Defu Lian

We know how you live: exploring the spectrum of urban lifestyles

An incisive understanding of human lifestyles is not only essential to many scientific disciplines, but also has a profound business impact for targeted marketing. In this paper, we present LifeSpec, a computational framework for exploring and hierarchically categorizing urban lifestyles. Specifically, we have developed an algorithm to connect multiple social network accounts of millions of individuals and collect their publicly available heterogeneous behavioral data as well as social links. In addition, a nonparametric Bayesian approach is developed to model the lifestyle spectrum of a group of individuals. To demonstrate the effectiveness of LifeSpec, we conducted extensive experiments and case studies, with a large dataset we collected covering 1 million individuals from 493 cities. Our results suggest that LifeSpec offers a powerful paradigm for 1) revealing an individuals lifestyle from multiple dimensions, and 2) uncovering lifestyle commonalities and variations of a group with various demographic attributes, such as vocation, education, gender, sexual orientation, and place of residence. The proposed method provides emerging implications for personalized recommendation and targeted advertising.

CEPR: A Collaborative Exploration and Periodically Returning Model for Location Prediction

With the growing popularity of location-based social networks, numerous location visiting records (e.g., check-ins) continue to accumulate over time. The more these records are collected, the better we can understand users’ mobility patterns and the more accurately we can predict their future locations. However, due to the personality trait of neophilia, people also show propensities of novelty seeking in human mobility, that is, exploring unvisited but tailored locations for them to visit. As such, the existing prediction algorithms, mainly relying on regular mobility patterns, face severe challenges because such behavior is beyond the reach of regularity. As a matter of fact, the prediction of this behavior not only relies on the forecast of novelty-seeking tendency but also depends on how to determine unvisited candidate locations. To this end, we put forward a Collaborative Exploration and Periodically Returning model (CEPR), based on a novel problem, Exploration Prediction (EP), which forecasts whether people will seek unvisited locations to visit, in the following. When people are predicted to do exploration, a state-of-the-art recommendation algorithm, armed with collaborative social knowledge and assisted by geographical influence, will be applied for seeking the suitable candidates; otherwise, a traditional prediction algorithm, incorporating both regularity and the Markov model, will be put into use for figuring out the most possible locations to visit. We then perform case studies on check-ins and evaluate them on two large-scale check-in datasets with 6M and 36M records, respectively. The evaluation results show that EP achieves a roughly 20p classification error rate on both datasets, greatly outperforming the baselines, and that CEPR improves performances by as much as 30p compared to the traditional location prediction algorithms.

Mining novelty-seeking trait across heterogeneous domains

An incisive understanding of personal psychological traits is not only essential to many scientific disciplines, but also has a profound business impact on online recommendation. Recent studies in psychology suggest that novelty-seeking trait is highly related to consumer behavior. In this paper, we focus on understanding individual novelty-seeking trait embodied at different levels and across heterogeneous domains. Unlike the questionnaire-based methods widely adopted in the past, we first present a computational framework, Novel Seeking Model (NSM), for exploring the novelty-seeking trait implied by observable activities. Then, we explore the novelty-seeking trait in two heterogeneous domains: check-in behavior in location based social networks, which reflects mobility patterns in the physical world, and online shopping behavior on e-commerce sites, which reflects consumption concepts in economic activities. To demonstrate the effectiveness of NSM, we conducted extensive experiments, with a large dataset covering the two-domain activities for hundreds of thousands of individuals. Our results suggest that NSM offers a powerful paradigm for 1) presenting an effective measurement of a personality trait that can explicitly explain the deviation of individuals from the habits of individuals and crowds; 2) uncovering the correlation of novelty-seeking trait at different levels and across heterogeneous domains. The proposed method provides emerging implications for personalized cross-domain recommendation and targeted advertising.

Collaborative filtering meets next check-in location prediction

With the increasing popularity of Location-based Social Networks, a vast amount of location check-ins have been accumulated. Though location prediction in terms of check-ins has been recently studied, the phenomena that users often check in novel locations has not been addressed. To this end, in this paper, we leveraged collaborative filtering techniques for check-in location prediction and proposed a short- and long-term preference model. We extensively evaluated it on two large-scale check-in datasets from Gowalla and Dianping with 6M and 1M check-ins, respectively, and showed that the proposed model can outperform the competing baselines.

Analyzing Location Predictability on Location-Based Social Networks

With the growing popularity of location-based social networks, vast amount of user check-in histories have been accumulated. Based on such historical data, predicting a user’s next check-in place is of much interest recently. There is, however, little study on the limit of predictability of this task and its correlation with users’ demographics. These studies can give deeper insight to the prediction task and bring valuable insights to the design of new prediction algorithms. In this paper, we carry out a thorough study on the limit of check-in location predictability, i.e., to what extent the next locations are predictable, in the presence of special properties of check-in traces. Specifically, we begin with estimating the entropy of an individual check-in trace and then leverage Fano’s inequality to transform it to predictability. Extensive analysis has then been performed on two large-scale check-in datasets from Jiepang and Gowalla with 36M and 6M check-ins, respectively. As a result, we find 25% and 38% potential predictability respectively. Finally, the correlation analysis between predictability and users’ demographics has been performed. The results show that the demographics, such as gender and age, are significantly correlated with location predictability.

