arxiv: 2604.27673 · v1 · submitted 2026-04-30 · 💻 cs.AI · cs.CY· cs.HC· cs.LG· cs.SI

Recognition: unknown

The TEA Nets framework combines AI and cognitive network science to model targets, events and actors in text

Sebastiano Franchini , Alexis Carrillo , Edoardo Sebastiano De Duro , Riccardo Improta , Ali Aghazadeh Ardebili , Massimo Stella

Authors on Pith no claims yet

Pith reviewed 2026-05-07 07:53 UTC · model grok-4.3

classification 💻 cs.AI cs.CYcs.HCcs.LGcs.SI

keywords target-event-agent networkscognitive network sciencetext analysisemotion detectionconspiracy narrativesLLM psychotherapy transcriptssemantic frame analysisAI text parsing

0 comments

The pith

Target-Event-Agent Networks extract agents, events, and targets from text to map emotional and semantic structures in narratives.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Target-Event-Agent Networks, or TEA Nets, which use artificial intelligence to identify agents who perform events on targets and then connect these elements into networks for further analysis. Grounded in cognitive network science, the framework turns raw text into structured representations that support interpretable measurements of emotion, syntax, and semantics. The authors demonstrate the approach on thousands of conspiracy texts and hundreds of psychotherapy transcripts written by humans and large language models. In the conspiracy corpus, highly conspiratorial narratives link personal pronouns to the same actions twice as often as less conspiratorial ones and connect people to anger-eliciting actions above random levels. The same networks detect that one language model expresses sadness at lower intensity than human writers do in therapy dialogues. A reader would care because the method supplies a concrete, network-based lens on who does what to whom that could make automated text analysis more aligned with human understanding of stories.

Core claim

We introduce Target-Event-Agent Networks (TEA Nets) as a computational framework to extract subjects (Agents), verbs (Events), and objects (Targets) from texts. Grounded in cognitive network science and artificial intelligence, TEA Nets are implemented as an open-source Python library. We test TEA Nets in three case studies, demonstrating the framework's ability to perform interpretable emotion detection, semantic frame analyses, and linguistic inquiries across conspiracy texts and textual responses generated by LLMs. In the LOCO conspiracy corpus, TEA Nets revealed that highly conspiratorial narratives linked personal pronouns with the same actions twice as frequently as low-similarity ones

What carries the argument

Target-Event-Agent Networks (TEA Nets), which use AI to identify agents, events, and targets in text and construct networks from these elements to enable statistical analysis of emotional and semantic patterns.

If this is right

Highly conspiratorial narratives link personal pronouns with the same actions twice as frequently as low-conspiracy narratives.
High-conspiracy texts connect person-focused elements through actions that elicit anger above random baseline levels while low-conspiracy texts emphasize scientific actors.
LLM-generated psychotherapy responses express sadness with higher frequency than random expectations but with lower emotional intensity than human responses.
The networks support semantic frame analyses and linguistic inquiries that distinguish emotional patterns across narrative types.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same network construction could be applied to political speeches or news articles to quantify how different sources emphasize personal versus institutional actions.
Pairing TEA Nets extraction with improved language models over time would likely raise the accuracy of agent and event detection and strengthen the reliability of the downstream statistics.
Testing the framework on non-English texts would reveal whether the observed emotional network patterns depend on language-specific features or hold more generally.

Load-bearing premise

The automated identification of targets, events, and agents in text must be accurate and meaningful enough that the reported statistical differences in the networks reflect real properties of the source narratives rather than extraction errors.

What would settle it

A manual annotation of a random sample of the conspiracy and therapy texts that finds the automated agent-event-target labels match human judgments in fewer than 60 percent of cases, or a recalculation of the network statistics with human-corrected labels that eliminates the reported differences in pronoun-action frequencies and emotional intensities.

read the original abstract

We introduce Target-Event-Agent Networks (TEA Nets) as a computational framework to extract subjects (``Agents"), verbs (``Events"), and objects (``Targets") from texts. Grounded in cognitive network science and artificial intelligence, TEA Nets are implemented as an open-source Python library. We test TEA Nets in three case studies, demonstrating the framework's ability to perform interpretable emotion detection, semantic frame analyses, and linguistic inquiries across conspiracy texts and textual responses generated by LLMs. In the LOCO conspiracy corpus, TEA Nets revealed that highly conspiratorial narratives (4,227 texts) linked personal pronouns (``I", ``you", ``we") with the same actions twice as frequently as low-similarity conspiracy narratives. High-conspiracy narratives connected person-focused elements (``you", ``people") through actions eliciting anger above the random baseline ($z = 2.63, p < .05$), a trend absent in low-similarity conspiracy narratives, which emphasized scientific actors (``researcher", ``scientist"). In the HOPE and CounseLLMe datasets of 212 (human) and 200 (LLM-based) psychotherapy transcripts, respectively, TEA Nets highlighted emotional differences. When expressing feelings, Claude 3 Haiku, GPT-3.5, and humans used sad words with higher frequency than random expectations but Haiku expressed sadness with lower emotional intensity than humans ($U = 1243.5, p = .036$). We discuss these differences in the context of psychotherapy training on LLM-simulated patients. Our results show that Target-Event-Agent Networks can extract relevant emotional, syntactic, and semantic insights from narratives, opening new avenues for text analysis with cognitive network science.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

TEA Nets provides a fresh open-source framework for agent-event-target networks in text analysis, with some solid applications, but the lack of extraction validation leaves the statistical claims on shaky ground.

