tables. Hence,
various benchmarks have been developed for jointly querying
tabular data and textual documents in domains such as finance,
scientific publications, and open domain. While benchmarks for
triple extraction from text for Knowledge Graph construction and
semantic annotation of tabular data exist in the community, there
remains a gap in benchmarks and tasks that specifically address
the joint extraction of triples from text and tables by leveraging
complementary clues across these different modalities.<br>
<br>
The TRIPLET 2026 challenge is proposing three sub-tasks and
benchmarks for understanding the complementarity between tables,
texts, and knowledge graphs, and in particular to propose a joint
knowledge extraction and reconciliation process.<br>
<br>
#Sub-Task 1: Assessing the Relatedness Between Tables and Textual
Passages<br>
The goal of this task is to assess the relatedness between tables
and textual passages (within documents and across documents). For
this purpose, we have constructed LATTE (Linking Across Table and
Text for Relatedness Evaluation), a human annotated dataset
comprising table–text pairs with relatedness labels. LATTE
consists of 7,674 unique tables and 41,880 unique textual
paragraphs originating from 3,826 distinct Wikipedia pages. Each
text paragraph is drawn from the same or contextually linked pages
as the corresponding table, rather than being artificially
generated. LATTE provides a challenging benchmark for cross-modal
reasoning by requiring classification of related and unrelated
table–text pairs. Unlike prior resources centered on table-to-text
generation or text retrieval, LATTE emphasizes fine-grained
semantic relatedness between structured and unstructured data.<br>
<br>
The Figure below provides an example, using a web-annotation tool
we developed, of how we identify the relatedness between the
sentence containing the entity AirPort Extreme 802.11n
(highlighted in Orange) and the data table providing information
about output power and frequency for this entity. Participants are
provided with tables and textual passages that would need to be
ranked. The evaluation will use metrics such as P@k, R@k and F1@k.<br>
<br>
Go to <a class="moz-txt-link-freetext"
href="https://www.codabench.org/competitions/12776/"
moz-do-not-send="true">https://www.codabench.org/competitions/12776/</a>
and enroll to participate in this Task.<br>
<br>
<br>
#Sub-Task 2: Joint Relation Extraction Between Texts and Tables<br>
The goal of this task is to automatically extract knowledge
jointly from tables and related texts. For this purpose, we
created ReTaT, a dataset that can be used to train and evaluate
systems for extracting such relations. This dataset is composed of
(table, surrounding text) pairs extracted from Wikipedia pages and
has been manually annotated with relation triples. ReTaT is
organized in three subsets with distinct characteristics: domain
(business, telecommunication and female celebrities), size (from
50 to 255 pairs), language (English vs French), type of relations
(data vs object properties), close vs open list of relation, size
of the surrounding text (paragraph vs full page). We then assessed
its quality and suitability for the joint table-text relation
extraction task using Large Language Models (LLMs).<br>
<br>
Given a Wikipedia page containing texts and tables and a list of
predicates defined in Wikidata, a participant system should
extract triples composed of mentions located partly in the text
and partly in the table and disambiguated with entities and
predicates identified in the Wikidata reference knowledge graph.
For example, in the Figure below, an annotation triple
<Q13567390, P2109, 24.57> is associated with mentions
highlighted in orange (subject), blue (predicate) and green
(object) to annotate the document available at <a
class="moz-txt-link-freetext"
href="https://en.wikipedia.org/wiki/AirPort_Extreme"
moz-do-not-send="true">https://en.wikipedia.org/wiki/AirPort_Extreme</a>.
