This year we get six main results: one related to how to integrate domain knowledge in a multi-view KDD process (cf. section ), two on new KDD methods involving clustering (cf. sections  and ), one on the construction of hierarchical structures of concepts in the field of e-tourism (cf. section ), one on partitioning objects taking into account simultaneously their relational descriptions given by multiple dissimilarity matrices (cf. section ), and finally improvment of our work on critical edition of Sanskrit texts (cf. section ).

This year, we pursued two main works on clustering methods:

With the expansion of the innovation community beyond the firm's boundaries (the so-called "open innovation") a lot of changes have been introduced in design and evaluation processes : the users can become co-designers, HCI design and evaluation focus is no longer placed on usability only but also on the whole user experience, experimentations take place out of lab with large number of heterogeneous people instead of carefully controlled panels of users ... All these deep changes require improvements of existing practices, methods and tools for the design / evaluation of information systems as well as for usage analysis. This evolution calls also for a structured user centered methodology (methods and ICT tools) to deal with open innovation. Various different disciplines and trends are dedicated in understanding user behavior on Internet and with Digital Technologies, notably Human Computer Interaction community (HCI), CSCW, Workplace Studies, Distributed Cognition and Data Mining. Our contribution to open innovation research keeps its focus on usage analysis for design, evaluation and maintenance of information systems and our activities this year, as indicated in our roadmap presented at the Inria theme evaluation (2011) have been conducted both breadth wise and in depth with two main objectives :

The research was conducted along three focus:

Let us note one research work related to Living labs done in 2011 and published in 2012 .

In , we introduce hard clustering algorithms that are able to partitioning objects taking into account simultaneously their relational descriptions (relations + values) given by multiple dissimilarity matrices. The aim is to obtain a collaborative role of the different dissimilarity matrices in order to obtain a final consensus partition. These matrices could have been generated using different sets of variables and a fixed dissimilarity function or using a fixed set of variables and different dissimilarity functions, or using different sets of variables and dissimilarity functions.

During 2012 we show interest and disadvantages of these approaches to classifying curves with a Urso and Vichi distance based on the mathematical properties of curves (first derivative and second). The curves are issued from temperature sensors placed in 40 offices during one year (See section ). This period was divided into the periods before and after challenge and the challenge period. During the challenge period the occupants had information by bonus / malus messages on energy consumption .

Extracting knowledge from Web user's access data in Web Usage Mining (WUM) process is a challenging task that is continuing to gain importance as the size of the Web and its user-base increase. That is why meaningful methods have been proposed in the literature in order to understand the behaviour of the user in the Web and improve the access modes to information. We pursued our previous work and defined a new approach of knowledge extraction using graph theory. which is described in .

This work is done in collaboration with the laboratory LRIA At the Ferhat Abbas University, Sétif, Algérie.

The method described in uses a nonlinear aggregation criterion, weighted Tchebycheff distances, more appropriate than linear combinations (such as weighted averages) for the construction of compromise solutions. We obtain a partition of the set of objects, the prototype of each cluster and a weight vector that indicates the relevance of each criterion in each cluster. Since this is a clustering algorithm for relational data, it is compatible with any distance function used to measure the dissimilarity between objects.

E.L. Moukhtar Zemmouri, in the context of his PhD thesis supervised by Hicham Behja, A. Marzark and B. Trousse pursued his work based on a Viewpoint Model in the context of a KDD process , .

Knowledge Discovery in Databases (KDD) is a highly complex, iterative and interactive process aimed at the extraction of previously unknown, potentially useful, and ultimately understandable patterns from data. In practice, a KDD process (data mining project according to CRISP-DM vocabulary) involves several actors (domain experts, data analysts, KDD experts, etc.) each with a particular viewpoint. We define a multi-view analysis as a KDD process held by several experts who analyze the same data with different viewpoints.

We propose to support users of multi-view analysis through the development of a set of semantic models to manage knowledge involved during such an analysis. Our objective is to enhance both the reusability of the process and coordination between users.

