pith. sign in

arxiv: 1907.10036 · v1 · pith:ZOYG5BLXnew · submitted 2019-07-23 · 💻 cs.CL · cs.HC

Happiness Entailment: Automating Suggestions for Well-Being

Sara Evensen , Yoshihiko Suhara , Alon Halevy , Vivian Li , Wang-Chiew Tan , Saran Mumick This is my paper

Pith reviewed 2026-05-24 17:20 UTC · model grok-4.3

classification 💻 cs.CL cs.HC
keywords happiness entailmentwell-being suggestionsneural network modeltext analysispsychological featuressustainable actionsevent descriptionentailment recognition
0
0 comments X

The pith

A neural network determines whether a candidate suggestion sustains well-being from a short description of a happy event.

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

The paper outlines a system that reads short texts users write about their happy moments and proposes actions meant to produce lasting improvements in well-being. It focuses on one component, the Happiness Entailment Recognition module, which receives both the event text and a possible suggestion and decides whether the suggestion is likely to help this particular user. The module is built as a neural network with separate encoders for the text and the suggestion plus extra layers that target psychologically relevant signals. A sympathetic reader would care because the approach offers an automated alternative to expert review of periodic self-reports, potentially turning everyday journaling into ongoing, personalized guidance.

Core claim

The Happiness Entailment Recognition (HER) module takes a short text describing an event and a candidate actionable suggestion as input and outputs a determination about whether the suggestion is more likely to be good for the user based on the event described. The module is implemented as a neural network with two encoders, one for the user input and one for the candidate suggestion, along with additional layers to capture psychologically significant features in the happy moment and suggestion.

What carries the argument

The Happiness Entailment Recognition (HER) module, a neural network with two encoders plus layers for psychologically significant features that decides if a suggestion fits a described happy event.

If this is right

  • A full system can analyze user journals and generate tailored suggestions without requiring expert analysis of assessments.
  • The module can serve as a reusable component inside larger applications aimed at behavior change for well-being.
  • Suggestions are chosen to encourage sustainable rather than short-term changes in user behavior.
  • The method complements existing self-reporting tools by shifting from periodic expert review to continuous automated analysis.

Where Pith is reading between the lines

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

  • Mobile apps that prompt daily journaling could feed the module in real time and surface suggestions immediately after an entry.
  • The same two-encoder structure might be adapted to other psychological goals such as reducing reported stress or increasing reported gratitude.
  • Longitudinal user studies could measure whether repeated use of the suggestions produces cumulative well-being gains beyond what random suggestions achieve.

Load-bearing premise

Short texts describing happy moments contain enough information for a neural model to judge which suggestions will produce sustainable improvements in well-being.

What would settle it

An experiment that tracks users' actual well-being scores after they follow or ignore the model's suggestions and checks whether the model's positive determinations reliably predict measured gains.

Figures

Figures reproduced from arXiv: 1907.10036 by Alon Halevy, Sara Evensen, Saran Mumick, Vivian Li, Wang-Chiew Tan, Yoshihiko Suhara.

Figure 1
Figure 1. Figure 1: Overview of our system. The system outputs a best followup suggestion based on an input happy moment provided by the user. The suggestion [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Manually crafted 36 sustainable suggestions. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: Network architecture of the model. The encoder (in gray boxes) [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Randomly chosen examples from our dataset, shown with both the selected gold labels (in bold) and the full set of annotations from the individual [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Examples of happy moments and their suggestibility labels. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
read the original abstract

Understanding what makes people happy is a central topic in psychology. Prior work has mostly focused on developing self-reporting assessment tools for individuals and relies on experts to analyze the periodic reported assessments. One of the goals of the analysis is to understand what actions are necessary to encourage modifications in the behaviors of the individuals to improve their overall well-being. In this paper, we outline a complementary approach; on the assumption that the user journals her happy moments as short texts, a system can analyze these texts and propose sustainable suggestions for the user that may lead to an overall improvement in her well-being. We prototype one necessary component of such a system, the Happiness Entailment Recognition (HER) module, which takes as input a short text describing an event, a candidate suggestion, and outputs a determination about whether the suggestion is more likely to be good for this user based on the event described. This component is implemented as a neural network model with two encoders, one for the user input and one for the candidate actionable suggestion, with additional layers to capture psychologically significant features in the happy moment and suggestion.

