Peifeng Yin

Location recommendation for location-based social networks

In this paper, we study the research issues in realizing location recommendation services for large-scale location-based social networks, by exploiting the social and geographical characteristics of users and locations/places. Through our analysis on a dataset collected from Foursquare, a popular location-based social networking system, we observe that there exists strong social and geospatial ties among users and their favorite locations/places in the system. Accordingly, we develop a friend-based collaborative filtering (FCF) approach for location recommendation based on collaborative ratings of places made by social friends. Moreover, we propose a variant of FCF technique, namely Geo-Measured FCF (GM-FCF), based on heuristics derived from observed geospatial characteristics in the Foursquare dataset. Finally, the evaluation results show that the proposed family of FCF techniques holds comparable recommendation effectiveness against the state-of-the-art recommendation algorithms, while incurring significantly lower computational overhead. Meanwhile, the GM-FCF provides additional flexibility in tradeoff between recommendation effectiveness and computational overhead.

On the semantic annotation of places in location-based social networks

In this paper, we develop a semantic annotation technique for location-based social networks to automatically annotate all places with category tags which are a crucial prerequisite for location search, recommendation services, or data cleaning. Our annotation algorithm learns a binary support vector machine (SVM) classifier for each tag in the tag space to support multi-label classification. Based on the check-in behavior of users, we extract features of places from i) explicit patterns (EP) of individual places and ii) implicit relatedness (IR) among similar places. The features extracted from EP are summarized from all check-ins at a specific place. The features from IR are derived by building a novel network of related places (NRP) where similar places are linked by virtual edges. Upon NRP, we determine the probability of a category tag for each place by exploring the relatedness of places. Finally, we conduct a comprehensive experimental study based on a real dataset collected from a location-based social network, Whrrl. The results demonstrate the suitability of our approach and show the strength of taking both EP and IR into account in feature extraction.

A straw shows which way the wind blows: ranking potentially popular items from early votes

Prediction of popular items in online content sharing systems has recently attracted a lot of attention due to the tremendous need of users and its commercial values. Different from previous works that make prediction by fitting a popularity growth model, we tackle this problem by exploiting the latent conforming and maverick personalities of those who vote to assess the quality of on-line items. We argue that the former personality prompts a user to cast her vote conforming to the majority of the service community while on the contrary the later personality makes her vote different from the community. We thus propose a Conformer-Maverick (CM) model to simulate the voting process and use it to rank top-k potentially popular items based on the early votes they received. Through an extensive experimental evaluation, we validate our ideas and find that our proposed CM model achieves better performance than baseline solutions, especially for smaller k.

On top-k social web search

To enhance the quality of document search, recent research studies have started to exploit the social networks of users by considering social influence (SI), measurement of the affinity between a query user and the publisher of a retrieved document, in addition to the commonly used textual relevance (TR). We refer to such document search that considers social networks as social web search. In this paper, we focus on efficient top-k social web search and propose two search strategies: (i) TR-based search and (ii) SI-based search that tailor document examination orders upon TR and SI, respectively. We evaluate the proposed strategies through experimentation.

Mining GPS Data for Trajectory Recommendation

The wide use of GPS sensors in smart phones encourages people to record their personal trajectories and share them with others in the Internet. A recommendation service is needed to help people process the large quantity of trajectories and select potentially interesting ones. The GPS trace data is a new format of information and few works focus on building user preference profiles on it. In this work we proposed a trajectory recommendation framework and developed three recommendation methods, namely, Activity-Based Recommendation (ABR), GPS-Based Recommendation (GBR) and Hybrid Recommendation. The ABR recommends trajectories purely relying on activity tags. For GBR, we proposed a generative model to construct user profiles based on GPS traces. The Hybrid recommendation combines the ABR and GBR. We finally conducted extensive experiments to evaluate these proposed solutions and it turned out the hybrid solution displays the best performance.

A Progress Advisor for IT Service Engagements

Monitoring the status of ongoing sales opportunities in IT service engagements is important for sales teams to improve the win rate of deals. Existing approaches aim at predicting the final outcome, i.e., The eventual chance of winning or losing a deal, given a snapshot of the deal data. While this type of prediction indirectly advises on the deal status, it offers limited guidance and insights. During the engagement progress, there occur numerous milestones and key events whose occurrence and status is important in achieving the desired outcome of the deal. These interim milestones and events may happen in different time intervals during the lifecycle of a deal, depending on the deal size and other parameters. In this paper, we describe a novel Bernoulli-Dirichlet predictive model for predicting the occurrence of key events and milestones within a service engagement process to assist in monitoring the progress of ongoing deals. This model enables predicting the timeline and status of the next event(s), given the current history of milestones activity in the engagement lifecycle. Through such a step-by-step guidance, sales teams may have a higher chance of success by knowing of upcoming events, and preparing to counter undesired events. We show the empirical evidences of significance and impact of such an approach in a real-world service provider environment.

