In various Web applications, users consume content in a series of sessions. That is prevalent in online music listening, where a session is a channel and channels are listened to in sequence, or in crowdsourcing, where a session is a set of tasks and task sets are completed in sequence. Content diversity can be defined in more than one way, e.g., based on artists or genres for music, or on requesters or rewards in crowdsourcing. A user may prefer to experience diversity within or across sessions. Naturally, intra-session diversity is set-based, whereas, inter-session diversity is sequence-based. This novel multi-session diversity gives rise to four bi-objective problems with the goal of minimizing or maximizing inter and intra diversities. Given the hardness of those problems, we propose to formulate a constrained optimization problem that optimizes inter diversity, subject to the constraint of intra diversity. We develop an efficient algorithm to solve our problem. Our experiments with human subjects on two real datasets, music and crowdsourcing, show our diversity formulations do serve different user needs, and yield high user satisfaction. Our large data experiments on real and synthetic data empirically demonstrate that our solution satisfy the theoretical bounds and is highly scalable, compared to baselines.

Workers are the most important resource in crowdsourcing. However, only investing in worker-centric needs, such as skill improvement, often conflicts with short-term platform-centric needs, such as task throughput. This paper studies learning strategies in task assignment in crowdsourcing and their impact on platform-centric needs. We formalize learning potential of individual tasks and collaborative tasks, and devise an iterative task assignment and completion approach that implements strategies grounded in learning theories. We conduct experiments to compare several learning strategies in terms of skill improvement, and in terms of task throughput and contribution quality. We discuss how our findings open new research directions in learning and collaboration.

International audience

International audience

We study task composition in crowdsourcing and the effect of personalization and diversity on performance. A central process in crowdsourcing is task assignment, the mechanism through which workers find tasks. On popular platforms such as Amazon Mechanical Turk, task assignment is facilitated by the ability to sort tasks by dimensions such as creation date or reward amount. Task composition improves task assignment by producing for each worker, a personalized summary of tasks, referred to as a Composite Task (CT). We propose different ways of producing CTs and formulate an optimization problem that finds for a worker, the most relevant and diverse CTs. We show empirically that workers' experience is greatly improved due to personalization that enforces an adequation of CTs with workers' skills and preferences. We also study and formalize various ways of diversifying tasks in each CT. Task diversity is grounded in organization studies that have shown its impact on worker motivation [33]. Our experiments show that diverse CTs contribute to improving outcome quality. More specifically, we show that while task throughput and worker retention are best with ranked lists, crowdwork quality reaches its best with CTs diversified by requesters, thereby confirming that workers look to expose their “good” work to many requesters.

We examine the applicability of Composite Items (CIs) for generating customized travel packages consisting of Points of Interest (POIs) in a given city. CIs have been shown to serve complex information needs such as selecting books for a reading club, identifying a set of products for a promotion, or planning a city tour. In the travel domain, a synthesized view of travel options in a city can be provided with a set of cohesive CIs, each of which is covering a different region in the city. In this paper, we attempt to understand the benefit of letting users customize travel packages, and examine the relationship between customization and personalization. For personalization, we gather user preferences on POI features when available or on latent topics extracted from POI tags. For customization, we develop a framework within which a user interacts with proposed travel packages and the system suggests new CIs according to refined user preferences. Our experiments reveal a tension between personalization and the cohesiveness of items forming each CI. As a result, customization is necessary to find a balance between POI personalization and CI cohesiveness. We also show that the refined user preferences obtained from customization in one city help build better travel packages in another city.

International audience

International audience

Crowdsourcing has gained popularity in a variety of domains as an increasing number of jobs are "taskified" and completed independently by a set of workers. A central process in crowdsourcing is the mechanism through which workers find tasks. On popular platforms such as Amazon Mechanical Turk, tasks can be sorted by dimensions such as creation date or reward amount. Research efforts on task assignment have focused on adopting a requester-centric approach whereby tasks are proposed to workers in order to maximize overall task throughput, result quality and cost. In this paper, we advocate the need to complement that with a worker-centric approach to task assignment, and examine the problem of producing, for each worker, a personalized summary of tasks that preserves overall task throughput. We formalize task composition for workers as an optimization problem that finds a representative set of k valid and relevant Composite Tasks (CTs). Validity enforces that a composite task complies with the task arrival rate and satisfies the worker's expected wage. Relevance imposes that tasks match the worker's qualifications. We show empirically that workers' experience is greatly improved due to task homogeneity in each CT and to the adequation of CTs with workers' skills. As a result task throughput is improved.

As the use of crowdsourcing spreads, the need to ensure the quality of crowdsourced work is magnified. While quality control in crowdsourcing has been widely studied, established mechanisms may still be improved to take into account other factors that affect quality. However, since crowdsourcing relies on humans, it is difficult to identify and consider all factors affecting quality. In this study, we conduct an initial investigation on the effect of crowd type and task complexity on work quality by crowdsourcing a simple and more complex version of a data extraction task to paid and unpaid crowds. We then measure the quality of the results in terms of its similarity to a gold standard data set. Our experiments show that the unpaid crowd produces results of high quality regardless of the type of task while the paid crowd yields better results in simple tasks. We intend to extend our work to integrate existing quality control mechanisms and perform more experiments with more varied crowd members.

Due to the amount of work needed in manual sentiment analysis of written texts, techniques in automatic sentiment analysis have been widely studied. However, compared to manual sentiment analysis, the accuracy of automatic systems range only from low to medium. In this study, we solve a sentiment analysis problem by crowdsourcing. Crowdsourcing is a problem solving approach that uses the cognitive power of people to achieve specific computational goals. It is implemented through an online platform, which can either be paid or volunteer-based. We deploy crowdsourcing applications in paid and volunteer-based platforms to classify teaching evaluation comments from students. We present a comparison of the results produced by crowdsourcing, manual sentiment analysis, and an existing automatic sentiment analysis system. Our findings show that the crowdsourced sentiment analysis in both paid and volunteer-based platforms are considerably more accurate than the automatic sentiment analysis algorithm but still fail to achieve high accuracy compared to the manual method. To improve accuracy, the effect of increasing the size of the crowd could be explored in the future.

SuperSQL is an extension of SQL that automatically formats data retrieved from the database into various kinds of application data (HTML, PDF...). Current developments lead us to identify improvement points and remodel the design of the SuperSQL architecture. Among them, in the current SuperSQL version, the emptiness of one single relation leads to the emptiness of the entire table forming the output data. This is because the process handling the retrieval of desired data does not consider the schema representation of the data and thus does not identify independence between data lists. In this paper, we propose a new process of data retrieval based on a three layers model: the definition layer, the equivalence layer and the optimisation layer. As a result, our proposed architecture is able to manage empty sets and allows easier integration to support future developments.