VISIT SPOT DETECTION DEVICE, RELEVANCE ANALYSIS DEVICE, VISIT SPOT DETECTION METHOD AND VISIT SPOT DETECTION PROCESSING PROGRAM

PROBLEM TO BE SOLVED: To accurately process information of a spot that is visited by a visitor.SOLUTION: A visit spot detection device in the embodiment is characterized by: detecting a visit spot corresponding to the beacon closest to a visitor for each of predetermined time, based on a threshold o...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: OTA MASAKATSU
Format: Patent
Sprache:eng ; jpn
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:PROBLEM TO BE SOLVED: To accurately process information of a spot that is visited by a visitor.SOLUTION: A visit spot detection device in the embodiment is characterized by: detecting a visit spot corresponding to the beacon closest to a visitor for each of predetermined time, based on a threshold of a received electric wave strength concerning at least one beacon and a detection parameter including the threshold of the radio wave receiving frequency; creating a character string including the character string obtained by arranging each detected spot in the visit order about the spot detected in the pattern of the parameter; finding perplexity serving as an index of prediction accuracy from a character just before that to the next character in the created character string, based on evaluation results according to a statistical language model to the continuity of the spot detected in this character string; and outputting the information of the visit spot detected by using the parameter used when the number of the detected spot type is the predetermined number or more in each of the pattern and the value of the perplexity is the smallest.SELECTED DRAWING: Figure 1 【課題】訪問者が訪れるスポットの情報を精度よく処理する。【解決手段】実施形態における訪問スポット検出装置は、少なくとも1つのビーコンに関する受信電波強度の閾値と電波受信頻度の閾値を含む検出パラメータに基づいて、来場者に最も近いビーコンに対応する訪問スポットを所定の時間ごとに検出し、パラメータのパターンにおいて検出したスポットについて、検出した各スポットを訪問順に並べた文字列を含む文字列を作成し、この文字列における検出したスポットの連続性に対する統計的言語モデルによる評価結果に基づいて、作成した文字列における直前の文字から次の文字への予測精度の指標となるパープレキシティを求め、パターンのそれぞれにおける、検出されたスポットの種類数が所定数以上でパープレキシティの値が最小の場合に用いたパラメータを用いて検出された訪問スポットの情報を出力する。【選択図】図1