In Zhang's thesis (supervised by F. Masseglia), which was partially founded by ANR MIDAS (cf. ), we present our study of the management and mining issues on data streams with evolving tuples, caused by model updates or tuple revisions. For instance, in an online auction system where bids on auction items are streaming, it is possible that some users may bid for more than one item within the user-specified time interval. As a result, the proles of the users can be updated or revised in such applications. Data streams having evolving tuples bring new challenges as well as research opportunity. In this work, he develops novel and efficient models and methods for managing and mining data streams with evolving tuples. (I) To model data streams with evolving tuples, we propose the Anti-Bouncing Streaming model (ABS) for usage streams. ABS fits data streams with evolving tuples and it enables methods for processing of data streams to handle tuple updates or revisions. (II) To find frequent itemsets from data streams with evolving tuples over pane-based sliding windows, we conduct theoretical analysis and propose theorems which can avoid scanning the past slides to check for possible itemsets that may become frequent. We also design novel data structures which can manage the data streams with evolving tuples efficiently and facilitate the mining of frequent itemsets. Moreover, we devise an efficient counting algorithm to verify the frequentness of the candidate frequent itemsets. We also propose two running frameworks for this problem. (III) To extract important feature set from data streams (including the ones with evolving tuples), based upon ABS, we devise the streaming feature set selection algorithm for data streams which is the first in the literature. This method is based on information theory to extract the informative feature sets. To further accelerate the extraction of the most informative feature set from high-dimensional data, we propose a framework that reduces the huge search space to a rather small subset while still guarantee the quality of the discovered feature sets.

In 2011, Chongsheng Zhang has mainly worked on a data stream mining method, intending to extract frequent itemsets. This method has not been published yet and is described in Chapter 5 (page 79) of his thesis document .

In we introduce hard clustering algorithms that are able to partitioning objects taking into account simultaneously their relational descriptions 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.

These methods, which are based on the dynamic hard clustering algorithm for relational data as well as on the dynamic clustering algorithm based on adaptive distances, are designed to furnish a partition and a prototype for each cluster as well as to learn a relevance weight for each dissimilarity matrix by optimizing an adequacy criterion that measures the fitting between clusters and their representatives.

These relevance weights change at each algorithm iteration and can either be the same for all clusters or different from one cluster to another. The usefulness of these partitioning hard clustering algorithms are shown on two time trajectory real world datasets.

In the context of our FACEPE collaboration with Brazil (cf. section  ), we have presented a method which allows clustering of symbolic descriptions constrained by presence rules in a polynomial time instead of a combinatorial one.This method allows to deal with "false missing values". Such a method can be applied on various classification problems .

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. In this work , we are interested in the analysis of the user browsing behavior. The objective is to understand the navigational practices of users (teachers, students and administrative staff). First we clean the data by removing irrelevant information and noise. During the second step, remaining data are arranged in a coherent way in order to identify user sessions. After we defined a new approach of knowledge extraction. This approach treats the data resulting from the preprocessing phase (first and second steps) as being a set of communities. Our approach extends the Modularity measure, proposed by Newman and Girvan , in the Web Mining context in order to benefit from their classifying capacity in the communities discovery.

This work is done in collaboration with the LRIA laboratory – Université Ferhat Abbas, Sétif, Algérie

New progresses concerning the computer assisted elaboration of Sanskrit texts have been made. First Nicolas Béchet and Marc Csernel have worked on the problem of moved texts. After an alignment between two versions of the texts, we discover that some parts of the text apppears to have been moved according to the technics developed in . Until now, we were not able to discover when a text has been moved in a manuscript.

Now using a words-grams technique proposed in , we were able to obtain quite good results on the moved texts problem and we were able to optimize the different possible parameters. A paper on the subject has been submitted to the Cicling 2012 conference ( http:// www. cicling. org/ 2012/ .

After the new treatment related to the moved text problem, we need to provide an interactive display of the critical edition. During his internship, Ehab Hassan has been working on the subject and obtained good results.These results need to be deeply examined by Sanskritists to see if they always fulfill their needs.

The automatic detection of communities in a social network can provide this kind of graph aggregation. The objective of graph aggregations is to produce small and understandable summaries and 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. The extracted network has in general a huge size which makes its analyses and visualization difficult tasks. In , we propose a social network extraction approach from relational database.

Often, the network has a large size which makes its analysis and visualization difficult.

The aggregation step is a necessary task, so we offer and an aggregation step based on the k-SNAP algorithm that produces a summary graph by grouping nodes based on attributes and relationships selected by the user.

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

The task of finding people who are experts on a given topic has cently attracted close attention. State-of-the-art expert finding algorithms uncover knowledge areas of candidate experts based on textual content of associated documents. While powerful, these models ignore social structure that might be available. Therefore, we develop a Bayesian hierarchical model for expert finding that accounts for both content and social relationships. The model assumes that social links are determined by expertise similarity between candidates. The results ofEGC experiments on UvT expert collection have demonstrated the effectiveness of our algorithm .

