探勘社交媒體之軌跡模式於旅程規劃

Abstract

近年智慧型手持裝志科技的進步與社交媒體成為使用者分享資訊的平台，使得 地理資訊相關的服務也逐漸普及，像是許多行動社群網路 (Location-based Social Networks) 相關的服務如Facebook、Foursquare 等都提供使用者在社群上分享個人所在 位置。許多人會在旅遊或日常生活時將自己所在的地理位置或是照片分享於社群之上 （如分享照片和打卡 (check-in)），增加與社群上成員的互動。目前已經有許多網站提供 使用者上傳個人的軌跡供其他人參考並增加更多互動，使用者得以分享其旅遊的軌跡資 訊供其他使用者參考。在本計劃中，我們預計提出一系列的資料探勘演算法於社交媒體 資料，找尋使用者的軌跡模式。透過這個軌跡模式，我們擬提出旅程規劃的平台，在此 平台上透過所探勘的軌跡模式，用以協助使用者更為方便地規劃旅程。 本計畫是為期三年之計畫 『探勘社交媒體之軌跡模式於旅程規劃』。在此計畫中， 將專注於從社交媒體中探勘使用者的移動模式，使用於旅程規劃推薦服務。其中社群媒 體探勘包含了含有地理資訊（經緯度座標）和拜訪該地點的時間的資料，如軌跡資訊、 打卡資訊、相片等資料中，探勘使用者移動模式，並根據探勘出的使用者移動模式和使 用者所輸入的旅遊路徑偏好，計算並推薦合適的旅遊路徑。其中使用者移動模式將考量 三個部分，順序因素 (Ordering factor)、時間因素 (Time factor) 和社群因素 (Social factor)。具體而言，在第一年，我們將透過社交媒體所包含之 GPS 軌跡資料、照片資料 以及打卡資料，探勘熱門的景點資訊。透過這些景點資訊與軌跡資料，提出以軌跡模式 為基礎之旅程規劃搜尋平台。在第二年，我們進一步透過社交媒體資料，探勘每個景點 地拜訪時間分布模式。在旅程規劃搜尋平台中，研發(1). 具有適合拜訪時間之旅程規劃 演算法 (2). 具有使用者查詢景點 (query points) 與適合拜訪時間之旅程規劃演算法。 有鑑於社交媒體上，我們將可探勘使用者社交活動行為模式，在第三年中，我們擬提出 具有社交行為模式的旅程規劃演算法。 隨著社交媒體與地理資訊服務近年的蓬勃發展，我們相信此計畫之執行，將可研 發出適用於軌跡模式為基礎之旅程規劃規劃搜尋平台，提供使用者更為便利的旅程規劃 服務。

The increasing availability of location-acquisition technology (e.g., GPS), has led to a huge volume of spatial trajectories that represent the movement routes of humans. Without loss of generality, a trajectory is a sequence of data points where each data point records location information and a time-stamp. For example, users could perform check-in services (e.g., Foursquare) to note their locations via a mobile phone and share their photos and activities. The time-ordered check-in records of a user are able to be expressed by trajectories. Moreover, on a photo sharing website (e.g., Flickr), people share geotagged photos whose time-stamps and geolocations can be represented as trajectories as well. In our project, we intend to mine trajectory patterns from social media and develop pattern-aware trajectory search platform for trip planning. This project is a three-year project 『Mining trajectory patterns from social media for trip planning』, and aims at designing a trip planning platform that consists of mining social media and trip planning algorithms. Our primary goals include (1) developing Pattern-Aware Trajectory Search (abbreviated as PATS); (2) proposing pattern-aware trajectory search with time constraint and user query points; (3) exploring social relationship for trip planning. More specifically, in the first year, given a spatial range and a user preference of depth/breadth specified by a user, we develop a Pattern-Aware Trajectory Search (PATS) framework to retrieve the top K trajectories passing through popular Regions Of Interests (abbreviated as ROIs). PATS is novel because the returned travel trajectories, discovered from travel patterns hidden in trip trajectories, may represent the most valuable travel experiences of other travelers fitting the user’s trip preference in terms of depth or breadth. In the second year, we aim at mining visiting time distribution of ROIs. Based on the mining results, we propose PATS with time constraint and PATS with both time constraint and user query points. Due to the social relationships among users, in the third year, we propose a social-aware trip planning framework. In view of the increasing attention on social media and location-based services, we strongly believe that this project is very timely and will deliver results of both theoretical and practical importance.