To do so, we propose first a formalization of viewpoint in KDD and a Knowledge Model that is “a specification of the information and knowledge structures and functions involved during a multi-view analysis”. Our formalization, using OWL ontologies, of viewpoint notion is based on CRISP-DM standard through the identification of a set of generic criteria that characterize a viewpoint in KDD. Once instantiated, these criteria define an analyst viewpoint. This viewpoint will guide the execution of the KDD process, and then keep trace of reasoning and major decisions made by the analyst.

Then, to formalize interaction and interdependence between various analyses according to different viewpoints, we propose a set of semantic relations between viewpoints based on goal-driven analysis. We have defined equivalence, inclusion, conflict, and requirement relations. These relations allow us to enhance coordination, knowledge sharing and mutual understanding between different actors of a multi-view analysis, and reusability in terms of viewpoint of successful data mining experiences within an organization.

With the help of Ehab Assan we improved the prototype made last year by Nicolas Bèchet (cf. 2011 AxIS activity report,). It is now included in the construction process of critical editions of Sanskrit texts. Ehab also added LaTeX output to the process, we now have paper as well as Web output. It was possible to present these new features , at the 13th International Conference on Intelligent Text Processing and Computational Linguistics (CICLing) in Delhi.

The work of Nicolas Bechet (AxIS member in 2011) and Yves Lechevallier in collaboration with Marie-Aude Aufaure (Ecole Centrale de Paris), was published in 2012 related to a method for the construction and the automatic settlement of hierarchical structures of concepts. We were particularly interested in the construction of a hierarchical structure of services offered in Hotels from a data set of an application in the field of e-tourism motivated by our contacts with the SME Addictrip. The goal is to associate to each service a concept that provides a common representation of all services. Our experiments are carried out using resources from partners specialized in online hotel booking, in particular from Addictrip. The establishment of a structure of concepts is essential to these partners that use their own terminologies description of hotel services. Indeed it provides a common representation space allowing the comparison of service coming from different resources. Our approach is based on proximity of literal terms in the service having a nearby measure based on n-grams of characters. The results during our experiments show the quality of this approach and its limitations.

The automatic detection of communities in a social network can provide a kind of graph aggregation. The objective of graph aggregations is to produce small and understandable summaries and it can highlight communities in the network, which greatly facilitates the interpretation.

Social networks allow having a global view of the different actors and different interactions between them, thus facilitating the analysis and information retrieval.

In the enterprise context, a considerable amount of information is stored in relational databases. Therefore, relational database can be a rich source to extract social network.

During this year many updates of the program developed by Louati Amine in 2011 were performed by Bruno Almeida Pimentel. A book chapter, included the new aggregation criteria proposed ans evaluted by Bruno Almeida Pimentel, was written and will be published in 2013.

This work is done in collaboration with Marie-Aude Aufaure, head of the Business Intelligence Team, Ecole Centrale Paris, MAS Laboratory.

In the work reported in in collaboration with Francisco de A.T. de Carvalho, we introduce an improvement of a clustering algorithm described in that is able to partition objects taking into account simultaneously their relational descriptions given by multiple dissimilarity matrices. In this version of the prototype clusters depend on the variables of the representation space. These matrices could have been generated using different sets of variables and dissimilarity functions. This method, which is based on the dynamic clustering algorithm for relational data, is designed to provided a partition and a vector of prototypes for each cluster as well as to learn a relevance weight for each dissimilarity matrix by optimizing an adequacy criterion that measures the fit between clusters and their representatives. These relevance weights change at each algorithm iteration and are different from one cluster to another. Moreover, various tools for the partition and cluster interpretation furnished by this new algorithm are also presented.

Two experiments demonstrate the usefulness of this clustering method and the merit of the partition and cluster interpretation tools. The first one uses a data set from UCI machine learning repository concerning handwritten numbers (digitalized pictures). The second uses a set of reports for which we have an expert classification given a priori. which we have an expert classification given a priori.