read the letter

The key takeaway is that this work gives us a Python library for turning text into Target-Event-Agent networks using a mix of AI and cognitive network ideas, with tests on conspiracy texts and LLM versus human psychotherapy responses. The stats they report look interesting on the surface, but the extraction quality isn't checked against human judgments. What stands out as new is the specific framing of TEA Nets and its application to these particular datasets. They show that in high-conspiracy narratives from the LOCO corpus, personal pronouns connect to the same actions twice as often as in low-similarity ones. They also find anger-eliciting links above baseline only in high-conspiracy texts, while low ones focus on scientific actors. In the therapy data, they note differences in sadness word intensity between Claude 3 Haiku and humans. The paper does a decent job making the library open-source and applying it to external corpora without obvious circularity. The results are presented with concrete numbers like z=2.63 and U=1243.5. The main weakness is the missing validation for the core step. The abstract and likely the methods don't include precision, recall, or agreement scores for how well the system identifies agents, events, and targets on these texts. If the underlying dependency parser or LLM prompt tends to over-link first-person pronouns or certain emotion words, then the network statistics and the reported differences could be artifacts rather than reflections of the narratives themselves. They should have run a small human annotation study on a sample to confirm the triples are reliable. This paper is aimed at computational linguists and researchers in social science who use text mining for topics like conspiracy beliefs or evaluating AI in clinical simulations. Someone looking for a new network-based approach to emotion and semantic analysis in stories would find it relevant. It has enough of a novel tool and clear applications that it should go to peer review rather than a desk reject. Referees will probably ask for the validation data and more implementation details, but the idea has merit.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces Target-Event-Agent Networks (TEA Nets), a computational framework that uses AI to extract Agents (subjects), Events (verbs), and Targets (objects) from text. Implemented as an open-source Python library and grounded in cognitive network science, the framework is evaluated in three case studies. On the LOCO conspiracy corpus (4,227 texts), it reports that high-conspiracy narratives link personal pronouns with actions twice as frequently as low-similarity ones and connect person-focused elements via anger-eliciting actions above baseline (z=2.63, p<.05), unlike low-conspiracy texts that emphasize scientific actors. On HOPE (212 human) and CounseLLMe (200 LLM) psychotherapy transcripts, it finds that Claude 3 Haiku, GPT-3.5, and humans use sad words above random expectations but Haiku expresses sadness with lower intensity than humans (U=1243.5, p=.036). The authors conclude that TEA Nets enable interpretable extraction of emotional, syntactic, and semantic insights from narratives.

Significance. If the core extraction step can be shown to be reliable, TEA Nets would constitute a useful bridge between AI/NLP pipelines and cognitive network science, offering an interpretable, graph-based lens on narrative structure with applications in computational social science, conspiracy research, and evaluation of LLM-generated therapeutic content. The open-source library is a positive feature for reproducibility and extension. The reported statistical contrasts are concrete and falsifiable in principle, which strengthens the contribution relative to purely qualitative text-analysis frameworks.

major comments (2)

[Abstract and Case Studies] Abstract and Case Studies sections: The headline claim that TEA Nets 'extract relevant emotional, syntactic, and semantic insights' rests on the premise that the automated identification of Agents, Events, and Targets is sufficiently accurate and unbiased. No precision, recall, inter-annotator agreement, or gold-standard validation figures are provided for the extraction pipeline on the LOCO, HOPE, or CounseLLMe corpora. Consequently the reported z=2.63 (anger links) and U=1243.5 (sadness intensity) statistics cannot be confidently interpreted as reflecting narrative properties rather than parser/LLM artifacts.
[Implementation] Implementation and Methods: The manuscript states that TEA Nets are realized as an open-source Python library but supplies insufficient technical detail on the concrete AI/NLP components (dependency parser, semantic role labeling model, or LLM prompt) used to produce the triples. Without this information it is impossible to diagnose potential systematic biases (e.g., over-linking of first-person pronouns or under-identification of scientific actors) that could drive the cross-condition differences reported in the case studies.

minor comments (2)

[Abstract] Abstract: A direct citation or URL for the open-source library repository would strengthen the reproducibility claim.
[Discussion] Discussion: The psychotherapy case study would benefit from reporting effect sizes or confidence intervals alongside the Mann-Whitney U statistic to allow readers to gauge practical significance of the sadness-intensity difference.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thoughtful and constructive review. The comments correctly identify areas where additional transparency and validation are needed to strengthen the interpretability of the TEA Nets results. We address each major comment below and commit to revisions that will incorporate the requested details without altering the core claims of the manuscript.

read point-by-point responses

Referee: [Abstract and Case Studies] Abstract and Case Studies sections: The headline claim that TEA Nets 'extract relevant emotional, syntactic, and semantic insights' rests on the premise that the automated identification of Agents, Events, and Targets is sufficiently accurate and unbiased. No precision, recall, inter-annotator agreement, or gold-standard validation figures are provided for the extraction pipeline on the LOCO, HOPE, or CounseLLMe corpora. Consequently the reported z=2.63 (anger links) and U=1243.5 (sadness intensity) statistics cannot be confidently interpreted as reflecting narrative properties rather than parser/LLM artifacts.