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.

Referee Report

2 major / 0 minor

Summary. The paper outlines a complementary approach to traditional psychology-based well-being assessment tools. On the assumption that users journal happy moments as short texts, it proposes analyzing these texts to generate sustainable suggestions for behavior modification. The central contribution is a prototype of the Happiness Entailment Recognition (HER) module: a neural network with two encoders (one for the event text, one for the candidate suggestion) plus additional layers to capture psychologically significant features; the module outputs a binary determination of whether the suggestion is likely good for the user given the described event.

Significance. If the HER module can be shown to extract relevant signals, the work could enable scalable, automated systems for personalized well-being suggestions that complement expert-driven self-report analysis. The high-level architecture is clearly described and directly addresses a practical gap, but the manuscript supplies no training data, loss function, or performance results, so the significance remains prospective rather than demonstrated.

major comments (2)
  1. [HER module description] HER module description (abstract and implementation outline): The claim that the two-encoder architecture with psychologically significant feature layers can determine whether a suggestion is good for the user is load-bearing for the prototype's utility, yet the manuscript provides no training corpus of (event, suggestion, label) triples, no label provenance, no loss function, and no accuracy/F1 or other metrics. Without these, the assertion that the model captures the relevant signals remains an untested design statement.
  2. [Introduction and approach] Introduction and approach section: The assumption that short texts describing happy moments contain sufficient information for reliable determination of sustainable well-being improvements is stated without supporting analysis, preliminary experiments, or discussion of potential information bottlenecks in the input texts.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed review and constructive comments. Our manuscript presents a high-level proposal and architectural outline for the HER module as a conceptual prototype, with the understanding that it is prospective rather than a completed empirical study. We respond to each major comment below.

read point-by-point responses

  1. Referee: [HER module description] HER module description (abstract and implementation outline): The claim that the two-encoder architecture with psychologically significant feature layers can determine whether a suggestion is good for the user is load-bearing for the prototype's utility, yet the manuscript provides no training corpus of (event, suggestion, label) triples, no label provenance, no loss function, and no accuracy/F1 or other metrics. Without these, the assertion that the model captures the relevant signals remains an untested design statement.

    Authors: The manuscript describes the proposed two-encoder architecture with psychologically motivated layers as a design for the HER module, without claiming that an implemented version has been trained or evaluated. The contribution is the high-level architecture itself, positioned as a necessary component for future systems. We do not assert that the model has captured signals in practice; the utility is framed as prospective, consistent with the referee's own summary. To improve clarity we can revise the abstract and introduction to more explicitly state that this is an untested architectural proposal. revision: partial

  2. Referee: [Introduction and approach] Introduction and approach section: The assumption that short texts describing happy moments contain sufficient information for reliable determination of sustainable well-being improvements is stated without supporting analysis, preliminary experiments, or discussion of potential information bottlenecks in the input texts.

    Authors: The assumption is presented as the foundational premise for the complementary approach. While the paper draws on established psychology literature regarding journaling and behavior modification, we acknowledge that the manuscript does not provide supporting analysis or discuss bottlenecks. We can revise the introduction to include relevant citations and a brief discussion of potential limitations arising from short input texts. revision: yes

Circularity Check

0 steps flagged

No derivation chain present; architectural proposal only

full rationale

The paper outlines a high-level system design and describes the HER module as a two-encoder neural network with psych-feature layers. No equations, parameters, derivations, predictions, or self-citations appear in the provided text. The central claim is a design statement rather than a result derived from inputs, so none of the enumerated circularity patterns apply. The absence of empirical validation is a separate concern about completeness, not circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The proposal rests on domain assumptions about user journaling behavior and model capability rather than on fitted parameters or new invented entities.

axioms (2)
  • domain assumption Users journal their happy moments as short texts that contain actionable psychological information
    Explicitly stated as the operating assumption for the system in the abstract.
  • ad hoc to paper A two-encoder neural network with additional layers can capture psychologically significant features relevant to well-being suggestions
    The implementation choice described in the abstract assumes this capability without further justification.