In the context of the ANR project PIMI (Personal Information Management through Internet see section ), ergonomic inspections have been carried out to evaluate the usability of the PIMI V0.1. prototype, in its PC and mobile versions , . Also, an experiment , was conducted on a mockup of a Personal Information Space. Users were asked to perform scenario-based data entry and retrieval tasks, then to modify the mockup according to their wishes and needs. The results allowed to validate the item content and structure for the future personal space, as well as to assess the role of user modifications as evaluation cues, and for the development of further ergonomic recommendations. Detailed information was obtained on how users enter and retrieve data, by modifying the interface settings and shows that the adaptable nature of a Personal Information Space can indeed influence its acceptance, and provides useful cues for ergonomic evaluation

A collaboration between UFRGS (Federal University of Rio Grande do Sul, Institute of Informatics), Brazil and Inria-AxIS led to a book chapter dealing with potential methodologies for including a user-centered approach into information visualization techniques. It starts by presenting the evolution of visualization techniques evaluation, briefly summarizing the main contributions in this area since its humble beginning as a collateral activity until the recent growth of interest. Then, the focus is on current issues related to such evaluations, particularly concerning the way they are designed and conducted, taking into account a background of well-known usability evaluation methods from HCI to help understanding why there are still open problems. A set of guidelines for a (more) user-centered usability evaluation of information visualization techniques is proposed and discussed.

A collaboration between University of Toulouse – IUT Tarbes, IRIT and Inria-AxIS led to a book chapter dealing with Mixed Interactive Systems (MIS) which denote an advanced form of interaction that aims at combining physical and digital worlds, such as mixed and augmented reality, tangible user interfaces, ubiquitous computing, etc. Their main interest relates to the use of physical artifacts from the user’s activity customary context. The book chapter first reports on a systematic review of the literature on MIS evaluation. From that review, usability recommendations were selected and deciphered before reformulating them under a common format. Finally, three different classification schemes of the usability recommendations obtained are proposed to facilitate search and retrieval, but also to better integrate them into the MIS development process.

Through an extensive review of the literature, a paper attempted to characterize a rather novel and popular view on human-computer interaction: User Experience (UX). After introducing its polysemous nature, this paper describes the origins of UX, its scope, underlying concepts and components, as well as its various definitions. Then, UX methods are surveyed and classified, distinguishing processes, frameworks, and specific methods. The paper identifies a set of issues about the needs for increased UX maturity. Even though UX can still be viewed as an extension of usability, its future may correspond to a paradigm evolution rather than simply a buzz word. The evolution is not drastic, but it adds complexity (including new measurements) by considering a few more user areas than traditional usability.

In 2011 we proposed a methodology coupling two methods (GenIoT a generative method based on probes (fake sensors and/or actuators) and ALoHa! a bodystorming method for designing service concepts in the specific paradigm of the Internet of Things (IoT). In the frame of the European project ELLIOT - Experiential Living Lab for Internet Of Things -, ICT Usage Lab (cf. section ) aims at co-creating “green” services, i.e. services based on air quality and noise measurement.

Both IoT ideation methods Aloha! and GenIoT were used for the co-creation of health related services (cf. section  ). The participants of the methods were Environment and Health professional. Results were very different than the workshops run with citizen in the frame of the mobility scenario in ELLIOT (see 2011 AxIS activity report). Comparison of these workshops shows that hybrid approaches –i.e. co-creation approaches mixing both real and virtual meetings are not working as well as pure face to face or pure online approaches. Moreover, GenIoT method seems to be more effective with citizen than with professional. Aloha! is effective in both groups but more efficient with professionals. However the participant experience of Aloha! is higher in the case of citizen (mainly because professionals are not used to practice creative thinking methods and do not appreciate to go out of their comfort zone).

The ECOFFICES project (cf. , and section ) was for AxIS project team our first step towards eco-behavior study. It provided us a very rich context to study how to analyse the evolution of the energy consuming of employees during an energy challenge. A qualitative analysis from questionnaires (before and after the challenge) has been done as well quantitative analysis. The data set for quantitative data is composed of heterogeneous data issued from around 400 sensors (temperature, presence, behavior in terms of opening doors, windows, bonus, malus, etc.). We made different studies related to data preprocessing and data analysis.

