text-mobility synergy

Text-mobility synergy is the assembly stage that combines written descriptions of a place with actual data on how people move to and from it, producing one numerical coordinate that a computer can work with.

Think of judging a person using both their resume and their real on-the-job behavior. A resume might just say 'cafe, downtown, 7am-9pm,' but it can't tell you whether that cafe is busy in the morning and quiet at night, or the reverse. Text-mobility synergy is the final step that compresses this written information together with actual visit records into a single card.

After this step, two places that share the same label, 'cafe,' end up with different coordinates if one is a morning spot and the other is an evening spot. Two places that looked identical from text alone finally become distinguishable the moment mobility data is blended in.

In the article, it appears as the third stage of Google Research's ME-POIs pipeline: "In the spatial multi-scale visit propagation and text-mobility synergy stage, this mobility signal is combined with the text-based representation into a single embedding." It's not the name of a separately sold AI model or product, but a term for the processing step that merges multiple types of data into one representation.