Maria Rigou

On-line consistent ranking on e-recruitment: seeking the truth behind a well-formed CV

In this work we present a novel approach for evaluating job applicants in online recruitment systems, using machine learning algorithms to solve the candidate ranking problem and performing semantic matching techniques. An application of our approach is implemented in the form of a prototype system, whose functionality is showcased and evaluated in a real-world recruitment scenario. The proposed system extracts a set of objective criteria from the applicants’ LinkedIn profile, and compares them semantically to the job’s prerequisites. It also infers their personality characteristics using linguistic analysis on their blog posts. Our system was found to perform consistently compared to human recruiters, thus it can be trusted for the automation of applicant ranking and personality mining.

The Role of Context in m-Commerce and the Personalization Dimension

The rapid development of mobile devices contributed to the introduction of m-commerce whose growth is expected to exceed that of e-commerce. The interface usability of mobile applications is a critical factor for the acceptance of m-commerce, as a good interface design allows users achieve high performance when using mobile applications. However, this it is especially difficult to achieve due to certain limitations of mobile devices; limited processing power, low-bandwidth communication, small display and overall device size, limited mobile Internet coverage, delays in standardization and poor quality of service. In this paper, we discuss the particular characteristics of the mobile setting, survey available design guidelines and stress the crucial role personalization can play for alleviating the aforementioned problems and allowing m-commerce reach its full potential

Full-Coverage Web Prediction based on Web Usage Mining and Site Topology

Understanding and modeling user online behavior, as well as predicting future requests remain an open challenge for researchers, analysts and marketers. In this paper, we propose an efficient prediction schema based on the extraction of sequential navigation patterns from server log files, combined with web site topology. Traversed paths are monitored, internally recorded and cleaned before being completed with cashed page views. After session and episode identification follows the construction of n-grams. Prediction is based upon a 5 + n-gram schema with all lower level n-grams participating, a procedure that resembles the construction of an All 5th-order Markov Model. The schema achieves full coverage while maintaining competitive prediction precision.

A method for personalized clustering in data intensive web applications

The paper introduces an algorithm for personalized clustering based on a range tree structure, used for identifying all web documents satisfying a set of predefined personal user preferences. The returned documents go through a clustering phase before reaching the end user, thus allowing more effective manipulation and supporting the decision making process. The proposed algorithm demonstrates increased applicability in semantic web settings, since they offer the infrastructure for the explicit declaration of web document attributes and their respective values, thus allowing for more automated retrieval. The proposed algorithm improves the k-means range algorithm, as it uses the already constructed range tree (i.e. during the personalized filtering phase) as the basic structure on which the clustering step is based, applying instead of the k-means, the k-windows algorithm. The total number of parameters used for modeling the web documents dictates the number of dimensions of the Euclidean space representation. The time complexity of the algorithm is O(logd-2n+v), where d is the number of dimensions, n is the total number of web documents and v is the size of the answer.

A computational geometry approach to Web personalization

In this paper we present an algorithm for efficient personalized clustering. The algorithm combines the orthogonal range search with the k-windows algorithm. It offers a real-time solution for the delivery of personalized services in online shopping environments, since it allows on-line consumers to model their preferences along multiple dimensions, search for product information, and then use the clustered list of products and services retrieved for making their purchase decisions.

Population-ethnic group specific genome variation allele frequency data: a querying and visualization journey.

National/ethnic mutation databases aim to document the genetic heterogeneity in various populations and ethnic groups worldwide. We have previously reported the development and upgrade of FINDbase (www.findbase.org), a database recording causative mutations and pharmacogenomic marker allele frequencies in various populations around the globe. Although this database has recently been upgraded, we continuously try to enhance its functionality by providing more advanced visualization tools that would further assist effective data querying and comparisons. We are currently experimenting in various visualization techniques on the existing FINDbase causative mutation data collection aiming to provide a dynamic research tool for the worldwide scientific community. We have developed an interactive web-based application for population-based mutation data retrieval. It supports sophisticated data exploration allowing users to apply advanced filtering criteria upon a set of multiple views of the underlying data collection and enables browsing the relationships between individual datasets in a novel and meaningful way.

Personalization in the semantic web era: a glance ahead

The problem of information overload when browsing and searching the web becomes more and more crucial as the web keeps growing exponentially and personalization features as the most popular remedy for it. The evolution of personalization as we experience it in recent years has been dramatically influenced by web mining, a research area developing around three main axes: web content, web usage and web structure mining. Lately most research efforts have moved towards combining techniques from more than one domain to achieve extreme personalization in large and complicated web structures. The

Designing for ease is designing for all: Experiences from a simplified office suite

Aims to present the work done in the development of a simplified office suite for disabled and focus on the use of technology applied to the area of “designing for all”. The paper presents an overview of the state‐of‐the‐art in the design for all. It provides practical references to techniques used. The main scope of the paper is to explore the developed technology and give details for the adopted mechanisms. It provides information about designing and implementing software applications for disabled and present a case study for mentally disabled. The paper presents a system that can be used by a specific target group. For this reason, it should be used as reference point for this group, although several techniques can be used for other user categories. The paper is a very useful presentation of an actual system that has been designed and implemented to cover the needs of disabled, useful for interaction with designers and researchers in assistive technology, and it fulfils the need for demonstrative technology in the area of designing for all.

An e-Recruitment System Exploiting Candidates’ Social Presence

Applicant personality is a crucial criterion in many job positions. Choosing applicants whose personality traits are compatible with job positions has been shown to increase their satisfaction levels, as well as the rate of employee retention. However, the task of assessing candidates’ personality is not addressed in today’s online recruitment systems, but is typically handled during the interview process. The rapid deployment of social web services has made candidates’ social activity much more transparent, giving us the opportunity to infer features of candidate personality with web mining techniques. In this work, a novel approach is proposed and evaluated for automatically extracting candidates’ personality traits based on their social media use.

Model Cloning: A Push to Reuse or a Disaster?

The paper focuses on evaluating and refactoring the conceptual schemas of Web applications. The authors introduce the notion of model clones, as partial conceptual schemas that are repeated within a broader application model and the notion of model smells, as certain blocks in the Web applications model, that imply the possibility of refactoring. A methodology is illustrated for detecting and eval- uating the existence of potential model clones, in order to identify problems in an applications conceptual schema by means of efficiency, consistency, usability and overall quality. The methodology can be deployed either in the process of design- ing an application or in the process of re-engineering it. Evaluation is performed according to a number of inspection steps. At first level the compositions used in the hypertext design are evaluated, followed by a second level evaluation concerning the manipulation and presentation of data to the user. Next, a number of metrics is defined to automate the detection and categorization of candidate model clones in order to facilitate potential model refactoring. Finally, the paper proposes a number of straightforward refactoring rules based on the categorization and discusses the aspects affecting the automation of the refactoring procedure.