In our first study , we cleaned the data set and selected reliable data for data analysis (only temperature of various equipments, user presence and bonus/malus points). We decided not to work with aggregated variables such as the initial ponderation (defined by partners) for the various bonus-malus rules and the energy consuming at the office level. We decided to use (office, day) as statistical unit (i.e. 9995 units) with a vectoral representation. Finally we realized that the three initial periods (before the challenge, during and after the challenge) on 379 days (2011-2012) should be in fact decomposed in five periods, due to the fact the first and the last periods were split into two subperiods (with and without heat). For the analysis, we apply for each (office,day) a first analysis on a vectorial representation of temperature with the MND method (cf. section focuslab) in order to identify the best partition of these. The MND method uses euclidean distance between each value of the vectorial representation and the prototypes are defined by the means. Second we did a clustering of these units based on bonus and malus and finally we made the correspondence between these two partitions. Three classes for (office, day) are obtained. The interpretation in terms of team relied difficult but we proposed various conclusions for a winner for managing a specific bonus, or in managing ambient temperature or in behavioral change.

In our second study in collaboration with Francisco de A.T. De Carvalho, our goal was to improve the interpretation task at the office and team level by applying AxIS advanced methods. To do this, we applied our hard clustering method presented in on this dataset where each office was characterized by two different representations:

Our research towards eco-behavior study started with the ECOFFICES project (cf. sections  and for more details) was recently complemented with a litterature review aiming at a deeper understanding of breaks and levers to eco behavior adoption. A first work was focused on the so-called "modal change problem", compiling methods and tools aimed at supporting people to use public transportation system rather than their personal car. A second work was initiated to get a better understanding of the role that users interface could play in encouraging people to adopt a specific behavior. This work is still in progress.

Eco mobility : prompting people to adopt public transportation mode rather than their personal car.

The first review of work conducted in the fields of Persuasive Communication, Commitment, Nudges and Persuasive Technology showed that behavioral change is a process with many steps requiring to support each step with specific means. For instance, if mass communication can support the public awareness of a problem, information is not sufficient to convince people to really change their behavior. It is necessary to push them to act and numerous well-known influence techniques are nowadays available. All recent technological development (geo localization, mobile devices, social networks) can provide very effective support for behavioral changes as far as they rely on design principles identified by research in Persuasive Technology. A presentation was done on this topic for GreenCode Forum (see the video on youtube).

We pursued our work on this topic and contributed to a white paper which is one of the main outcomes of the FIREBALL project [cf. section ), a Coordination Action within the 7th Framework Programme for ICT, running in the period 2010-2012. The aim of this project was to bring together communities and stakeholders who are active in three areas, namely: research and experimentation on the Future Internet (FIRE); open and user-driven innovation in Living Labs; and urban development for smarter cities. The goal was to develop a common vision on how the different methodologies and concepts in these areas can be aligned for cities as playgrounds of open and user driven innovation related to the Future Internet.

The white paper addresses several aspects that are critical for understanding the ‘smart city’ concept and the current progress in this area. Based on cases studies and foresight reports we aim to shed light on how the concept of smart city is currently adopted by European Cities and what the ambitions and expectations are in using this concept. It investigates the drivers and bottlenecks that influence the transformation towards a “smart city”. Underlying approaches to smart cities are discussed, both in terms of the strategies and planning approaches. From this point of view, this paper explores the conditions that must be established to stimulate the transformation towards smart cities, and the resources that are available or should be made available such as investments in broadband networks and in smart applications, as well as in the capabilities to innovate. This also points to the changing structures and processes of innovation and city development. Interestingly, we see a tendency towards more decentralized and bottom-up approaches to planning and innovation. Innovation ecosystems are characterized by a combination of top down and bottom up initiatives, leading to networking and collaboration among stakeholders, which eventually extend to real innovation communities. Increasingly, citizens, advanced businesses and local governments act as proactive catalysers of innovation, shaping cities as “agents of change”.

FocusLab v1.3 (software component) done inside the ELLIOT project (cf. section ) and for the purposes of the CPER Telius (cf. section ) corresponds to the design and the implementation of a set of web-services providing basic and advanced functionalities for data analysis and some other tools supporting the living lab process.

In this version, five data analysis web services are proposed including three generic web services: a classical linear regression and two AxIs methods:

Other data analysis services and tools have been added for Living Labs needs. We propose also two clustering methods which must be downloaded as standalone software and used for mining data from living labs:

The application of the services provided by FocusLab 1.3 and other AxIS data mining methods for the purposes of ELLIOT use cases and other experimental projects are under study.