iSee journey so far…

May 2020: Short Proposal Accepted

November 2020: Full Proposal Accepted

March 2021: Official iSee Kick-off

March 2021: The website is now online

April 2021: Welcome meeting

May 2021: Y1Q1 Deliverables completed (community created, XE evaluation instruments survey)

June 2021: iSee application at SICSA XAI workshop

July 2021 : Milestone0 deliverables completed !

October 2021: Data Management Plan completed!

October 2021: XAI Usability and Evaluation Workshops at TF Network Conference on “Data Science – The beating heart of AI”

November 2021: Case base with foundational explainers

December 2021: Milestone1 deliverables completed ! iSee v1. iSeeOnto model tested on simple cases, Retrieval-only CBR engine, Cockpit interface

December 2021 : new use case = Connected Home Care (BT), TotalEnergies, result from Engagement plan and co-creation activities

January 2022 : iSeeOnto ontology v1 is now online !

February 2022 : Y1 Consortium meeting in Madrid.
Telecom use case represented in iSee

March 2022 : iSee Cockpit tool V1 and Explainers catalogue V1 completed !

March 2022 : Chistera annual webinar
Launch of XAI-CBR challenge 

July 2022 : Use case Design cockpit online on iSee platform
Explainer library of iSee released online
Explanation strategy retrieval first iteration
Healthcare and Telecom UC entered into the cockpit
New use cases identified with external partners

September 2022 : iSee consortium meeting in Nancy
XCBR workshop and XCBR challenge organised with iSee team
3 papers from iSee team at ICCBR 2022 conference

October – November 2022 : We are working on the integration of all the components into one single platform: front-end cockpit, behaviour tree editor and visualiser, CBR retrieval engine returning a recommendation of explanation strategies in Behaviour tree format to the cockpit, Behaviour tree executor to generate explanation from the recommended strategy, and link concepts across iSeeOnto, use case and CBR ontology.

December 2022 : Deliverables for December 2022 – Milestone 2 – are published on Zenodo DOI 10.5281/zenodo.7472464 !

January 2023 : New Release of iSee cockpit including user guidelines and tooltips.
2 real world use cases added to iSee : Home resident care monitoring (BT) using time series dataset, Radiology diagnosis using images dataset (JIVA).

iSee organises the XCBR challenge and tutorial at ICCBR 2023 (July at Aberdeen, Scotland)
We propose a tutorial on iSee as part of the Industry day at ICCBR’23 (ICCBR 23 (rgu.ac.uk))

We won the bronze medal in the VIDEO contest of CHISTERA annual seminar in

 Bratislava, April 2023

17th July 2023. ICCBR 2023 in Aberdeen. XCBR challenge, iSee tutorial and online course 

on how to use iSee… and in person project meeting

July 2023. Published papers at ICCBR 2023 in Aberdeen, higlighting the ground structure (BT) used at the reuse of explanation strategy for similar intent, and the future of iSee to the CBR community.

September 2023. New use cases  Production  line anomaly detection with Bosch(images data) and Brain guard (UCM)

September 2023. New Release of iSee : cockpit, questionnaires, BT editor, extension of the ontology of Explanation experience with evaluating user question. New function allowing the reuse of explainers within a Behaviour Tree (process to reuse an end-user explanation strategy within an existing strategy), and allowing the reuse of behaviour Tree (explainer+evaluating user question).

October 2023. We’ve included 55 reusable explainers in our iSee library ! check the library link under “project”

1st October 2023. Milestone 3 delivered on Zenodo https://doi.org/10.5281/zenodo.8414582

24th October 2023. Integration meeting (on line)

November 2023.  New metrics dashboard showing the compliance based on a new scale for the purpose of evaluating the quality of explanation experiences. Content Validation Study with RGU and leading XAI practitioners confirms the new scale. The study asks participants to review a curated list of statements based on their Relevance and Clarity.

1st January 2024. New use cases on Health :Brainguard, ECG(time series)

15th January 2024 – New reuse functionalities were created and released in the BT editor. Users can edit explanation strategies as they need and want. Recommendation of suitable explainers for substitution is based on explainer previously used and positively positively.

31/01/2024 – New functionalities on testing BTs were created and released in the BT editor (Automatic reuse). Users can check their strategies’ applicability for their use case and check the BT structure’s suitability.

12/02/2024 –Final state of the library of explainers. We have 70 explainers exposed as a service.

1st March 2024 – we completed 3 real world use cases. Evaluation experiments invite :
Bosch Link
Telecom link 
Jiva link 

12/02/2024 to 30/03/2024 Evaluation experiments with BT France IT community in BT. Participants were asked to walkthrough the iSee chatbot to see an AI recommendation on data instances, request explanations from iSee and evaluation their meaningfulness. 2 real use cases : Health radiograph fracture and Sensor anomaly use case (image data type)

Final project date. 31st March 2024. 3 months extension.

March- May 2024 – 3 Satellite projects start.
RGU-IIT-BTF : Explaining CBR-RAG model in LLM for legal question answering knowledge extraction
BTF : Generative AI(LLM) to build human intelligible explanation beyond iSee Chatbot, to improve explanation experience.
UCM-NTNU : Evaluation metrics for counterfactual and feature-based explainers including PertCF

March 2024-  Certification path

14th April 2024 iSee Major Release 1.0 (academic product) , including terms of  use  , video guidelines

15th April2024-  Chistera annual  seminar, poster participation and  video contest  Winner. published on youtube (click on the image to watch it)

March – May 2024 Members of iSee consortium (UCM, UCC, BTF) with Instituto Tecnologico de Merida organise the ICCBR 2024 international conference (July 2024 in Merida, Mexico). iSee hosts the 6th XCBR workshop there.

4th July 2024 We managed to wind the Best Paper Award of ICCBR 2024 ! The authors team is composed of Robert Gordon University, Informatics Institute of Technology (IIT Campus) and BT France. The conference is the 32nd International Conference on Case-Based Reasoning, in Mexico.
We were the first to integrate episodic memory in the form of a legal case base with retrieval augmented generation to improve accuracy for LLMs.