Elliott Ash Profile picture
Aug 15, 2021 15 tweets 12 min read Read on X
A short thread introducing new work with @PinchOfData and @phinifa:

“Text Semantics Capture Political and Economic Narratives”

Paper: arxiv.org/abs/2108.01720
Repo: github.com/relatio-nlp/re…
Demo: colab.research.google.com/drive/1Zychi2O…
Human beings are storytellers.

It’s no wonder then that social scientists are increasingly interested in narratives -- the stories we tell in fiction, politics, and life -- and how they shape beliefs, behavior, and government policies.

e.g. @RobertJShiller
Narratives are obscure to social scientists because they consist of information, so the physical manifestations are spoken or written language.
More specifically, a narrative is an “account of a series of events, facts, etc., given in order and with the establishing of connections between them” (@OED).

Yet existing text-as-data approaches do not account for "who" does "what" to "whom".
We provide an approach for extracting narratives from text.

First, we use semantic role labeling (@ai2_allennlp) to extract the semantic roles of agent, verb, and patient. The agent is the entity that performs an action, while the patient is the entity acted upon.
The set of agents and patients is high-dimensional (typically millions of plain-text phrases).

We use named entity recognition (@spacy_io) to identify specific individuals and organizations. The remaining phrases are embedded (@gensim_py) and then clustered (@scikit_learn).
The resulting unsupervised pipeline takes in a plain-text corpus and outputs interpretable narratives representing the core claims.

In the paper, we construct narratives from floor speeches in U.S. Congress.
Some narratives are simple (e.g. “immigrants steal jobs”), but others are complex and interconnected.

We use a graph-based approach to build networks of connected entities, representing the larger narrative structures — or worldviews — expressed in a corpus.
Check out this interactive worldview graph constructed from Trump’s tweet archive: sites.google.com/view/trump-nar…

#networkx #pyvis
The pipeline has a lot going on under the hood. We provide a python package ʀᴇʟᴀᴛɪᴏ that makes it easy to use.

Repo: github.com/relatio-nlp/re…
Demo Notebook: colab.research.google.com/drive/1Zychi2O…

Special thanks to @AndreiPlamada and @ETH_SIS for indispensable contributions to the package!
In the paper, we apply the method to over a million speeches given in U.S. Congress for the period 1994-2015. We show dynamics, sentiment, and partisanship in the narratives.
In particular, we show the most divisive policy narratives.

For example, “Oil”: Democrats say “oil makes profit” while Republicans say “oil creates jobs”.

Or “Jobs”: Democrats say “companies ship jobs” while Republicans say “taxes kill jobs”.
Section 4 discusses the potential and limitations of the approach. One thing we are excited about is how ʀᴇʟᴀᴛɪᴏ could be used to support qualitative analysis of narratives, not just in social science but also in history and the humanities.

Feedback welcome!
A special shout-out to teammates @phinifa and @PinchOfData, talented upcoming economists, grand co-authors, and a delight to work with.

The project originated at #SICSS Zurich 2019!
@msalganik @chris_bail

And thanks to @snsf_ch for Spark funding.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Elliott Ash

Elliott Ash Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @ellliottt

May 19, 2023
Q: Give an example NLP task where transformers don't give the best performance.
A: Extractive summarization on legal documents!

Just posted: "Legal Extractive Summarization of U.S. Court Opinions":
arxiv.org/abs/2305.08428
(with E. Bauer, @dominsta_nlp & N. Gu, all of @ETH_en)

Repo: github.com/bauerem/legal_…

Web demo: huggingface.co/spaces/bauerem… ImageImageImage
"Extractive summarization" means excerpting the most important passages from a long document.

That is preferred to "abstractive summarization" (which paraphrases the documents in shorter form) for technical fields like law, where the wording really matters.
Read 6 tweets
May 18, 2023
Recently posted at @cepr_org: "Text Algorithms in Economics" with @StephenEKHansen (forthcoming in ARE @AnnualReviews).

Ungated version: elliottash.com/wp-content/upl…

Companion code notebooks: github.com/sekhansen/text…

& now a 🧵 on the highlights... Image
Methods covered in the "Text Algorithms" section:
1) preliminaries/ pre-processing
2) bag-of-words representation of docs
3) dimensionality reduction, e.g. topic models
4) word embeddings with local context
5) embedding sequences with attention
6) supervised learning Image
In taking these methods to economics, we organize things around "Four Measurement Problems":
Read 16 tweets
Feb 7, 2021
“In politics, when reason and emotion collide, emotion invariably wins.” -- Drew Westen, The Political Brain.

Interesting idea! What do the data say about that?

New working paper: "Emotion and Reason in Political Language", with @gloriagennaro.

PDF: bit.ly/gennaro-ash-em…
We use computational linguistics tools ("word embeddings") to map out a dimension for emotion on one pole and cognition on another pole.
The resulting geometric emotion scale is continuous and doesn't rely on the presence of particular words. In a human validation where annotators ranked pairs of sentences as more or less emotive, our metric agreed with human judgment much more often than a word-based measure.
Read 11 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(