|
|
ARElight-0.24.0: Context Sampling of Large Text for Deep Learning RE
Nicolay Rusnachenko,
Huizhi Liang,
Maxim Kolomeets,
Lei Shi
46th European Conference on Information Retrieval, 24th-28th March, 2024
Glasgow, Scotland
(Demonstration Paper, Oral Presentation, Offline)
paper /
demo /
code /
poster /
old-landing-page
Old pre-release Version 0.22.0 related materials:
presentation /
ml-trainings-youtube /
video[rus]
AREkit-based application for a granular view onto sentiments
between entities for large document collections, including books, mass-media, Twitter/X, and more.
See our online demo
|
|
|
Advances in Sentiment Analysis of the Large Mass-Media Documents
Nicolay Rusnachenko
Glasgow IR Research Group, Sir Alwyn Williams Building, Glasgow,
G12 8QN, 22 January 2024
Scotland, United Kingdom (UK)
(Oral Presentation)
event /
video /
slides
Unlike to the similar talks in past,
in this one we overview the capabilities the most-recent instructive Large Language Models (LLM).
This list includes: Mistral, Mixtral-7B, Flan-T5, Microsoft-Phi2, LLama2, etc.
We also cover advances of LLM-fine tuning by exploiting Chain-of-Thoughts techniques to get the most out of the LLM capabilities.
|
|
|
Dialogue Agents with Literary Character Personality Traits
Junzhe Zhao,
Huizhi Liang,
Nicolay Rusnachenko
Proceedings of the The 22nd IEEE/WIC International Conference on Web Intelligence and Intelligent Agent Technology October 26-29, 2023
Venice, Italy
(Oral Presentation, Online)
doi /
paper /
eprint /
code
To enhance the engagement of chatbots, we need to imbue them with unique personalities and speech patterns.
We propose an automated system that uses character dialogues from literary works.
We used the Chinese classic, Dream of the Red Chamber.
Our system efficiently extracts dialogues and personality traits from the book, creates a personality map for each character, generates responses that reflect these traits.
|
|
|
RuSentNE-2023: Named Entity Oriented Sentiment Analysis Task
Anton Golubev,
Nicolay Rusnachenko,
Natalia Loukachevitch
Dialogue-2023, Codalab platform, 2023
(Compebtitions)
bibtex /
doi /
preprint-arXiv /
paper /
code /
codalab /
info
Participants are offered the task of extracting sentiments of three classes
(neg, pos, neutral) from news texts in relation to pre-marked entities such as
PERSON, ORG, PROFESSION, COUNTRY, NATIONALITY within a separate sentence.
|
|
|
The Rivers Trust: The application of GNNs for river water quality sensor networks
Nicolay Rusnachenko,
Andrew Johnson,
Al-Amin Bashir Bugaje,
Lingyi Yang,
Tai-Ying Lee,
Mohammad Matin Saddiqi,
Mansi Mungee (PI)
The Catalyst (Newcastle University), Newcastle Upon Tyne, 21-24'th March, 2023
England, United Kingdom (UK)
(Oral Presentation)
youtube-presentation /
event-link
Taking a facilitator Role in the
The Rivers Trust project.
The Rivers Trust company leads the
Catchment Systems Thinking Cooperative (CaSTCo) project,
which has been set up to provide a national framework for sharing water environment data
across a range of stakeholders.
To the best of our personal knowledge we were the first who experiment with the
Graph Neural Networks application (GNN)
for water quality assessment.
|
|
|
Advances in Sentiment Analysis of the Large Mass-Media Documents
Nicolay Rusnachenko
Scalalable Research Group Seminars
Urban Sciences Building (Newcastle University), Newcastle Upon Tyne, 11'th March, 2023
England, United Kingdom (UK)
(Oral Presentation)
slides
In this talk we cover the advances of machine-learning approaches in sentiment
analysis of large mass-media documents. Complements the past talk at Wolfson College (Oxford)
with the detailed description of long-input transformers, RLHF training overview.
|
|
|
Paper Overview: ETC -- Encoding Long and Structured Inputs in Transformers
Nicolay Rusnachenko
The Alan Turing Institute, London, 3'rd March, 2023
England, United Kingdom (UK)
(Oral Presentation)
/
(Impromptu)
/
(Skimming Session)
slides /
video
In this talk we examine the limitations of BERT model from the input size perspective.
To address the shortcommings, authors propose a related position embeddings in
order to implement local-window based sparse attention. To attend the distant tokens,
authors propose a Global-Attention mechanism in addition to the sparsed one for the main input.
Another main contribution is input structuring in attention mechanism.
|
|
|
Sentiment Attitude Extraction Resources Translation
Nicolay Rusnachenko
(Github Repository)
code /
task
As a task, Sentiment Attitude Extraction is devoted to extraction of the sentiment connections
from subjects towards objects mentioned in texts, usually analytical articles.
This task has been originally proposed and becomes a part of the studies in RuSentRel dataset,
in which texts are written in Russian.
|
|
Awesome Sentiment Attitude Extraction
This is a curated list of works devoted to sentiment attitude extraction domain.
The latter considered as an area of studies, in which sentiment analysis
and relation extraction are inextricably linked
|
|
Official RuSentRel-1.1 Leaderboard
Nicolay Rusnachenko
(Github Repository)
This repository is an official results benchmark for automatic sentiment attitude
extraction task within RuSentRel-1.1
collection, for the following models:
conventional approaches (SVM, RandomForest, kNN, NB, Gradient Boosting),
CNN-based networks,
RNN-based networks,
Attentive models,
BERT-based language models.
|
|
Advances in Sentiment Attitude Extraction Task
Nicolay Rusnachenko
Introduction to Humanistic Informatics course
Denmark, Odense, November 26'th, 2020
(Online Lecture)
video /
slides
This lecture is devoted to the works that were done within last 4 years,
which is related sentiment attitude extraction task.
|
|