Authors: We agree that the absence of quantitative validation metrics for the extraction pipeline limits the strength of the interpretive claims. The current manuscript presents the case studies as exploratory demonstrations rather than definitive causal analyses, but the referee is correct that readers require evidence of extraction reliability to evaluate the reported statistical contrasts. In the revised manuscript we will add a dedicated Validation subsection in Methods. This will report precision, recall, and F1 scores for Agent, Event, and Target identification on a manually annotated random sample of 200 texts drawn proportionally from the LOCO, HOPE, and CounseLLMe corpora. We will also include inter-annotator agreement (Cohen’s kappa) between two independent annotators and a brief error analysis highlighting common failure modes. The statistical results will be reframed with explicit caveats referencing these performance figures, and we will note that the observed patterns are conditional on the accuracy of the underlying triples. revision: yes
Referee: [Implementation] Implementation and Methods: The manuscript states that TEA Nets are realized as an open-source Python library but supplies insufficient technical detail on the concrete AI/NLP components (dependency parser, semantic role labeling model, or LLM prompt) used to produce the triples. Without this information it is impossible to diagnose potential systematic biases (e.g., over-linking of first-person pronouns or under-identification of scientific actors) that could drive the cross-condition differences reported in the case studies.

Authors: We accept that the technical description of the extraction pipeline is currently too high-level. The revised Implementation section will specify the exact components: dependency parsing is performed with spaCy (en_core_web_lg model, version 3.7), semantic role labeling uses the AllenNLP SRL model, and an optional GPT-4o-mini refinement step is applied only to sentences where SRL confidence falls below 0.7. We will include the full prompt template used for the LLM component, the rule-based post-processing steps that convert SRL output into Agent-Event-Target triples, and a pseudocode overview of the pipeline. The open-source repository will be updated with these specifications in the README and a new technical appendix. These additions will allow independent assessment of potential biases such as pronoun over-representation or domain-specific actor under-detection. revision: yes

Circularity Check

0 steps flagged

TEA Nets applies independent extraction to external corpora with no definitional or fitted circularity

full rationale

The paper defines TEA Nets as a framework for extracting Agent-Event-Target triples via AI/NLP methods, implements it as an open-source library, and applies the resulting networks to three independent external datasets (LOCO conspiracy texts, HOPE human psychotherapy transcripts, CounseLLMe LLM-generated transcripts). Reported statistics such as pronoun-action co-occurrence rates, z-scores for anger-eliciting links, and Mann-Whitney U tests on sadness intensity are computed directly from the processed external texts rather than from any fitted parameters or self-referential definitions. No equations, uniqueness theorems, or self-citations are invoked as load-bearing premises, and no ansatz or known result is smuggled in or renamed. The derivation chain remains self-contained against external benchmarks without reducing any central claim to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The central claim rests on the assumption that automated extraction of agent-event-target triples accurately reflects cognitive and semantic structures. Since only the abstract is available, the ledger is limited to elements explicitly invoked in the summary. Full implementation details would likely reveal additional free parameters in network construction or thresholding.

axioms (2)

domain assumption The extraction of agents, events, and targets from text accurately reflects cognitive structures.
Central to applying network science metrics to the extracted triples for emotion and semantic analysis.
standard math Statistical comparisons to random baselines are appropriate for identifying significant patterns in the networks.
Invoked in the reported z-score and Mann-Whitney U test calculations.

invented entities (1)

TEA Nets framework no independent evidence
purpose: To model and analyze targets, events, and actors in text using cognitive networks and AI.
Newly introduced computational framework without external validation mentioned beyond the case studies.

pith-pipeline@v0.9.0 · 5651 in / 1657 out tokens · 80819 ms · 2026-05-07T07:53:19.486722+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

91 extracted references · 15 canonical work pages · 1 internal anchor

[1]

write newline

" write newline " cite write " FUNCTION editor.postfix editor num.names #1 > "( )" "( )" if FUNCTION editor.trans.postfix editor num.names #1 > "( )" "( )" if FUNCTION trans.postfix translator num.names #1 > "( )" "( )" if FUNCTION authors.editors.reflist.apa5 'field := 'dot := field num.names 'numnames := numnames 'format.num.names := format.num.names na...
[2]

, Ciringione, L

abramski2024voices APACrefauthors Abramski, K. , Ciringione, L. , Rossetti, G. Stella, M. APACrefauthors \ 2024 . Voices of rape: Cognitive networks link passive voice usage to psychological distress in online narratives Voices of rape: Cognitive networks link passive voice usage to psychological distress in online narratives . Computers in Human Behavior...

2024
[3]

, Mata, R

aeschbach2025measuring APACrefauthors Aeschbach, S. , Mata, R. Wulff, D.U. APACrefauthors \ 2025 . Measuring individual semantic networks: A simulation study Measuring individual semantic networks: A simulation study . PLoS One 20 8 e0328712,

2025
[4]

APACrefauthors \ 2012

aitchison2012words APACrefauthors Aitchison, J. APACrefauthors \ 2012 . Words in the mind: An introduction to the mental lexicon Words in the mind: An introduction to the mental lexicon . John Wiley & Sons

2012
[5]

, Farrell, L

barrett2008evidence APACrefauthors Barrett, P.M. , Farrell, L. , Pina, A.A. , Peris, T.S. Piacentini, J. APACrefauthors \ 2008 . Evidence-based psychosocial treatments for child and adolescent obsessive--compulsive disorder Evidence-based psychosocial treatments for child and adolescent obsessive--compulsive disorder . Journal of Clinical Child & Adolesce...