Similar to the Text2KGBench evaluation (<a
class="moz-txt-link-freetext"
href="https://link.springer.com/chapter/10.1007/978-3-031-47243-5_14"
moz-do-not-send="true">https://link.springer.com/chapter/10.1007/978-3-031-47243-5_14</a>),
and because the set of triples are not exhaustive for a given
sentence, to avoid false negatives, we follow a locally closed
approach by only considering the relations that are part of the
ground truth. The evaluation then uses metrics such as P, R and
F1.<br>
<br>
Go to <a class="moz-txt-link-freetext"
href="https://www.codabench.org/competitions/12936/"
moz-do-not-send="true">https://www.codabench.org/competitions/12936/</a>
and enroll to participate in this Task.<br>
<br>
# Sub-Task 3: Detecting Inconsistencies Between Texts, Tables and
Knowledge Graphs<br>
The goal of this task is to check the consistency of knowledge
extracted from tables and texts with existing triples in the
Wikidata knowledge graph. Different kind of inconsistencies will
be considered in this task. Participants to this task will be able
to report on their findings in their system paper.<br>
<br>
See the Figure at <a class="moz-txt-link-freetext"
href="https://ecladatta.github.io/images/triplet_annotation_tool.png"
moz-do-not-send="true">https://ecladatta.github.io/images/triplet_annotation_tool.png</a><br>
<br>
# Data & Evaluation:<br>
For the first 2 sub-tasks, we have released a training dataset
with ground-truth annotations, enabling participant teams to
develop machine learning-based systems, and in particular for
training purposes and for hyperparameter optimizations and
internal validations.<br>
<br>
A separate blind test dataset will remain private and be used for
ranking the submissions.<br>
<br>
Participants should register on Codabench and then enroll for each
sub-task separately (Task 1: <a class="moz-txt-link-freetext"
href="https://www.codabench.org/competitions/12776/"
moz-do-not-send="true">https://www.codabench.org/competitions/12776/</a>
and Task 2: <a class="moz-txt-link-freetext"
href="https://www.codabench.org/competitions/12936/"
moz-do-not-send="true">https://www.codabench.org/competitions/12936/</a>).
Each team are allowed a limited number of daily submissions, and
the highest achieved accuracy will be reported as the team's final
result. We encourage participants to develop open-source
solutions, to utilise and fine-tune pre-trained language models
and to experiment with LLMs of various size in zero-shot or
few-shot settings.<br>
<br>
# Challenge Important Dates:<br>
- Release of training set: 13 February 2026<br>
- Deadline for registering to the challenge: 15 March 2026<br>
- Release of test set: 24 March 2026<br>
- Submission of results: 10 April 2026<br>
- System Results & Notification of Acceptance: 17 April 2026<br>
- Submission of System Papers: 28 April 2026<br>
- Presentations @ TRIPLET Workshop: May 2026<br>
<br>
Workshop Organizers<br>
- Raphael Troncy (EURECOM, France)<br>
- Yoan Chabot (Orange, France)<br>
- Véronique Moriceau (IRIT, France)<br>
- Nathalie Aussenac-Gilles (IRIT, France)<br>
- Mouna Kamel(IRIT, France)<br>
<br>
Contact:<br>
For discussions, please use our Google Group, <a
class="moz-txt-link-freetext"
href="https://groups.google.com/g/triplet-challenge"
moz-do-not-send="true">https://groups.google.com/g/triplet-challenge</a><br>
<br>
The workshop is supported by the ECLADATTA project funded by the
French National Funding Agency ANR under the grant
ANR-22-CE23-0020.<br>
<br>
<pre class="moz-signature">--
Raphaël Troncy
EURECOM, Campus SophiaTech
Data Science Department
450 route des Chappes, 06410 Biot, France.
e-mail: <a class="moz-txt-link-abbreviated moz-txt-link-freetext"
href="mailto:raphael.troncy@eurecom.fr" moz-do-not-send="true">raphael.troncy@eurecom.fr</a> & <a
class="moz-txt-link-abbreviated moz-txt-link-freetext"
href="mailto:raphael.troncy@gmail.com" moz-do-not-send="true">raphael.troncy@gmail.com</a>
Tel: +33 (0)4 - 9300 8242
Fax: +33 (0)4 - 9000 8200
Web: <a class="moz-txt-link-freetext"
href="http://www.eurecom.fr/~troncy/" moz-do-not-send="true">http://www.eurecom.fr/~troncy/</a>
</pre>
</div>
</body>
</html>