Created by Malireddy Chanakya & Srivenkata N Mounika Somisetty & Malireddy Chaitanya
- 200 short stories
- 200 corresponding telegraphic summaries
- 50 selected abstractive summaries
- 50 selected extractive summaries each by SMMRY and RESOOMER
- 45 MCQ questions for 15 stories (3 questions/story)
- index.txt
index.txt contains a table listing out each:
1. story’s id
2. name
3. author
4. word count
5. length of the telegraphic summary
6. length of the abstractive summary
Corresponding to each story an <id>.txt file is present in the stories directory, telegraphic directory and possibly in the abstractive directory
1. A segment is defined as a continuous span of words in the source, chosen as a part of the summary.
2. A word should not be fragmented. Eg - if the word "breaking" appears in the source, the entire word should be a part of the segment and not fragments like "break".
3. Each segment should be relevant to the plot, try to advance the story and have some continuity with the preceding and the following segment.
4. Each segment extracted from a dialogue should be enclosed in quotes.
5. Each segment extracted from parentheses should be enclosed in parentheses.
6. Segments should be arranged in the same order as they appear in the story.
7. The summary should be minimal. If multiple segments mean the same thing, pick the shortest. Adjectives, adverbs, and modifiers are not to be included if they are not relevant to the plot. Extraneous facts and long descriptions are to be ignored.
8. When the segments are read in sequence the plot should be apparent and unambiguous.
1. Summaries should be written from a third party perspective. Eg - "This story is about a girl..."
2. Summaries should only discuss the plot and try to avoid inferences and opinions not immediately apparent from the story.
3. Summaries should maintain the same order of events as they occur in the source text.
@inproceedings{malireddy-etal-2018-gold,
title = "Gold Corpus for Telegraphic Summarization",
author = "Malireddy, Chanakya and
Somisetty, Srivenkata N M and
Shrivastava, Manish",
booktitle = "Proceedings of the First Workshop on Linguistic Resources for Natural Language Processing",
month = aug,
year = "2018",
address = "Santa Fe, New Mexico, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/W18-3810",
pages = "71--77",
abstract = "Most extractive summarization techniques operate by ranking all the source sentences and then select the top ranked sentences as the summary. Such methods are known to produce good summaries, especially when applied to news articles and scientific texts. However, they don{'}t fare so well when applied to texts such as fictional narratives, which don{'}t have a single central or recurrent theme. This is because usually the information or plot of the story is spread across several sentences. In this paper, we discuss a different summarization technique called Telegraphic Summarization. Here, we don{'}t select whole sentences, rather pick short segments of text spread across sentences, as the summary. We have tailored a set of guidelines to create such summaries and, using the same, annotate a gold corpus of 200 English short stories.",
}