2008
[6]

, Smith, S.G

basile2022national APACrefauthors Basile, K.C. , Smith, S.G. , Kresnow, M j. , Khatiwada, S. Leemis, R.W. APACrefauthors \ 2022 . The national intimate partner and sexual violence survey: 2016/2017 report on sexual violence The national intimate partner and sexual violence survey: 2016/2017 report on sexual violence . Centers for Disease Control and Preve...

2022
[7]

, Musacchio, E

basile2023llamantino APACrefauthors Basile, P. , Musacchio, E. , Polignano, M. , Siciliani, L. , Fiameni, G. Semeraro, G. APACrefauthors \ 2023 . Llamantino: Llama 2 models for effective text generation in italian language Llamantino: Llama 2 models for effective text generation in italian language . arXiv preprint arXiv:2312.09993 ,

work page arXiv 2023
[8]

\ Schulz, E

binz2023using APACrefauthors Binz, M. \ Schulz, E. APACrefauthors \ 2023 . Using cognitive psychology to understand GPT-3 Using cognitive psychology to understand gpt-3 . Proceedings of the National Academy of Sciences 120 6 e2218523120,

2023
[9]

\ McGann, M

birhane2024large APACrefauthors Birhane, A. \ McGann, M. APACrefauthors \ 2024 . Large models of what? Mistaking engineering achievements for human linguistic agency Large models of what? mistaking engineering achievements for human linguistic agency . Language Sciences 106 101672,

2024
[10]

APACrefauthors \ 2001

bohner2001writing APACrefauthors Bohner, G. APACrefauthors \ 2001 . Writing about rape: Use of the passive voice and other distancing text features as an expression of perceived responsibility of the victim Writing about rape: Use of the passive voice and other distancing text features as an expression of perceived responsibility of the victim . British J...

2001
[11]

, Ashokkumar, A

boyd2022liwc APACrefauthors Boyd, R.L. , Ashokkumar, A. , Seraj, S. Pennebaker, J.W. APACrefauthors \ 2022 1 . The development and psychometric properties of LIWC-22. The development and psychometric properties of liwc-22. Austin, TX University of Texas at Austin . APACrefURL https://www.liwc.app APACrefURL

2022
[12]

, Ashokkumar, A

boyd2022development APACrefauthors Boyd, R.L. , Ashokkumar, A. , Seraj, S. Pennebaker, J.W. APACrefauthors \ 2022 2 . The development and psychometric properties of LIWC-22 The development and psychometric properties of liwc-22 . Austin, TX: University of Texas at Austin 10 1--47,

2022
[13]

, Warriner, A.B

brysbaert2014concreteness APACrefauthors Brysbaert, M. , Warriner, A.B. Kuperman, V. APACrefauthors \ 2014 . Concreteness ratings for 40 thousand generally known English word lemmas Concreteness ratings for 40 thousand generally known english word lemmas . Behavior research methods 46 904--911,

2014
[14]

, Roske, S.F

carrillo2025textual APACrefauthors Carrillo, A. , Roske, S.F. , Ianov-Vitanov, R. , Perinelli, E. , Grecucci, A. Stella, M. APACrefauthors \ 2025 . Textual forma mentis networks bridge language structure, emotional content and psychopathology levels in adolescents Textual forma mentis networks bridge language structure, emotional content and psychopatholo...

work page arXiv 2025
[15]

chen2023llm APACrefauthors Chen, S. , Wu, M. , Zhu, K.Q. , Lan, K. , Zhang, Z. Cui, L. APACrefauthors \ 2023 . LLM-empowered chatbots for psychiatrist and patient simulation: application and evaluation Llm-empowered chatbots for psychiatrist and patient simulation: application and evaluation . arXiv preprint arXiv:2305.13614 ,

work page arXiv 2023
[16]

, Yan, Z

cheng2024potential APACrefauthors Cheng, N. , Yan, Z. , Wang, Z. , Li, Z. , Yu, J. , Zheng, Z. Han, W. APACrefauthors \ 2024 . Potential and Limitations of LLMs in Capturing Structured Semantics: A Case Study on SRL Potential and limitations of llms in capturing structured semantics: A case study on srl . International Conference on Intelligent Computing ...

2024
[17]

, Warner-Willich, J

citraro2023hypergraph APACrefauthors Citraro, S. , Warner-Willich, J. , Battiston, F. , Siew, C.S. , Rossetti, G. Stella, M. APACrefauthors \ 2023 . Hypergraph models of the mental lexicon capture greater information than pairwise networks for predicting language learning Hypergraph models of the mental lexicon capture greater information than pairwise ne...

2023
[18]

\ Pleydell-Pearce, C.W

conway2000construction APACrefauthors Conway, M.A. \ Pleydell-Pearce, C.W. APACrefauthors \ 2000 . The construction of autobiographical memories in the self-memory system. The construction of autobiographical memories in the self-memory system. Psychological review 107 2 261,

2000
[19]

, Pendse, S.R

de2023benefits APACrefauthors De Choudhury, M. , Pendse, S.R. Kumar, N. APACrefauthors \ 2023 . Benefits and harms of large language models in digital mental health Benefits and harms of large language models in digital mental health . arXiv preprint arXiv:2311.14693 ,

work page arXiv 2023
[20]

, Improta, R

de2024network APACrefauthors De Duro, E.S. , Improta, R. Stella, M. APACrefauthors \ 2024 . Network science highlights the emotional structure of counselling conversations simulated by Large Language Models and humans Network science highlights the emotional structure of counselling conversations simulated by large language models and humans . Proceedings...