E. Smirnova visited Intellius, people search technology company (Aug 8 - Oct 5, 2011): the goal of this visit was to validate the research on expert finding in social networks on real dataset and further advance it. As a real dataset, we have taken a sample of United States LinkedIn public profiles. We built an organizational network by connecting a LinkedIn user and his collegues at different workplaces. We also constructed a geographical network from user's current location in the United States. We used Amazon's Mechanical Turk framework ( http:// aws. amazon. com/ code/ 923) to collect user-oriented judgements for model evaluation. We found that the user-oriented model is statistically significantly prefered to the baseline model on 72,5% of queries.

Her work on name disambiguation done in 2010 has been integrated in an article related to the problem of quick detection of top-k Personnalized PageRank (PPR)in . The effectiveness of the chosen approach based on Monte Carlo methods for quick detection of top-k PPR lists has been demonstrated on the Web and Wikipedia graphs.

Yi-Ling Kuo during her internship has worked on Person Name Disambiguation and started by managing the analysis of the very huge Yahoo! Web graph.

This topic has been done in the context of Smirnova's thesis which has been defended on december 15 (thesis supervised by B. Trousse (AxIS) and K.Avrachenkov (Maestro)).

Tweets exchanged over the Internet represent an important source of information, even if their characteristics make them difficult to analyze (a maximum of 140 characters, etc.). In , we define a data warehouse model to analyze large volumes of tweets by proposing measures relevant in the context of knowledge discovery. The use of data warehouses as a tool for the storage and analysis of textual documents is not new but current measures are not well-suited to the specificities of the manipulated data. We also propose a new way for extracting the context of a concept in a hierarchy. Experiments carried out on real data underline the relevance of our proposal.

This work is done inside a collaboration with LIRMM and CEMAGREF.

We pursued our work on articulation of usage mining approach and human factors expertise for the design and evaluation of information systems. Namely, collaborative clustering techniques were used to analyze data issued from users via a card sorting technique, with respect to an a priori ("expert") clustering. Considering the difficulties that people have in managing large information sets in their everyday life, for either professional or non-professional purposes (administration, social relationships, leisure, etc.), our recent research focusses on personal information space for which information bits are currently scattered many places.

In this PIMs field, there is little research with a user-centric approach, with the view that users-based knowledge might help specifying computer-based tools and a state-of-the-art showed little work specifically on usability. Studies address a variety of questions from user needs to accessibility (including studies on older people) or user acceptance, among others. In the context of user-centered and long-term studies to understand the evolution of user information practice, we looked in a study at the intuitive way people organize their personal information, with or without computer systems, in order to help the design of future systems. Also, we recently surveyed 15 tools that claim to support personal information management.

In order to provide to the research community a comprehensive landscape of research streams in the Living Lab domain, we launched a study on the state-of-the-art about the ubiquitous notion of User Experience. During this continuous study, a landscape , has progressively emerged that we organized through 4 main axes: focus granularity (individual/group), user’s role in the design process (observed subject/value creator), collaboration style (structured/unstructured), and evaluation purpose (reliability/adoptability). Our landscape of research streams has been used by Finnish colleagues who conducted an empirical study on the use of the Living Lab research domain landscape as a tool for assessing the maturity level of 16 Finnish Living Labs . The Living Lab research domain Landscape has allowed the study team to identify four categories of Living Labs.

There are many different Internet research areas and corresponding technologies that were already investigated, experimented and progressively deployed such as peer-to-peer, autonomous, cognitive and ad hoc networking, that have already demonstrated how to improve network performance and user experience. Peer-to-peer networking for large-scale distributed systems and widely used applications has proved both the feasibility and economic potential for delivering services to millions of users. Others emerged more recently in the context of the future Internet (FI), such as Cloud Computing for transparently sharing among users scalable elastic resources over a limitless network. As it remains difficult to visualise the conceptual evolution and articulate the various Internet research areas, we conducted a study for identifying the appropriate concepts that could populate the FI domain landscape , over three different periods of time (1990-1999, 2000-2005 and 2006-2011). Several INRIA research teams are involved in FIRE (Future Internet Research Experimentation) Testbed projects, namely: PlanetLab, OneLab, TEFIS, SensLAB, and BonFIRE whose scientific leaders were interviewed during the development of the FI domain landscape. Four dimensions were used for landscaping the Future Internet research domain: evolution approaches (from incremental evolution design to Clean Slate re-design or radical evolution), Internet routing (from the basic data packet delivery towards more sophisticated content distribution and retrieval capacities such as content Centric Networking), network type (from wired communication to wireless communication networks), evolution trend (from computer network towards network computing). The resulting tentative landscape of FI research areas shown in Figure  is intended to provide a faster and broader understanding of the different Internet research streams and related topics.

The goal of the Fireball project is to bring together three different constituencies: user driven open innovation, Future Internet, and Smart Cities , . It aims at defining a roadmap , based on analysis of needs, opportunities and gaps, to benefit a wide scale implementation of the methodologies and concepts elaborated. A first objective in the project was to get a clear picture of the state-of-the-art in each domain. During the review, progressively emerges a landscape that we organized along 4 main axes: wiring (wired/wireless), user's role (subject/actor), Internet evolution approach (structured/unstructured), evaluation purpose (reliability/adoptability). A large variety of FI research have been engaged. If initial efforts in Future Internet research have been directed towards the goal of providing the technical infrastructure supporting the next network generation, a rising trend in this research field is to consider now a higher level layer, the layer of services.

