Using word embeddings to generate data-driven human agent decision-making from natural language

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Generating replicable and empirically valid models of human decision-making is crucial for the scientific accuracy and reproducibility of agent-based models. A two-fold challenge in developing models of decision-making is a lack of high resolution and high quality behavioral data and the need for more transparent means of translating these data into models. A common and largely successful approach to modeling is hand-crafting agent decision heuristics from qualitative field interviews. This empirically-based, qualitative approach successfully incorporates contextual decision making, heterogeneous preferences, and decision strategies. However, it is labor intensive and often leads to models that are hard to replicate, thereby limiting the scale and scope over which such methods can be usefully applied. A potential solution to these problems is provided by new approaches in natural language processing, which can use textual sources ranging from field interview transcripts to unstructured data from the web to capture and represent human cognition. Here we use word embeddings, a vector-based representation of language, to create agents that reason using similarity comparison. This approach proves to be effective at mirroring theoretical expectations for human decision biases across a range of natural language decision-making tasks. We provide a proof-of-concept agent-based model that illustrates how the agents we create can be readily deployed to study cultural diffusion. The agent-based model replicates previously found results with the added benefit of qualitative interpretability. The agent architecture we propose is able to mirror human likelihood assessments from natural language and offers a new way to model agent cognitive processes for a broad array of agent-based modeling use cases.

Original languageEnglish (US)
Pages (from-to)221-242
Number of pages22
JournalGeoInformatica
Volume23
Issue number2
DOIs
StatePublished - Apr 15 2019

Fingerprint

Decision making
decision making
language
cognition
interview
cultural studies
heuristics
modeling
Mirrors
labor
Personnel
fold
lack
trend
Processing
decision

Keywords

  • Agent-based model
  • Decision-making
  • Natural language processing
  • Word embedding

Cite this

@article{686e57e3b6894ddba68b71e8f25df804,
title = "Using word embeddings to generate data-driven human agent decision-making from natural language",
abstract = "Generating replicable and empirically valid models of human decision-making is crucial for the scientific accuracy and reproducibility of agent-based models. A two-fold challenge in developing models of decision-making is a lack of high resolution and high quality behavioral data and the need for more transparent means of translating these data into models. A common and largely successful approach to modeling is hand-crafting agent decision heuristics from qualitative field interviews. This empirically-based, qualitative approach successfully incorporates contextual decision making, heterogeneous preferences, and decision strategies. However, it is labor intensive and often leads to models that are hard to replicate, thereby limiting the scale and scope over which such methods can be usefully applied. A potential solution to these problems is provided by new approaches in natural language processing, which can use textual sources ranging from field interview transcripts to unstructured data from the web to capture and represent human cognition. Here we use word embeddings, a vector-based representation of language, to create agents that reason using similarity comparison. This approach proves to be effective at mirroring theoretical expectations for human decision biases across a range of natural language decision-making tasks. We provide a proof-of-concept agent-based model that illustrates how the agents we create can be readily deployed to study cultural diffusion. The agent-based model replicates previously found results with the added benefit of qualitative interpretability. The agent architecture we propose is able to mirror human likelihood assessments from natural language and offers a new way to model agent cognitive processes for a broad array of agent-based modeling use cases.",
keywords = "Agent-based model, Decision-making, Natural language processing, Word embedding",
author = "Runck, {Bryan C} and Manson, {Steven M} and Shook, {Eric A} and Gini, {Maria L} and Jordan, {Nicholas R}",
year = "2019",
month = "4",
day = "15",
doi = "10.1007/s10707-019-00345-2",
language = "English (US)",
volume = "23",
pages = "221--242",
journal = "GeoInformatica",
issn = "1384-6175",
publisher = "Kluwer Academic Publishers",
number = "2",

}

TY - JOUR

T1 - Using word embeddings to generate data-driven human agent decision-making from natural language

AU - Runck, Bryan C

AU - Manson, Steven M

AU - Shook, Eric A

AU - Gini, Maria L

AU - Jordan, Nicholas R

PY - 2019/4/15

Y1 - 2019/4/15

N2 - Generating replicable and empirically valid models of human decision-making is crucial for the scientific accuracy and reproducibility of agent-based models. A two-fold challenge in developing models of decision-making is a lack of high resolution and high quality behavioral data and the need for more transparent means of translating these data into models. A common and largely successful approach to modeling is hand-crafting agent decision heuristics from qualitative field interviews. This empirically-based, qualitative approach successfully incorporates contextual decision making, heterogeneous preferences, and decision strategies. However, it is labor intensive and often leads to models that are hard to replicate, thereby limiting the scale and scope over which such methods can be usefully applied. A potential solution to these problems is provided by new approaches in natural language processing, which can use textual sources ranging from field interview transcripts to unstructured data from the web to capture and represent human cognition. Here we use word embeddings, a vector-based representation of language, to create agents that reason using similarity comparison. This approach proves to be effective at mirroring theoretical expectations for human decision biases across a range of natural language decision-making tasks. We provide a proof-of-concept agent-based model that illustrates how the agents we create can be readily deployed to study cultural diffusion. The agent-based model replicates previously found results with the added benefit of qualitative interpretability. The agent architecture we propose is able to mirror human likelihood assessments from natural language and offers a new way to model agent cognitive processes for a broad array of agent-based modeling use cases.

AB - Generating replicable and empirically valid models of human decision-making is crucial for the scientific accuracy and reproducibility of agent-based models. A two-fold challenge in developing models of decision-making is a lack of high resolution and high quality behavioral data and the need for more transparent means of translating these data into models. A common and largely successful approach to modeling is hand-crafting agent decision heuristics from qualitative field interviews. This empirically-based, qualitative approach successfully incorporates contextual decision making, heterogeneous preferences, and decision strategies. However, it is labor intensive and often leads to models that are hard to replicate, thereby limiting the scale and scope over which such methods can be usefully applied. A potential solution to these problems is provided by new approaches in natural language processing, which can use textual sources ranging from field interview transcripts to unstructured data from the web to capture and represent human cognition. Here we use word embeddings, a vector-based representation of language, to create agents that reason using similarity comparison. This approach proves to be effective at mirroring theoretical expectations for human decision biases across a range of natural language decision-making tasks. We provide a proof-of-concept agent-based model that illustrates how the agents we create can be readily deployed to study cultural diffusion. The agent-based model replicates previously found results with the added benefit of qualitative interpretability. The agent architecture we propose is able to mirror human likelihood assessments from natural language and offers a new way to model agent cognitive processes for a broad array of agent-based modeling use cases.

KW - Agent-based model

KW - Decision-making

KW - Natural language processing

KW - Word embedding

UR - http://www.scopus.com/inward/record.url?scp=85064219236&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064219236&partnerID=8YFLogxK

U2 - 10.1007/s10707-019-00345-2

DO - 10.1007/s10707-019-00345-2

M3 - Article

AN - SCOPUS:85064219236

VL - 23

SP - 221

EP - 242

JO - GeoInformatica

JF - GeoInformatica

SN - 1384-6175

IS - 2

ER -