2024
[21]

, Improta, R

deduro2025counsellme APACrefauthors De Duro, E.S. , Improta, R. Stella, M. APACrefauthors \ 2025 . Introducing CounseLLMe: A dataset of simulated mental health dialogues for comparing LLMs like Haiku, LLaMAntino and ChatGPT against humans Introducing counsellme: A dataset of simulated mental health dialogues for comparing llms like haiku, llamantino and c...

2025
[22]

, Bhensdadia, C

Detroja_2023_relation_extraction APACrefauthors Detroja, K. , Bhensdadia, C. Bhatt, B.S. APACrefauthors \ 2023 . A survey on relation extraction A survey on relation extraction . Intelligent Systems with Applications 19 200244,

2023
[23]

, Chang, M W

devlin2019bert APACrefauthors Devlin, J. , Chang, M W. , Lee, K. Toutanova, K. APACrefauthors \ 2019 . Bert: Pre-training of deep bidirectional transformers for language understanding Bert: Pre-training of deep bidirectional transformers for language understanding . Proceedings of the 2019 conference of the North American chapter of the association for co...

2019
[24]

APACrefauthors \ 2019

doczi2019overview APACrefauthors D \'o czi, B. APACrefauthors \ 2019 . An overview of conceptual models and theories of lexical representation in the mental lexicon An overview of conceptual models and theories of lexical representation in the mental lexicon . The Routledge handbook of vocabulary studies 46--65,

2019
[25]

APACrefauthors \ 2013

doell2013word APACrefauthors Doell, K. APACrefauthors \ 2013 . The word feel as an indicator of enacted social support in personal relationships The word feel as an indicator of enacted social support in personal relationships . International Journal of Psychological Studies 5 4 107,

2013
[26]

\ Holtzman, N.S

edwards2017meta APACrefauthors Edwards, T. \ Holtzman, N.S. APACrefauthors \ 2017 . A meta-analysis of correlations between depression and first person singular pronoun use A meta-analysis of correlations between depression and first person singular pronoun use . Journal of Research in Personality 68 63--68,

2017
[27]

spaCy : Industrial-strength Natural Language Processing in Python

spacy_website APACrefauthors Explosion APACrefauthors \ 2025 . spaCy : Industrial-strength Natural Language Processing in Python. spaCy : Industrial-strength natural language processing in python. https://spacy.io/ . Accessed April 23, 2025

2025
[28]

, Skoven Hinge, J

fazzi2025don APACrefauthors Fazzi, G.F. , Skoven Hinge, J. , Heinrich, S. Burelli, P. APACrefauthors \ 2025 . Don't Get Too Excited-Eliciting Emotions in LLMs Don't get too excited-eliciting emotions in llms . Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems Proceedings of the extended abstracts of the chi ...

2025
[29]

, Keith, K.A

feder2022causal APACrefauthors Feder, A. , Keith, K.A. , Manzoor, E. , Pryzant, R. , Sridhar, D. , Wood-Doughty, Z. others APACrefauthors \ 2022 . Causal inference in natural language processing: Estimation, prediction, interpretation and beyond Causal inference in natural language processing: Estimation, prediction, interpretation and beyond . Transactio...

2022
[30]

a r, D. , Geissler, D. , Schweisthal, J. , Pr \

feuerriegel2025using APACrefauthors Feuerriegel, S. , Maarouf, A. , B \"a r, D. , Geissler, D. , Schweisthal, J. , Pr \"o llochs, N. others APACrefauthors \ 2025 . Using natural language processing to analyse text data in behavioural science Using natural language processing to analyse text data in behavioural science . Nature Reviews Psychology 1--16,

2025
[31]

, Roozenbeek, J

Fong_2021_language_conspiracy APACrefauthors Fong, A. , Roozenbeek, J. , Goldwert, D. , Rathje, S. Van Der Linden, S. APACrefauthors \ 2021 . The language of conspiracy: A psychological analysis of speech used by conspiracy theorists and their followers on Twitter The language of conspiracy: A psychological analysis of speech used by conspiracy theorists ...

2021
[32]

APACrefauthors \ 2025

gagliardi_2023_conspiracy_bias_beliefs APACrefauthors Gagliardi, L. APACrefauthors \ 2025 . The role of cognitive biases in conspiracy beliefs: A literature review The role of cognitive biases in conspiracy beliefs: A literature review . Journal of Economic Surveys 39 1 32--65,

2025
[33]

, Tardelli, S

Gambini_2024_anatomy_conspiracy_theorists APACrefauthors Gambini, M. , Tardelli, S. Tesconi, M. APACrefauthors \ 2024 . The anatomy of conspiracy theorists: unveiling traits using a comprehensive twitter dataset The anatomy of conspiracy theorists: unveiling traits using a comprehensive twitter dataset . Computer Communications 217 25--40,

2024
[34]

, Vicente Ripoll, M.A

garcia2024enhancing APACrefauthors Garc \' a-Torres, D. , Vicente Ripoll, M.A. , Fern \'a ndez Peris, C. Mira Solves, J.J. APACrefauthors \ 2024 . Enhancing Clinical Reasoning with Virtual Patients: A Hybrid Systematic Review Combining Human Reviewers and ChatGPT Enhancing clinical reasoning with virtual patients: A hybrid systematic review combining huma...