pith-pipeline@v0.9.0 · 5734 in / 1153 out tokens · 19917 ms · 2026-05-24T17:20:15.073495+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

29 extracted references · 29 canonical work pages · 1 internal anchor

  1. [1]

    Positive psychology in clinical practice,

    A. L. Duckworth, T. A. Steen, and M. E. Seligman, “Positive psychology in clinical practice,” Annu. Rev. Clin. Psychol., vol. 1, pp. 629–651, 2005

    work page 2005

  2. [2]

    M. E. Seligman, Flourish: A visionary new understand- ing of happiness and well-being . Simon and Schuster, 2012

    work page 2012

  3. [3]

    What good are positive emotions?

    B. L. Fredrickson, “What good are positive emotions?” Review of general psychology, vol. 2, no. 3, pp. 300–319, 1998

    work page 1998

  4. [4]

    The benefits of frequent positive affect: Does happiness lead to success?

    S. Lyubomirsky, L. King, and E. Diener, “The benefits of frequent positive affect: Does happiness lead to success?” Psychological bulletin, vol. 131, no. 6, p. 803, 2005

    work page 2005

  5. [5]

    Long- running german panel survey shows that personal and economic choices, not just genes, matter for happiness,

    B. Headey, R. Muffels, and G. G. Wagner, “Long- running german panel survey shows that personal and economic choices, not just genes, matter for happiness,” Proceedings of the National Academy of Sciences , 2010

    work page 2010

  6. [6]

    How do simple positive activities increase well-being?

    S. Lyubomirsky and K. Layous, “How do simple positive activities increase well-being?” Current directions in psychological science, vol. 22, no. 1, pp. 57–62, 2013

    work page 2013

  7. [7]

    A survey method for char- acterizing daily life experience: The day reconstruction method,

    D. Kahneman, A. B. Krueger, D. A. Schkade, N. Schwarz, and A. A. Stone, “A survey method for char- acterizing daily life experience: The day reconstruction method,” Science, vol. 306, no. 5702, pp. 1776–1780, 2004

    work page 2004

  8. [8]

    The experience sampling method,

    R. Larson and M. Csikszentmihalyi, “The experience sampling method,” in Flow and the foundations of posi- tive psychology. Springer, 2014, pp. 21–34

    work page 2014

  9. [9]

    Ecolog- ical momentary assessment,

    S. Shiffman, A. A. Stone, and M. R. Hufford, “Ecolog- ical momentary assessment,” Annu. Rev. Clin. Psychol. , vol. 4, pp. 1–32, 2008

    work page 2008

  10. [10]

    Subjective well-being: The science of hap- piness and a proposal for a national index

    E. Diener, “Subjective well-being: The science of hap- piness and a proposal for a national index.” American psychologist, vol. 55, no. 1, p. 34, 2000

    work page 2000

  11. [11]

    Jo: The Smart Journal

    V . Li, A. Halevy, A. Zief-Balteriski, W.-C. Tan, G. Mi- haila, J. Morales, N. Nuno, H. Liu, C. Chen, X. Ma, S. Robins, and J. Johnson, “Jo: The smart journal,” arXiv:1907.07861, 2019

    work page internal anchor Pith review Pith/arXiv arXiv 1907

  12. [12]

    HappyDB: A Corpus of 100,000 Crowdsourced Happy Moments,

    A. Asai, S. Evensen, B. Golshan, A. Halevy, V . Li, A. Lopatenko, D. Stepanov, Y . Suhara, W.-C. Tan, and Y . Xu, “HappyDB: A Corpus of 100,000 Crowdsourced Happy Moments,” in Proc. LREC ’18 , 2018

    work page 2018

  13. [13]

    The CL-Aff happiness shared task: Results and key insights,

    K. Jaidka, S. Mumick, N. Chhaya, and L. Ungar, “The CL-Aff happiness shared task: Results and key insights,” in Proc. AffCon ’19 , 2019

    work page 2019

  14. [14]

    Lyubomirsky, The how of happiness: A scientific approach to getting the life you want

    S. Lyubomirsky, The how of happiness: A scientific approach to getting the life you want . Penguin, 2008

    work page 2008

  15. [15]