Tone Analyzer for Online Customer Service: An Unsupervised Model with Interfered Training

Emotion analysis of online customer service conservation is important for good user experience and customer satisfaction. However, conventional metrics do not fit this application scenario. In this work, by collecting and labeling online conversations of customer service on Twitter, we identify 8 new metrics, named as tones, to describe emotional information. To better interpret each tone, we extend the Latent Dirichlet Allocation (LDA) model to Tone LDA (T-LDA). In T-LDA, each latent topic is explicitly associated with one of three semantic categories, i.e., tone-related, domain-specific and auxiliary. By integrating tone label into learning, T-LDA can interfere the original unsupervised training process and thus is able to identify representative tone-related words. In evaluation, T-LDA shows better performance than baselines in predicting tone intensity. Also, a case study is conducted to analyze each tone via T-LDA output.

A Discrete Constraint-Based Method for Pipeline Build-Up Aware Services Sales Forecasting

Services organizations maintain a pipeline of sales opportunities with different maturity level (belonging to progressive sales stages), lifespan (time to close) and contract values at any time point. As time goes, some opportunities close (contract signed, or lost) and new opportunities are added to the pipeline. Accurate forecasting of contract signing by the end of a time period (e.g., quarterly) is highly desirable to make appropriate sales activity management with respect to the projected revenue. While the problem of sales forecasting has been investigated in general, two specific aspects of sales engagement for services organizations, which entail additional complexity, have not been thoroughly investigated: (i) capturing the growth trend of current pipeline, and (ii) incorporating current pipeline build-up in updating the prediction model. We formulate these two issues as a dynamic curve-fitting problem in which we build a sales forecasting model by balancing the effect of current pipeline data and the model trained based on historical data. There are two challenges in doing so, (i) how to mathematically define such a balance and (ii) how to dynamically update the balance as more new data become available. To address these two issues, we propose a novel discrete-constraint method (DCM). It achieves the balance via fixing the value of certain model parameters and applying a leave-one-out algorithm to determine an optimal free parameter number. By conducting experiments on real business data, we demonstrate the superiority of DCM in sales pipeline forecasting.

An Optimization Approach to Services Sales Forecasting in a Multi-staged Sales Pipeline.

Services organization manage a pipeline of sales opportunities with variable enterprise sales engagement lifespan, maturity levels (belonging to progressive sales stages), and contract values at any given point in time. Accurate forecasting of contract signings by the end of a time period (e.g., a quarter) is a desire for many services organizations in order to get an accurate projection of incoming revenues, and to provide support for delivery planning, resource allocation, budgeting, and effective sales opportunity management. While the problem of sales forecasting has been investigated in its generic context, sales forecasting for services organizations entails the consideration of additional complexities, which has not been thoroughly investigated: (i) considering opportunities in multi-staged sales pipeline, which means providing stage-specific treatment of sales opportunities in each group, and (ii) using the information of the current pipeline build-up, as well as the projection of the pipeline growth over the remaining time period before the end of the target time period in order to make predictions. In this paper, we formulate this problem, considering the service-specific context, as a machine learning problem over the set of historical services sales data. We introduce a novel optimization approach for finding the optimized weights of a sales forecasting function. The objective value of our optimization model minimizes the average error rates for predicting sales based on two factors of conversion rates and growth factors for any given point in time in a sales period over historical data. Our model also optimally determines the number of historical periods that should be used in the machine learning framework to predict the future revenue. We have evaluated the presented method, and the results demonstrate superior performance (in terms of absolute and relative errors) compared to a baseline state of the art method.

Deal or deceit: detecting cheating in distribution channels

Distribution channel is a system that partners move products from manufacturer to end users. To increase sales, it is quite common for manufacturers to adjust the product prices to partners according to the product volume per deal. However, the price adjustment is like a double-edged sword. It also spurs some partners to form a cheating alliance, where a cheating seller applies for a falsified big deal with a low price and then re-sells the products to the cheating buyers. Since these cheating behaviors are harmful to a healthy ecosystem of distribution channel, we need the automatic method to guide the tedious audit process. Thus, in this study we propose the method to rank all partners by the degree of cheating, either as seller or buyer. It is mainly motivated by the observation that the sales volumes of a cheating seller and its corresponding cheating buyer are often negatively correlated with each other. Specifically, the proposed framework consists of three parts: 1) an asymmetric correlation measure which is needed to distinguish cheating sellers from cheating buyers; 2) a systematic approach which is needed to remove false positive pairs, i.e., two partners whose sale correlation is purely coincident; 3) finally, a probabilistic model to measure the degree of cheating behaviors for each partner. Based on the 4-year channel data of an IT company we empirically show how the proposed method outperforms the other baseline ones. It is worth mentioning that with the proposed unsupervised method more than half of the partners in the resultant top-30 ranking list are true cheating partners.