At present, most research on the fairness of recommender systems is conducted either from the perspective of customers or from the perspective of product (or service) providers. However, such a practice ignores the fact that when fairness is guaranteed to one side, the fairness and rights of the other side are likely to reduce. In this paper, we consider recommendation scenarios from the perspective of two sides (customers and providers). From the perspective of providers, we consider the fairness of the providers' exposure in recommender system. For customers, we consider the fairness of the reduced quality of recommendation results due to the introduction of fairness measures. We theoretically analyzed the relationship between recommendation quality, customers fairness, and provider fairness, and design a two-sided fairness-aware recommendation model (TFROM) for both customers and providers. Specifically, we design two versions of TFROM for offline and online recommendation. The effectiveness of the model is verified on three real-world data sets. The experimental results show that TFROM provides better two-sided fairness while still maintaining a higher level of personalization than the baseline algorithms.

Ctrip.com is a leading online travel company in China, providing travel services including accommodation and flight reservations. With high quality customer service in mind, Ctrip.com has made substantial investments in establishing its own in-house call center. In order to optimize customer satisfaction while controlling operational costs, call center management must determine staffing level at least a week in advance. Our paper addresses this issue by proposing a methodology that predicts the upcoming week's call volume in half-hour increments. The proposed method is a hybrid of a time series TBATS model (Trigonometric seasonal formulation, Box-Cox transform, ARIMA errors, Trend, and Seasonal components), a model of processing time series, and boosting tree theory. For high prediction accuracy, we also considered the effects of related events such as flight delays and ticket sales, and illustrated how to incorporate them as exogenous variables in the tree model part. With this approach, call center management can arrange staffing availability efficiently based on call volume predictions.

When a passenger is going to book an airline flight, after he inputs his query conditions, he has to search from a long flight list to find the one that best matchs his preferences. Therefore, a friendly personalized flight recommender system is a necessary tool for online travel service companies. However, the majority of current approaches do not work well since in most cases, there is no enough historical order information on the route the customers may plan to take. In this paper, we propose an approach for building a Personalize Flight recommendation Service (PFS). In PFS, the attributes' information of flight orders is organized into several domains, and a tensor factorization method is used to recommend flights to users. Specifically, PFS applies a history-based weighting strategy to better model users' biases. PFS is also implemented in a distributed and parallel way to increase its execution speed when dealing with a huge amount of flight data. Extensive experiments on both a real flight order dataset and a synthetic large-scale dataset show the superiority of our model against other comparative ones.

Online periodic prediction is very common in practical problems like stock index analysis or traffic status estimation, where a prediction service is invoked at every time point. However, due to the potential communication cost, service overheads or the expenditure incurred by every prediction service invocation, it is valuable to reduce the number of invocation times without sacrificing too much precision. Therefore, we present an approach based on reinforcement learning to select the timing for prediction service invocations, which still guarantees high quality in capturing the regularity in a future time series. Extensive experiments are conducted on a real-world airfare search volume dataset of multiple routes. The experiment results prove that our algorithms show superiority over other solutions.

Personalized flight recommendation is useful when a traveler wants to find the most suitable flight ticket efficiently. It is hard to learn latent factors behind personal choices directly from users' preferred flight tickets whose features change over time, which leads to the ineffectiveness of traditional recommendation approaches. Therefore, we propose a personalized flight ticket recommendation approach based on paired choice modeling. In this approach, an optimization problem is formulated to maximize a user's choice utility over flight tickets through a paired-choice analysis of users' historical orders. The experiments conducted on a real world dataset show that our approach outperforms other state-of-the-art methods.

In order for travelers to arrive at their destination comfortably, it is important that they plan their trip carefully. Our motivation in this paper is to free the traveler from tedious planning tasks. Several studies and applications have been proposed to help travelers plan their trip according to explicit preferences, but most ignore implicit preferences. For the first time, to the best of our knowledge, we propose a Personalized and Optimized Door-to-door TRavel Planning two-stage approach named PD-TRP, which combines different kinds of transportation services to provide high quality and cost-effective travel plan recommendations, taking into consideration individual explicit and implicit preferences in real time. In order to recommend a set of suitable plans with explicit attribute values, we propose an efficient heuristic search algorithm to generate the top-K candidate plans, and rank them according to the information entropy of implicit preferences discovered from the travelers' previous travel behaviour. We simulate information on transportation services, crawled from the Internet, on 2000 travelers and their past travel behaviour over the last three years, collected from an online travel company, and extensive experimental results show the two-stage approach PD-TRP can effectively improve the degree of satisfaction of the travelers.