Mining Check-In History for Personalized Location Naming

Many innovative location-based services have been established to offer users greater convenience in their everyday lives. These services usually cannot map users physical locations into semantic names automatically. The semantic names of locations provide important context for mobile recommendations and advertisements. In this article, we proposed a novel location naming approach which can automatically provide semantic names for users given their locations and time. In particular, when a user opens a GPS device and submits a query with her physical location and time, she will be returned the most appropriate semantic name. In our approach, we drew an analogy between location naming and local search, and designed a local search framework to propose a spatiotemporal and user preference (STUP) model for location naming. STUP combined three components, user preference (UP), spatial preference (SP), and temporal preference (TP), by leveraging learning-to-rank techniques. We evaluated STUP on 466,190 check-ins of 5,805 users from Shanghai and 135,052 check-ins of 1,361 users from Beijing. The results showed that SP was most effective among three components and that UP can provide personalized semantic names, and thus it was a necessity for location naming. Although TP was not as discriminative as the others, it can still be beneficial when integrated with SP and UP. Finally, according to the experimental results, STUP outperformed the proposed baselines and returned accurate semantic names for 23.6% and 26.6% of the testing queries from Beijing and Shanghai, respectively.

Geo-social media data analytic for user modeling and location-based services

More and more geo-tagged social media data is generated, nowadays, from the geo-tagged tweets, geo-tagged photos to check-ins. Analyzing this flourish data enables the possibility for us to discover users daily mobility patterns, profiles and preferences. As a result, based on the analyzed results, new types of location-based services emerge. In this article, we first introduce the recent advances in location-based user preferences modeling, which includes: 1) inferring users demographics, 2) identifying users novelty-seeking characteristics and 3) discovering users shopping impulsiveness. After that, we present a comprehensive summary on the state-of-arts of the location-based services, which take advantage of the geo-social media, including: 1) location-based recommendations, 2) location-based predication.

GeoMF++: Scalable Location Recommendation via Joint Geographical Modeling and Matrix Factorization

Location recommendation is an important means to help people discover attractive locations. However, extreme sparsity of user-location matrices leads to a severe challenge, so it is necessary to take implicit feedback characteristics of user mobility data into account and leverage the location’s spatial information. To this end, based on previously developed GeoMF, we propose a scalable and flexible framework, dubbed GeoMF++, for joint geographical modeling and implicit feedback-based matrix factorization. We then develop an efficient optimization algorithm for parameter learning, which scales linearly with data size and the total number of neighbor grids of all locations. GeoMF++ can be well explained from two perspectives. First, it subsumes two-dimensional kernel density estimation so that it captures spatial clustering phenomenon in user mobility data; Second, it is strongly connected with widely used neighbor additive models, graph Laplacian regularized models, and collective matrix factorization. Finally, we extensively evaluate GeoMF++ on two large-scale LBSN datasets. The experimental results show that GeoMF++ consistently outperforms the state-of-the-art and other competing baselines on both datasets in terms of NDCG and Recall. Besides, the efficiency studies show that GeoMF++ is much more scalable with the increase of data size and the dimension of latent space.

Lifting the Predictability of Human Mobility on Activity Trajectories

Mobility prediction has recently attracted plenty of attention since it plays an important part in many applications ranging from urban planning and traffic forecasting to location-based services, including mobile recommendation and mobile advertisement. However, there is little study on exploiting the activity information, being often associated with the trajectories on which prediction is based, for assisting location prediction. To this end, in this paper, we propose a Time-stamped Activity INference Enhanced Predictor (TAINEP) for forecasting next location on activity trajectories. In TAINEP, we propose to leverage topic models for dimension reduction so as to capture co-occurrences of different time-stamped activities. It is then extended to incorporate temporal dependence between topics of consecutive time-stamped activities to infer the activity which may be conducted at the next location and the time when it will happen. Based on the inferred time-stamped activities, a probabilistic mixture model is further put forward to integrate them with commonly-used Markov predictors for forecasting the next locations. We finally evaluate the proposed model on two real-world datasets. The results show that the proposed method outperforms the competing predictors without inferring time-stamped activities. In other words, it lifts the predictability of human mobility.

Discrete Deep Learning for Fast Content-Aware Recommendation

Cold-start problem and recommendation efficiency have been regarded as two crucial challenges in the recommender system. In this paper, we propose a hashing based deep learning framework called Discrete Deep Learning (DDL), to map users and items to Hamming space, where a user»s preference for an item can be efficiently calculated by Hamming distance, and this computation scheme significantly improves the efficiency of online recommendation. Besides, DDL unifies the user-item interaction information and the item content information to overcome the issues of data sparsity and cold-start. To be more specific, to integrate content information into our DDL framework, a deep learning model, Deep Belief Network (DBN), is applied to extract effective item representation from the item content information. Besides, the framework imposes balance and irrelevant constraints on binary codes to derive compact but informative binary codes. Due to the discrete constraints in DDL, we propose an efficient alternating optimization method consisting of iteratively solving a series of mixed-integer programming subproblems. Extensive experiments have been conducted to evaluate the performance of our DDL framework on two different Amazon datasets, and the experimental results demonstrate the superiority of DDL over the state-of-the-art methods regarding online recommendation efficiency and cold-start recommendation accuracy.