2024
[35]

, Ghanem, B

Giachanou2023consp_propagators_linguistic APACrefauthors Giachanou, A. , Ghanem, B. Rosso, P. APACrefauthors \ 2023 . Detection of conspiracy propagators using psycho-linguistic characteristics Detection of conspiracy propagators using psycho-linguistic characteristics . Journal of Information Science 49 1 3--17,

2023
[36]

APACrefauthors \ 1974

gollob1974subject APACrefauthors Gollob, H.F. APACrefauthors \ 1974 . The subject-verb-object approach to social cognition. The subject-verb-object approach to social cognition. Psychological Review 81 4 286,

1974
[37]

greenberg1963some APACrefauthors Greenberg, J.H. \ . APACrefauthors \ 1963 . Some universals of grammar with particular reference to the order of meaningful elements Some universals of grammar with particular reference to the order of meaningful elements . Universals of language 2 73--113,

1963
[38]

, Shang, G

guo2024curious APACrefauthors Guo, Y. , Shang, G. Clavel, C. APACrefauthors \ 2024 . Benchmarking linguistic diversity of large language models Benchmarking linguistic diversity of large language models . arXiv preprint arXiv:2412.10271 ,

work page arXiv 2024
[39]

, Schult, D.A

Aric2008NetworkX APACrefauthors Hagberg, A.A. , Schult, D.A. Swart, P.J. APACrefauthors \ 2008 . Exploring Network Structure, Dynamics, and Function using NetworkX Exploring network structure, dynamics, and function using networkx . G. Varoquaux, T. Vaught \ J. Millman\ ( ), Proceedings of the 7th Python in Science Conference Proceedings of the 7th python...

2008
[40]

arXiv preprint arXiv:2303.13988 , year=

hagendorff2023machine APACrefauthors Hagendorff, T. APACrefauthors \ 2023 . Machine psychology: Investigating emergent capabilities and behavior in large language models using psychological methods Machine psychology: Investigating emergent capabilities and behavior in large language models using psychological methods . arXiv preprint arXiv:2303.13988 1 ,

work page arXiv 2023
[41]

, Fischer, N

haim_2024_formamentis APACrefauthors Haim, E. , Fischer, N. , Citraro, S. , Rossetti, G. Stella, M. APACrefauthors \ 2026 . Forma mentis networks predict creativity ratings of short texts via interpretable artificial intelligence in human and AI-simulated raters Forma mentis networks predict creativity ratings of short texts via interpretable artificial i...

2026
[42]

\ Stella, M

haim2026cognitive APACrefauthors Haim, E. \ Stella, M. APACrefauthors \ 2026 . Cognitive Networks for Knowledge Modeling: A Gentle Introduction for Data-and Cognitive Scientists Cognitive networks for knowledge modeling: A gentle introduction for data-and cognitive scientists . Wiley Interdisciplinary Reviews: Cognitive Science 17 2 e70026,

2026
[43]

APACrefauthors \ 2017

harnad2017cognize APACrefauthors Harnad, S. APACrefauthors \ 2017 . To cognize is to categorize: cognition is categorization To cognize is to categorize: cognition is categorization . Handbook of categorization in cognitive science Handbook of categorization in cognitive science \ ( \ 21--54). Elsevier

2017
[44]

, J \"a rvikivi, J

hert2024importance APACrefauthors Hert, R. , J \"a rvikivi, J. Arnhold, A. APACrefauthors \ 2024 . The importance of linguistic factors: He likes subject referents The importance of linguistic factors: He likes subject referents . Cognitive Science 48 4 e13436,

2024
[45]

APACrefauthors \ 2024

hills2024behavioral APACrefauthors Hills, T.T. APACrefauthors \ 2024 . Behavioral Network Science: Language, Mind, and Society Behavioral network science: Language, mind, and society . Cambridge University Press

2024
[46]

\ Kenett, Y.N

hills2022mind APACrefauthors Hills, T.T. \ Kenett, Y.N. APACrefauthors \ 2022 . Is the mind a network? Maps, vehicles, and skyhooks in cognitive network science Is the mind a network? maps, vehicles, and skyhooks in cognitive network science . Topics in Cognitive Science 14 1 189--208,

2022
[47]

\ Rijkhoff, J

holmberg1998word APACrefauthors Holmberg, A. \ Rijkhoff, J. APACrefauthors \ 1998 . Word order in the Germanic languages Word order in the germanic languages . EMPIRICAL APPROACHES TO LANGUAGE TYPOLOGY 20 75--104,

1998
[48]

, Lam, M.H

huang2024apathetic APACrefauthors Huang, J t. , Lam, M.H. , Li, E.J. , Ren, S. , Wang, W. , Jiao, W. Lyu, M.R. APACrefauthors \ 2024 . Apathetic or empathetic? evaluating llms' emotional alignments with humans Apathetic or empathetic? evaluating llms' emotional alignments with humans . Advances in Neural Information Processing Systems 37 97053--97087,

2024
[49]

\ Gilbert, E

hutto2014vader APACrefauthors Hutto, C. \ Gilbert, E. APACrefauthors \ 2014 . Vader: A parsimonious rule-based model for sentiment analysis of social media text Vader: A parsimonious rule-based model for sentiment analysis of social media text . Proceedings of the international AAAI conference on web and social media Proceedings of the international aaai ...

