Balancing Goals, Health, and Cost: A Food Information System for Managing Complex Choices and Fostering Sustained Food Agency

Paper Title

Balancing Goals, Health, and Cost: A Food Information System for Managing Complex Choices and Fostering Sustained Food Agency

Publication Info

• Topic area: Technology-supported grocery planning for low-income communities
• Keywords: Food agency, grocery planning, multi-objective optimization, nutrition awareness, self-regulated learning, low-income communities, dietary goals, food literacy, healthy eating index, personalized recommendations

Background and Problem

• Problem / challenge: Individuals in low-income communities face systemic barriers to obtaining nutritious and affordable groceries. Existing grocery technologies often fail to address the complexity of balancing cost, nutrition, and personal dietary goals, and they lack integration across the planning, shopping, and reflection phases.
• Significance: Addressing these challenges can help mitigate health disparities, improve dietary quality, and empower individuals to make informed food choices.
• Motivation and related work: Prior research has explored food agency, self-tracking, and optimization for food procurement, but these efforts are often fragmented, focusing on single phases of the grocery cycle. Existing systems lack holistic approaches that integrate real-world constraints like cost and inventory with user-specific dietary goals.

Solution

• Proposed approach: The Food Information System, a grocery planning tool that uses multi-objective optimization to recommend products aligned with users’ dietary goals, store inventory, and cost constraints.
• Novelty:
  1. Introduction of a conceptual model treating grocery planning as a multi-objective optimization problem.
  2. Integration of Zimmerman’s Self-Regulated Learning framework to support sustained behavior change across planning, shopping, and reflection phases.
  3. Implementation of personalized feedback reports using the Healthy Eating Index (HEI) and natural language explanations.
  4. Focus on low-income communities with real-world constraints, including cost and inventory.
• Procedure and key techniques:
  • Users set dietary goals and build grocery lists.
  • The system applies a multi-objective optimization algorithm to recommend products, balancing cost, nutrition, and user goals.
  • Recommendations include alternatives and explanations to support informed decision-making.
  • Users receive biweekly feedback reports summarizing their grocery purchases and offering actionable suggestions for improvement.

Results

• Concrete findings:
  • Average HEI scores showed slight, non-significant improvement from 51.91 to 53.13 during the intervention.
  • Participants reported increased comfort with mobile grocery tools, with significant improvements in planning (+15.85), product selection (+17.45), information comparison (+19.30), and budgeting (+16.70) on a 0–100 scale.
  • The purchase rate for recommended items was 11.29%, with higher uptake for vegetables (23.85%) and dairy (22.31%).
• Advantage over baselines:
  • Participants reported improved nutritional awareness, food literacy, and intentional grocery choices compared to pre-intervention behaviors.
  • The system supported increased perceived food agency, even though overall HEI improvements were not statistically significant.
• Experiments / evaluation:
  • Eight-week within-subjects intervention with 55 participants from a food-insecure community.
  • Data collected through grocery receipts, app usage logs, and focus groups.
  • HEI scores used to evaluate dietary quality; qualitative feedback analyzed thematically.
• Limitations and future work:
  • Limited to a single store (Walmart) and short intervention period (8 weeks).
  • Challenges with real-time inventory accuracy and incomplete nutritional data.
  • Future work should explore multi-store integration, longer-term deployments, and more robust evaluation of recommendation quality.

Summary

The Food Information System introduces a novel approach to grocery planning, integrating multi-objective optimization with Zimmerman’s Self-Regulated Learning framework to support healthier food choices in low-income communities. While HEI improvements were not statistically significant, participants reported increased nutritional awareness, food literacy, and comfort with mobile grocery tools. The study highlights the importance of transparency, personalization, and user control in fostering food agency. Future research should address data limitations, expand to multi-store contexts, and explore long-term impacts on dietary behavior.