In spite of an important number of tools supporting open innovation, there is few comparative evaluation and no grid or evaluation criteria helping to choose a product. A 2011 review in the Computer Aided Innovation field  Hüsig, University Regensburg & Khon, Otto Beisheim School of Managementprovides a large overview of available tools in relation with a wide range of innovation cycle features. This useful top-down categorization approach is of little help to choose a specific tool. To find the best OI tool supporting idea exchanges among a community of participants, we built an exploratory method on the Web and we elaborate an assessment grid of OI tools based on the QSOS method which is a method designed to qualify, select and compare free and open source software in an objective, traceable and argued way. It publicly available under the terms of the GNU Free Documentation License. In our QSOS method, evaluation criteria are organized in a tree-hierarchy grid with, a scoring method procedure of each tree-leaf criteria (from 0 - not covered to 2 - completely covered). To achieve the construction of the OI tools criteria assessment grid (first step), we used different mining tools for Web crawling, network analysis, criteria classification and from the 29 top rated OI tools, we finally selected 6 of them from which we extracted the tree-map categorization used to build the reference software criteria sheet. In the following steps, after appropriate weighting, we used the provided OS3 Web application and we were able to compare 4 top selected OI tools .

To support our method, we developed a QSOS-based OI Grid for supporting the OI tool selection. The QSOS Grid to compare OI tools, which is in an XML format, has been translated into a MindMap. As a first mock-up, the QSOS Web Interface O3S will be installed at the beginning of 2012 on our server. For the future, we will extend our own QSOS comparative method to other softwares and to other categories interesting for the Living Lab community.

Internet of Things (IoT) is supposed to be a physical world where everyday objects, rooms and machines are connected to one another and to the larger digital world. In this web of people and objects, individuals as well as things will have their own unique URL and, according to interactions between all these entities, social networks will be articulated with a ring of connected objects. This mash up of "hyper groups" and "hyper objects" will be the next step towards a deeper level of automation in which the user interface has vanished, the explicit control over the world is no longer the rule and where the users will have to understand the dynamic changes of their environment in relation to their social interactions and to their physical behaviour. The design of IoT based services therefore raises many challenges related to the user experience and requires a deep understanding of users' needs in their real life environment as well as many field experiments; this is why the Living Lab approach appears as the qualification device which has been lacking so far. As described in a paper accepted for ServDes2012 conference (Fevruary 2012) among 80 papers submitted, we developed and tested two methods designed for Internet of Things service idea generation: GenIOT (Generative Technique for the Internet of Things) and Aloha! (Animation Lens: Object/Human Actoring!) methods.

For the exploration step in the context of ELLIOT, we developed NiceAir, an android mock-up for visualizing air quality data on Google map in Nice as well as some interest points (free bikes, free park places, bus stop, etc.). Such a mock-up will be available in the co-creation step of green services in the ELLIOT project and during the european Mobility Week.

In the context of two contracts (TICTAC and ELLIOT) related to user oriented innovation process, we have developped two mock-ups

Progressive advances in hardware and software technologies have enabled the production and storage of system monitoring data streams in a wide range of fields (e.g. telecommunications, sensor networks, etc.). Traditional clustering methods are unable to deal with data of such a voluminous and dynamic nature. In this work , we propose an efficient clustering approach (ATWUEDA) for monitoring massive time-changing data streams. This work considers a real case study on condition monitoring data streams of an electric power plant provided by EDF.

This work is done in collaboration with Alzennyr Da Silva of BILab laboratory (Telecom ParisTech and EDF R&D Common Laboratory)

The FocusLab platform aims to be a major delivery mechanism of previous and current work in AxIS. It is a way to make methodological contributions (including software) available for the scientific community, but also a way for stimulating further research. This work has slowly started on the software part due to the absence of human resource funding and due to the absence of engineers in the team until 2010. Mid 2011, we started the specification and the development of a first version of a platform with three parts (hardware, software and methods) with the arrival of two engineers on the ELLIOT and TIC TAC contracts. A first version is available since the end of september as a Web portal and a second one is planned for the end of 2012 with advanced features for the software part (cf. service oriented platform, SOA architecture and interoperability).

Related to the software part we are in the process of developing several AxIS methods as Web services: we started with SCDS (cf. ) which was demonstrated in the context of the MIDAS project on two applications (Orange labs mobile portal and vehicle trajectories) and in the ELLIOT platform (linked to the San rafaelle Hospital media use case) at the first review meeting (cf. ) applied on data issued from San Rafaelle Hospital use case. ATWUEDA (cf. ), GEAR (Marascu's thesis) are under development as web services.

Our work on mining evolutive data (ATWUEDA) and data streams (such as SCDS) have been used in real applications in the context of Internet of things and sensors: ATWUEDA (system monitoring for EDF (cf. section ) and SDMS (cf. section ).