The allocation of class seats on a flight in airline industry is closely connected to multiple correlating factors, including yield management and airfare strategy from airlines, policy regulation and price modification from travel agencies, booking and reservation behavior from customers. Different from various machine learning methods targeting at direct fare or price prediction, we constructed a state predictor of class seats by applying a Naïve Bayes algorithm based on Multinomial Event Model on the core flight reservations inventory big data, to tell the probability of class availability within the next several hours or days. Four fundamental models and one integrated model are developed to propose an optimal decision to the airfare search engine layer, which makes the engine be capable of forecasting a smart buy-or-wait suggestion to customers. In our experimental route from SHA to TYO, the integrated model reaches an average of 95.42% accuracy.

Modern airlines use sophisticated pricing models to maximize their revenue, which results in highly volatile airfares. Without sufficient information, it is usually difficult for ordinary customers to estimate future price changes. Over the last few years, several studies have tried to solve the problem of optimal purchase timing for flight tickets, in which the prediction task is described as a binary classification concerning to buy or wait at a given point. However, forecasting the real-time price changes has never received much attention from the research community. In this paper, we address the problem of airfare forecast and present a systematic approach that covers the most important aspects of building a forecast service, including data modelling, forecast algorithm and long-term prediction strategies. A novel matrix-like data schema is first introduced to organize price series and extract temporal features. For the prediction task, we specifically investigate Learn++.NSE, an incremental ensemble classifier designed for learning in nonstationary environments. We propose a modification of the original algorithm to make a regressor that is capable of learning incrementally from streaming price series, with an extra ability of multi-step ahead forecasting. We further evaluate the forecast model on real-world price data collected from diverse routes and discuss its performance with respect to short-term and long-term prediction.

In this paper, we propose an approach to estimate the worst-case response time (WCRT) of each task in a preemptive multitasking single-processor real-time system utilizing an L1 cache. The approach combines intertask cache-eviction analysis and intratask cache-access analysis to estimate the number of cache lines that can possibly be evicted by the preempting task and also be accessed again by the preempted task after preemptions (thus requiring the preempted task to reload the cache line(s)). This cache-reload delay caused by preempting task(s) is then incorporated into WCRT analysis. Three sets of applications with up to six concurrent tasks running are used to test our approach. The experimental results show that our approach can tighten the WCRT estimate by up to 32% (1.4X) over prior state-of-the-art.

In this paper, we investigate the problem of inter-task cache interference in preemptive multi-tasking real-time systems. A prioritized cache is used to reduce cache conflicts among tasks by partitioning the cache. Cache partitions are assigned to tasks according to their priorities. We extend a known tool, SYMTA, in order to estimate the Worst Case Execution Time of tasks executing on a uniprocessor with a unified prioritized L1 cache. Furthermore, we apply a formal timing analysis approach to estimate the Worst Case Response Time of tasks using the prioritized cache. The prioritized cache is compared to a conventional set associative cache of the same size. Our experiments show that the WCRT estimate can be reduced up to 50% when a prioritized cache is used.

In this paper, we propose an approach to estimate the worst case response time (WCRT) of tasks in a preemptive multi-tasking single-processor real-time system with a set associative cache. The approach focuses on analyzing the cache reload overhead caused by preemptions. We combine inter-task cache eviction behavior analysis and path analysis of the preempted task to reduce, in our analysis, the estimate of the number of cache lines that can possibly be evicted by the preempting task (thus requiring a reload by the preempted task). A mobile robot application which contains three tasks is used to test our approach. The experimental results show that our approach can tighten the WCRT estimate by up to 73% over prior state-of-the-art.