Against the backdrop of the “Healthy China” initiative and ongoing urban regeneration, the adaptive relationship between public space and residents’ health behaviors in aging urban neighborhoods has become a key focus of urban planning research. This study selects eight typical aging neighborhoods in downtown Jinan as empirical cases and constructs an “Emotion–Behavior–Space” analytical framework grounded in emotional geography. By applying SnowNLP sentiment analysis and LDA topic modeling to 340 semi-structured interview transcripts, and constructing an emotional keyword network using Gephi, the study accurately reveals the emotional interactions between residents’ outdoor activity patterns and spatial perceptions, and identifies key environmental factors influencing spatial adaptability. Findings show that middle-aged and elderly residents are the main activity participants, with diverse but low-intensity activity types. Emotional experience is significantly associated with activity frequency and spatial characteristics. Positive emotional perceptions are driven by diverse and decentralized factors such as ecological quality, place ambiance, and service accessibility, whereas negative emotions are concentrated around a limited number of high-centrality issues, including narrow roads, traffic conflicts, aging infrastructure, and spatial disorder. By introducing natural language processing and network analysis, this research overcomes the limitations of traditional qualitative approaches, unveils the interaction logic among resident behavior, emotion, and space, and proposes a dual-core renewal strategy focusing on “diversified positive experience space” and “precise intervention at negative nodes,” offering theoretical support and methodological guidance for fine-grained renewal of aging neighborhoods under the healthy city agenda.