Datasets:

feyninc
/

sencha

id stringlengths 10 10	title stringlengths 28 136	text stringlengths 5.62k 98.6k	num_sections int64 4 44
1601.02403	Argumentation Mining in User-Generated Web Discourse	# Argumentation Mining in User-Generated Web Discourse ## Abstract The goal of argumentation mining, an evolving research field in computational linguistics, is to design methods capable of analyzing people's argumentation. In this article, we go beyond the state of the art in several ways. (i) We deal with actual We...	19
1603.00968	MGNC-CNN: A Simple Approach to Exploiting Multiple Word Embeddings for Sentence Classification	# MGNC-CNN: A Simple Approach to Exploiting Multiple Word Embeddings for Sentence Classification ## Abstract We introduce a novel, simple convolution neural network (CNN) architecture - multi-group norm constraint CNN (MGNC-CNN) that capitalizes on multiple sets of word embeddings for sentence classification. MGNC-CN...	9
1603.01417	Dynamic Memory Networks for Visual and Textual Question Answering	# Dynamic Memory Networks for Visual and Textual Question Answering ## Abstract Neural network architectures with memory and attention mechanisms exhibit certain reasoning capabilities required for question answering. One such architecture, the dynamic memory network (DMN), obtained high accuracy on a variety of lang...	15
1603.01514	A Bayesian Model of Multilingual Unsupervised Semantic Role Induction	# A Bayesian Model of Multilingual Unsupervised Semantic Role Induction ## Abstract We propose a Bayesian model of unsupervised semantic role induction in multiple languages, and use it to explore the usefulness of parallel corpora for this task. Our joint Bayesian model consists of individual models for each languag...	15
1603.04513	Multichannel Variable-Size Convolution for Sentence Classification	"# Multichannel Variable-Size Convolution for Sentence Classification\n\n## Abstract\n\nWe propose M(...TRUNCATED)	10
1604.00400	Revisiting Summarization Evaluation for Scientific Articles	"# Revisiting Summarization Evaluation for Scientific Articles\n\n## Abstract\n\nEvaluation of text (...TRUNCATED)	14
1604.05781	"What we write about when we write about causality: Features of causal statements across large-scale(...TRUNCATED)	"# What we write about when we write about causality: Features of causal statements across large-sca(...TRUNCATED)	9
1605.03481	Tweet2Vec: Character-Based Distributed Representations for Social Media	"# Tweet2Vec: Character-Based Distributed Representations for Social Media\n\n## Abstract\n\nText fr(...TRUNCATED)	9
1605.07333	Combining Recurrent and Convolutional Neural Networks for Relation Classification	"# Combining Recurrent and Convolutional Neural Networks for Relation Classification\n\n## Abstract\(...TRUNCATED)	17
1606.05320	Increasing the Interpretability of Recurrent Neural Networks Using Hidden Markov Models	"# Increasing the Interpretability of Recurrent Neural Networks Using Hidden Markov Models\n\n## Abs(...TRUNCATED)	7

End of preview. Expand in Data Studio

🍵 Sencha: Scientific Paper Chunking Assessment

Scientific Challenges - A dataset for evaluating chunking algorithms on academic papers.

Overview

Sencha is designed to test how well chunking algorithms handle long-form scientific documents. It contains full-text NLP research papers with questions that require finding specific information across multiple sections.

Key Challenges

Handling structured sections (Abstract, Methods, Results, etc.)
Preserving citation context (BIBREF tags)
Managing hierarchical section headers
Chunking technical content with equations and terminology

Dataset Structure

Corpus

The corpus config contains 250 full-text NLP papers.

Column	Type	Description
`id`	string	ArXiv paper ID
`title`	string	Paper title
`text`	string	Full paper text in markdown format
`num_sections`	int	Number of sections in the paper

Questions

The questions config contains 1,146 questions about paper content.

Column	Type	Description
`id`	string	Unique question identifier
`paper_id`	string	Reference to corpus document (ArXiv ID)
`question`	string	Question about the paper content
`answer`	string	Answer to the question
`chunk-must-contain`	string	Evidence passage that answers the question

Statistics

Metric	Value
Papers	250
Questions	1,146
Avg paper length	~~26,400 chars (~~5,300 words)
Min paper length	~5,600 chars
Max paper length	~98,500 chars
Avg must-contain length	613 chars
Domain	NLP/Computational Linguistics

Usage

from datasets import load_dataset

# Load the corpus
corpus = load_dataset("chonkie-ai/sencha", "corpus", split="train")

# Load the questions
questions = load_dataset("chonkie-ai/sencha", "questions", split="train")

# Use with MTCB evaluator
from mtcb import SenchaEvaluator
from chonkie import RecursiveChunker

evaluator = SenchaEvaluator(
    chunker=RecursiveChunker(chunk_size=512),
    embedding_model="voyage-3-large"
)
result = evaluator.evaluate(k=[1, 3, 5, 10])