2014
[50]

, Lindblom, K.M

jaeger2014trauma APACrefauthors Jaeger, J. , Lindblom, K.M. , Parker-Guilbert, K. Zoellner, L.A. APACrefauthors \ 2014 . Trauma narratives: It’s what you say, not how you say it. Trauma narratives: It’s what you say, not how you say it. Psychological Trauma: Theory, Research, Practice, and Policy 6 5 473,

2014
[51]

\ Hills, T.T

Kenett_2022_Networks APACrefauthors Kenett, Y.N. \ Hills, T.T. APACrefauthors \ 2022 . Editors' introduction to networks of the mind: How can network science elucidate our understanding of cognition? Editors' introduction to networks of the mind: How can network science elucidate our understanding of cognition? \ ( 14)\ ( 1). Wiley Online Library

2022
[52]

, Clutton, P

klein2018topic APACrefauthors Klein, C. , Clutton, P. Polito, V. APACrefauthors \ 2018 . Topic modeling reveals distinct interests within an online conspiracy forum Topic modeling reveals distinct interests within an online conspiracy forum . Frontiers in psychology 9 189,

2018
[53]

, McKillop, N

labhardt2025intervene APACrefauthors Labhardt, D. , McKillop, N. , Holdsworth, E. , Brown, S. , Howat, D. Jones, C. APACrefauthors \ 2025 . To Intervene or Not to Intervene: An Experimental Methodology Measuring Actual Bystander Behaviour To intervene or not to intervene: An experimental methodology measuring actual bystander behaviour . Behavioral Scienc...

2025
[54]

arXiv preprint arXiv:2310.10701 , year=

li2023theory APACrefauthors Li, H. , Chong, Y.Q. , Stepputtis, S. , Campbell, J. , Hughes, D. , Lewis, M. Sycara, K. APACrefauthors \ 2023 . Theory of mind for multi-agent collaboration via large language models Theory of mind for multi-agent collaboration via large language models . arXiv preprint arXiv:2310.10701 ,

work page arXiv 2023
[55]

, Zeng, C

li2024leveraging APACrefauthors Li, Y. , Zeng, C. , Zhong, J. , Zhang, R. , Zhang, M. Zou, L. APACrefauthors \ 2024 . Leveraging large language model as simulated patients for clinical education Leveraging large language model as simulated patients for clinical education . arXiv preprint arXiv:2404.13066 ,

work page arXiv 2024
[56]

, Conway, M

lozoya2025new APACrefauthors Lozoya, D. , Conway, M. , De Duro, E. D'Alfonso, S. APACrefauthors \ 2025 June 5 . Leveraging Large Language Models for Simulated Psychotherapy Client Interactions: Development and Usability Study of Client101 Leveraging Large Language Models for Simulated Psychotherapy Client Interactions: Development and Usability Study of C...

work page doi:10.2196/68056 2025
[57]

, Waheed, A

malhotra2022speaker APACrefauthors Malhotra, G. , Waheed, A. , Srivastava, A. , Akhtar, M.S. Chakraborty, T. APACrefauthors \ 2022 . Speaker and time-aware joint contextual learning for dialogue-act classification in counselling conversations Speaker and time-aware joint contextual learning for dialogue-act classification in counselling conversations . Pr...

2022
[58]

APACrefauthors \ 2011

mcadams2011narrative APACrefauthors McAdams, D.P. APACrefauthors \ 2011 . Narrative identity Narrative identity . Handbook of identity theory and research Handbook of identity theory and research \ ( \ 99--115). Springer

2011
[59]

\ Roth, C

menezes2019semantic APACrefauthors Menezes, T. \ Roth, C. APACrefauthors \ 2019 . Semantic hypergraphs Semantic hypergraphs . arXiv preprint arXiv:1908.10784 ,

work page arXiv 2019
[60]

, Hills, T

Miani2022LOCO APACrefauthors Miani, A. , Hills, T. Bangerter, A. APACrefauthors \ 2022 . LOCO: The 88-million-word language of conspiracy corpus Loco: The 88-million-word language of conspiracy corpus . Behavior research methods 54 4 1794--1817,

2022
[61]

Efficient Estimation of Word Representations in Vector Space

mikolov2013efficient APACrefauthors Mikolov, T. , Chen, K. , Corrado, G. Dean, J. APACrefauthors \ 2013 . Efficient estimation of word representations in vector space Efficient estimation of word representations in vector space . arXiv preprint arXiv:1301.3781 ,

work page internal anchor Pith review arXiv 2013
[62]

APACrefauthors \ 1995

miller1995wordnet APACrefauthors Miller, G.A. APACrefauthors \ 1995 . WordNet: a lexical database for English Wordnet: a lexical database for english . Communications of the ACM 38 11 39--41,

1995
[63]

APACrefauthors \ 2018

mohammad2018intensities APACrefauthors Mohammad, S.M. APACrefauthors \ 2018 . Word Affect Intensities Word affect intensities . Proceedings of the 11th Edition of the Language Resources and Evaluation Conference (LREC-2018). Proceedings of the 11th edition of the language resources and evaluation conference (lrec-2018). Miyazaki, Japan

2018
[64]

APACrefauthors \ 2018

newman_2018_networks APACrefauthors Newman, M. APACrefauthors \ 2018 . Networks: An Introduction Networks: An introduction . London (UK) Oxford University Press

2018
[65]

, Lysaker, P.H

nikzad2025evolution APACrefauthors Nikzad, A.H. , Lysaker, P.H. , Minor, K.S. , Leonhardt, B. , Liberman, M.Y. , Vohs, J. Tang, S.X. APACrefauthors \ 2025 . Evolution of Linguistic Markers of Agency, Centrality and Content During Metacognitive Therapy for Psychosis: A Pilot Exploratory Study Evolution of linguistic markers of agency, centrality and conten...

