A multi-agent framework for detecting and forecasting silent resistance in organizational communication: Implications for open innovation dynamics

A central yet understudied issue to the idea of change in organizations is its silent resistance that comprises it. It does not deal with active resistance, in contrast, like in case with overt resistance, it manifests in more subtle patterns of behavior leadership communication withdrawal, lack of engagement, or strategic silence. Such kinds of resistance can hardly be identified with a conventional tool such as a survey, an interview or sentiment analysis. The concept of this study involves a multi-agent communication mining model, a new model that aims at sensing and predicting the silent resistance by treating each actor in an organization as an agent. The model measures three important facets of communication including the interaction frequency that can be used to identify silence, semantic content to measure disengagement, and network centrality that can be used to measure the moving around in the course of communication structures. These points are grouped in an active resistance risk score that adjusts with time. The framework also integrates a forecasting module, which predicts the future risk of resistance following recent behavioral trends. The assessment is performed using a synthetic dataset that represents a scenario of organizational transformation. The findings demonstrate that the suggested model is substantially better than the sentiment-only and engagement-only benchmark models, with an F1-score of 0.862 and AUC of 0.968. The effect of each group of features and their combination can be verified with ablation and temporal analyses, and the forecasting experiment proves that resistance risk can be predicted up to five weeks prior. This research introduces a quantifiable, data-derived persistence mechanism of behavior tracking in change dynamics. It can give us valuable information of action that can be taken as organizational leaders will get some early signs of resistance being built, thus having a better chance of being able to make good decisions when it comes to using strategies to change. Their findings support the introduction of behavioral analytics into business intelligence.Future research could extend the proposed framework to encompass broader open innovation dynamics, including the management of collective intelligence, the adoption of behavioral analytics in Small and Medium-sized Enterprises (SMEs), and the development of platform-based ecosystems.