    Mental health manpower trends: A report to the staff director, jack r. ewalt

    G. W. Albee, “Mental health manpower trends: A report to the staff director, jack r. ewalt.” 1959

    work page 1959

  16. [16]

    Confronting a trau- matic event: toward an understanding of inhibition and disease

    J. W. Pennebaker and S. K. Beall, “Confronting a trau- matic event: toward an understanding of inhibition and disease.” Journal of abnormal psychology, vol. 95, no. 3, p. 274, 1986

    work page 1986

  17. [17]

    The pascal recognising textual entailment challenge,

    I. Dagan, O. Glickman, and B. Magnini, “The pascal recognising textual entailment challenge,” in Machine Learning Challenges Workshop, 2005, pp. 177–190

    work page 2005

  18. [18]

    Cheap and fast—but is it good?: evaluating non-expert annotations for natural language tasks,

    R. Snow, B. O’Connor, D. Jurafsky, and A. Y . Ng, “Cheap and fast—but is it good?: evaluating non-expert annotations for natural language tasks,” in Proc. EMNLP ’08. Association for Computational Linguistics, 2008, pp. 254–263

    work page 2008

  19. [19]

    Supervised learning of universal sentence representations from natural language inference data,

    A. Conneau, D. Kiela, H. Schwenk, L. Barrault, and A. Bordes, “Supervised learning of universal sentence representations from natural language inference data,” in Proc. EMNLP ’17 , 2017, pp. 670–680

    work page 2017

  20. [20]

    M. A. Gluck, C. E. Myers, and E. Mercado, Learning and Memory: From Brain To Behavior (3rd edition) . Worth Publishers, 2016

    work page 2016

  21. [21]

    The need to belong: desire for interpersonal attachments as a fundamental hu- man motivation

    R. F. Baumeister and M. R. Leary, “The need to belong: desire for interpersonal attachments as a fundamental hu- man motivation.” Psychological bulletin, vol. 117, no. 3, p. 497, 1995

    work page 1995

  22. [22]

    12 life task participation and well-being: The importance of taking part in daily life,

    N. Cantor and C. A. Sanderson, “12 life task participation and well-being: The importance of taking part in daily life,” Well-being: Foundations of hedonic psychology , p. 230, 2003

    work page 2003

  23. [23]

    A decomposable attention model for natural language inference,

    A. P. Parikh, O. T ¨ackstr¨om, D. Das, and J. Uszkoreit, “A decomposable attention model for natural language inference,” in Proc. EMNLP ’16 , 2016, pp. 2249–2255

    work page 2016

  24. [24]

    A large annotated corpus for learning natural language inference,

    S. R. Bowman, G. Angeli, C. Potts, and C. D. Manning, “A large annotated corpus for learning natural language inference,” in Proc. EMNLP ’15 , 2015

    work page 2015

  25. [25]

    Random search for hyper- parameter optimization,

    J. Bergstra and Y . Bengio, “Random search for hyper- parameter optimization,” Journal of Machine Learning Research, vol. 13, no. Feb, pp. 281–305, 2012

    work page 2012

  26. [26]

    A minimal span- based neural constituency parser,

    M. Stern, J. Andreas, and D. Klein, “A minimal span- based neural constituency parser,” in Proc. ACL ’17 , 2017, pp. 818–827

    work page 2017

  27. [27]

    A structured self-attentive sentence embedding,

    Z. Lin, M. Feng, C. N. d. Santos, M. Yu, B. Xiang, B. Zhou, and Y . Bengio, “A structured self-attentive sentence embedding,” in Proc. ICLR ’17 , 2017

    work page 2017

  28. [28]

    Happy To- gether: Learning and understanding appraisal from natu- ral language,

    M. A.-M. Arun Rajendran, Chiyu Zhang, “Happy To- gether: Learning and understanding appraisal from natu- ral language,” in Proc. AffCon ’19 , 2019

    work page 2019

  29. [29]

    What constitutes happiness? predicting and characterizing the ingredients of happiness using emotion intensity analy- sis,

    R. K. Gupta, P. Bhattacharya, and Y . Yang, “What constitutes happiness? predicting and characterizing the ingredients of happiness using emotion intensity analy- sis,” in Proc. AffCon ’19 , 2019

    work page 2019