2025
[66]

APACrefauthors \ 2018

o2018today APACrefauthors O'Neill, T. APACrefauthors \ 2018 . ‘Today I speak’: Exploring how victim-survivors use Reddit ‘today i speak’: Exploring how victim-survivors use reddit . International journal for crime, justice and social democracy 7 1 44--59,

2018
[67]

, Cattan, A

Otmazgin2022FcorefFA APACrefauthors Otmazgin, S. , Cattan, A. Goldberg, Y. APACrefauthors \ 2022 . F-coref: Fast, Accurate and Easy to Use Coreference Resolution F-coref: Fast, accurate and easy to use coreference resolution . AACL. Aacl

2022
[68]

APACrefauthors \ 2023

pavlick2023symbols APACrefauthors Pavlick, E. APACrefauthors \ 2023 . Symbols and grounding in large language models Symbols and grounding in large language models . Philosophical Transactions of the Royal Society A 381 2251 20220041,

2023
[69]

\ Freund, R.R

peluso2018therapist APACrefauthors Peluso, P.R. \ Freund, R.R. APACrefauthors \ 2018 . Therapist and client emotional expression and psychotherapy outcomes: A meta-analysis. Therapist and client emotional expression and psychotherapy outcomes: A meta-analysis. Psychotherapy 55 4 461,

2018
[70]

, Francis, M.E

pennebaker2001linguistic APACrefauthors Pennebaker, J.W. , Francis, M.E. Booth, R.J. APACrefauthors \ 2001 . Linguistic inquiry and word count: LIWC 2001 Linguistic inquiry and word count: Liwc 2001 . Mahway: Lawrence Erlbaum Associates 71 2001 2001,

2001
[71]

, Zhang, Y

qi2020stanza APACrefauthors Qi, P. , Zhang, Y. , Zhang, Y. , Bolton, J. Manning, C.D. APACrefauthors \ 2020 . Stanza: A Python natural language processing toolkit for many human languages Stanza: A python natural language processing toolkit for many human languages . arXiv preprint arXiv:2003.07082 ,

work page arXiv 2020
[72]

, Cebrian, M

rahwan2019machine APACrefauthors Rahwan, I. , Cebrian, M. , Obradovich, N. , Bongard, J. , Bonnefon, J F. , Breazeal, C. others APACrefauthors \ 2019 . Machine behaviour Machine behaviour . Nature 568 7753 477--486,

2019
[73]

, Cola, G

recordare_2024_conspiracy_theorists APACrefauthors Recordare, A. , Cola, G. , Fagni, T. Tesconi, M. APACrefauthors \ 2024 . Unveiling Online Conspiracy Theorists: a Text-Based Approach and Characterization Unveiling online conspiracy theorists: a text-based approach and characterization . arXiv preprint arXiv:2405.12566 ,

work page arXiv 2024
[74]

, Norr, A.M

reger2021virtual APACrefauthors Reger, G.M. , Norr, A.M. , Gramlich, M.A. Buchman, J.M. APACrefauthors \ 2021 . Virtual standardized patients for mental health education Virtual standardized patients for mental health education . Current psychiatry reports 23 9 57,

2021
[75]

, Klösch, B

reiter_haas_2024_analysis_health_narratives APACrefauthors Reiter-Haas, M. , Klösch, B. , Hadler, M. Lex, E. APACrefauthors \ 2024 . Framing Analysis of Health-Related Narratives: Conspiracy versus Mainstream Media. Framing analysis of health-related narratives: Conspiracy versus mainstream media. APACrefURL https://arxiv.org/abs/2401.10030 APACrefURL

work page arXiv 2024
[76]

APACrefauthors \ 2012

rogers2012client APACrefauthors Rogers, C. APACrefauthors \ 2012 . Client centered therapy (new ed) Client centered therapy (new ed) . Hachette UK

2012
[77]

APACrefauthors \ 2019

rudin2019stop APACrefauthors Rudin, C. APACrefauthors \ 2019 . Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead . Nature machine intelligence 1 5 206--215,

2019
[78]

APACrefauthors \ 1986

sarbin1986narrative APACrefauthors Sarbin, T.R. APACrefauthors \ 1986 . Narrative psychology: The Storied Nature of Human Conduct Narrative psychology: The storied nature of human conduct . Bloomsbury Publishing

1986
[79]

, Vilella, S

Semeraro_2025_EmoAtlas APACrefauthors Semeraro, A. , Vilella, S. , Improta, R. , De Duro, E.S. , Mohammad, S.M. , Ruffo, G. Stella, M. APACrefauthors \ 2025 . EmoAtlas: An emotional network analyzer of texts that merges psychological lexicons, artificial intelligence, and network science Emoatlas: An emotional network analyzer of texts that merges psychol...

2025
[80]

, Vilella, S

Semeraro_2022_Emotional APACrefauthors Semeraro, A. , Vilella, S. , Ruffo, G. Stella, M. APACrefauthors \ 2022 . Emotional profiling and cognitive networks unravel how mainstream and alternative press framed AstraZeneca, Pfizer and COVID-19 vaccination campaigns Emotional profiling and cognitive networks unravel how mainstream and alternative press framed...

2022

Showing first 80 references.