Towards Preventing Intimate Partner Violence by Detecting Disagreements in SMS Communications

Mahesh Babu Kommalapati, Xiao Gu, Harshit Pandey, Christie J. Rizzo, Charlene Collibee, Silvio Amir, Aarti Sathyanarayana
Proceedings of the 4th Machine Learning for Health Symposium, PMLR 259:610-622, 2025.

Abstract

Intimate Partner Violence (IPV) among adolescents is a major public health concern, particularly for justice-involved adolescents which are at higher risk of experiencing IPV. Early detection of disagreements between romantic partners can provide an opportunity for just-in-time interventions to prevent escalation. Prior work has proposed methods for early detection of disagreements based on metadata features of text message conversations. In this work, we build on these prior efforts and investigate the impact of explicitly modeling the contents of text conversations for disagreement detection. We develop and evaluate supervised classifiers that combine metadata features with sentiment and semantic features of texts and compare their performance against few-shot learning with instruction-tuned Large Language Models (LLMs). We conduct experiments on a dataset collected to study the communication patterns and risk factors associated with IPV among justice-involved adolescents. In addition, we measure models’ generalization to out-of-distribution samples using an external dataset comprising adolescents enrolled in child welfare services. We find that: (i) text-based features improve predictive performance but do not help models generalize to other populations; and (ii) LLMs struggle in this setting but can outperform supervised classifiers in out-of-distribution samples.

Cite this Paper


BibTeX
@InProceedings{pmlr-v259-kommalapati25a, title = {Towards Preventing Intimate Partner Violence by Detecting Disagreements in SMS Communications}, author = {Kommalapati, Mahesh Babu and Gu, Xiao and Pandey, Harshit and Rizzo, Christie J. and Collibee, Charlene and Amir, Silvio and Sathyanarayana, Aarti}, booktitle = {Proceedings of the 4th Machine Learning for Health Symposium}, pages = {610--622}, year = {2025}, editor = {Hegselmann, Stefan and Zhou, Helen and Healey, Elizabeth and Chang, Trenton and Ellington, Caleb and Mhasawade, Vishwali and Tonekaboni, Sana and Argaw, Peniel and Zhang, Haoran}, volume = {259}, series = {Proceedings of Machine Learning Research}, month = {15--16 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v259/main/assets/kommalapati25a/kommalapati25a.pdf}, url = {https://proceedings.mlr.press/v259/kommalapati25a.html}, abstract = {Intimate Partner Violence (IPV) among adolescents is a major public health concern, particularly for justice-involved adolescents which are at higher risk of experiencing IPV. Early detection of disagreements between romantic partners can provide an opportunity for just-in-time interventions to prevent escalation. Prior work has proposed methods for early detection of disagreements based on metadata features of text message conversations. In this work, we build on these prior efforts and investigate the impact of explicitly modeling the contents of text conversations for disagreement detection. We develop and evaluate supervised classifiers that combine metadata features with sentiment and semantic features of texts and compare their performance against few-shot learning with instruction-tuned Large Language Models (LLMs). We conduct experiments on a dataset collected to study the communication patterns and risk factors associated with IPV among justice-involved adolescents. In addition, we measure models’ generalization to out-of-distribution samples using an external dataset comprising adolescents enrolled in child welfare services. We find that: (i) text-based features improve predictive performance but do not help models generalize to other populations; and (ii) LLMs struggle in this setting but can outperform supervised classifiers in out-of-distribution samples.} }
Endnote
%0 Conference Paper %T Towards Preventing Intimate Partner Violence by Detecting Disagreements in SMS Communications %A Mahesh Babu Kommalapati %A Xiao Gu %A Harshit Pandey %A Christie J. Rizzo %A Charlene Collibee %A Silvio Amir %A Aarti Sathyanarayana %B Proceedings of the 4th Machine Learning for Health Symposium %C Proceedings of Machine Learning Research %D 2025 %E Stefan Hegselmann %E Helen Zhou %E Elizabeth Healey %E Trenton Chang %E Caleb Ellington %E Vishwali Mhasawade %E Sana Tonekaboni %E Peniel Argaw %E Haoran Zhang %F pmlr-v259-kommalapati25a %I PMLR %P 610--622 %U https://proceedings.mlr.press/v259/kommalapati25a.html %V 259 %X Intimate Partner Violence (IPV) among adolescents is a major public health concern, particularly for justice-involved adolescents which are at higher risk of experiencing IPV. Early detection of disagreements between romantic partners can provide an opportunity for just-in-time interventions to prevent escalation. Prior work has proposed methods for early detection of disagreements based on metadata features of text message conversations. In this work, we build on these prior efforts and investigate the impact of explicitly modeling the contents of text conversations for disagreement detection. We develop and evaluate supervised classifiers that combine metadata features with sentiment and semantic features of texts and compare their performance against few-shot learning with instruction-tuned Large Language Models (LLMs). We conduct experiments on a dataset collected to study the communication patterns and risk factors associated with IPV among justice-involved adolescents. In addition, we measure models’ generalization to out-of-distribution samples using an external dataset comprising adolescents enrolled in child welfare services. We find that: (i) text-based features improve predictive performance but do not help models generalize to other populations; and (ii) LLMs struggle in this setting but can outperform supervised classifiers in out-of-distribution samples.
APA
Kommalapati, M.B., Gu, X., Pandey, H., Rizzo, C.J., Collibee, C., Amir, S. & Sathyanarayana, A.. (2025). Towards Preventing Intimate Partner Violence by Detecting Disagreements in SMS Communications. Proceedings of the 4th Machine Learning for Health Symposium, in Proceedings of Machine Learning Research 259:610-622 Available from https://proceedings.mlr.press/v259/kommalapati25a.html.

Related Material