Indoor fingerprint localization technology has emerged as a promising solution for indoor location services, which have become increasingly important in various applications. However, the accuracy and coverage of the indoor fingerprint database are critical factors that affect the performance of this technology. In this paper, we propose a novel method to verify and match the home broadband address based on a finite state machine and web crawler technology. We first construct a basic address information database by crawling online data sources. Then we use analytic hierarchy process (AHP) to assign word position weights and extract effective feature words from the home broadband address. Finally, we employ cosine distance to find the most similar basic address data to the home broadband address. We conduct experiments to evaluate our method and show that it can improve the feature word extraction accuracy and similarity calculation accuracy by more than 30%. As a result, our method can enhance the quality of big data positioning and achieve a significant improvement in the quality and data volume of the indoor fingerprint database.