Lightweight adaptation to situational changes in classifiers of multimodal human data multimodal human data: Dissertation

    Research output: ThesisDissertationCollection of Articles

    Abstract

    Intelligent computer applications need to adapt their behaviour to contexts and users, but conventional methods to train multimodal classifiers do not suit to this purpose because they require acquiring large sets of labelled data for each situation. Due to large variety of usage contexts of personal applications, no developer can predict all these situations, to say nothing of collecting adequate training databases for them. Hence personal applications require new methods for adapting to changing runtime contexts. As runtime adaptation largely relies on interaction with end users, these methods should be fairly lightweight with respect to standard ones, i.e. they should require much less domain knowledge and explicitly acquired data.This thesis introduces lightweight solutions for adapting reasoning models to situations at runtime, identifies important context and application characteristics and provides guidelines for considering these factors in adaptation design. The proposed solutions have been validated experimentally with realistic data sets, and the results have confirmed that they considerably reduce the dependence of context- and user-adaptive classifiers on domain knowledge and explicit interaction efforts. Studies with personal assistive applications have also demonstrated that users can accept the proposed lightweight adaptation even when its accuracy is relatively low.
    Original languageEnglish
    QualificationDoctor Degree
    Awarding Institution
    • University of Oulu
    Supervisors/Advisors
    • Seppänen, Tapio, Supervisor, External person
    • Gimel’farb, Georgy, Supervisor, External person
    Award date20 May 2016
    Place of PublicationEspoo
    Publisher
    Print ISBNs978-951-38-8410-9
    Electronic ISBNs978-951-38-8411-6
    Publication statusPublished - 2016
    MoE publication typeG5 Doctoral dissertation (article)

    Fingerprint

    Classifiers
    Computer applications

    Keywords

    • human-computer interaction
    • context adaptation
    • multimodal fusion

    Cite this

    @phdthesis{9911946c2ea24db9a91a9b8e4dc034c5,
    title = "Lightweight adaptation to situational changes in classifiers of multimodal human data multimodal human data: Dissertation",
    abstract = "Intelligent computer applications need to adapt their behaviour to contexts and users, but conventional methods to train multimodal classifiers do not suit to this purpose because they require acquiring large sets of labelled data for each situation. Due to large variety of usage contexts of personal applications, no developer can predict all these situations, to say nothing of collecting adequate training databases for them. Hence personal applications require new methods for adapting to changing runtime contexts. As runtime adaptation largely relies on interaction with end users, these methods should be fairly lightweight with respect to standard ones, i.e. they should require much less domain knowledge and explicitly acquired data.This thesis introduces lightweight solutions for adapting reasoning models to situations at runtime, identifies important context and application characteristics and provides guidelines for considering these factors in adaptation design. The proposed solutions have been validated experimentally with realistic data sets, and the results have confirmed that they considerably reduce the dependence of context- and user-adaptive classifiers on domain knowledge and explicit interaction efforts. Studies with personal assistive applications have also demonstrated that users can accept the proposed lightweight adaptation even when its accuracy is relatively low.",
    keywords = "human-computer interaction, context adaptation, multimodal fusion",
    author = "Elena Vildjiounaite",
    note = "OH: Dissertation",
    year = "2016",
    language = "English",
    isbn = "978-951-38-8410-9",
    series = "VTT Science",
    publisher = "VTT Technical Research Centre of Finland",
    number = "125",
    address = "Finland",
    school = "University of Oulu",

    }

    Lightweight adaptation to situational changes in classifiers of multimodal human data multimodal human data : Dissertation. / Vildjiounaite, Elena.

    Espoo : VTT Technical Research Centre of Finland, 2016. 89 p.

    Research output: ThesisDissertationCollection of Articles

    TY - THES

    T1 - Lightweight adaptation to situational changes in classifiers of multimodal human data multimodal human data

    T2 - Dissertation

    AU - Vildjiounaite, Elena

    N1 - OH: Dissertation

    PY - 2016

    Y1 - 2016

    N2 - Intelligent computer applications need to adapt their behaviour to contexts and users, but conventional methods to train multimodal classifiers do not suit to this purpose because they require acquiring large sets of labelled data for each situation. Due to large variety of usage contexts of personal applications, no developer can predict all these situations, to say nothing of collecting adequate training databases for them. Hence personal applications require new methods for adapting to changing runtime contexts. As runtime adaptation largely relies on interaction with end users, these methods should be fairly lightweight with respect to standard ones, i.e. they should require much less domain knowledge and explicitly acquired data.This thesis introduces lightweight solutions for adapting reasoning models to situations at runtime, identifies important context and application characteristics and provides guidelines for considering these factors in adaptation design. The proposed solutions have been validated experimentally with realistic data sets, and the results have confirmed that they considerably reduce the dependence of context- and user-adaptive classifiers on domain knowledge and explicit interaction efforts. Studies with personal assistive applications have also demonstrated that users can accept the proposed lightweight adaptation even when its accuracy is relatively low.

    AB - Intelligent computer applications need to adapt their behaviour to contexts and users, but conventional methods to train multimodal classifiers do not suit to this purpose because they require acquiring large sets of labelled data for each situation. Due to large variety of usage contexts of personal applications, no developer can predict all these situations, to say nothing of collecting adequate training databases for them. Hence personal applications require new methods for adapting to changing runtime contexts. As runtime adaptation largely relies on interaction with end users, these methods should be fairly lightweight with respect to standard ones, i.e. they should require much less domain knowledge and explicitly acquired data.This thesis introduces lightweight solutions for adapting reasoning models to situations at runtime, identifies important context and application characteristics and provides guidelines for considering these factors in adaptation design. The proposed solutions have been validated experimentally with realistic data sets, and the results have confirmed that they considerably reduce the dependence of context- and user-adaptive classifiers on domain knowledge and explicit interaction efforts. Studies with personal assistive applications have also demonstrated that users can accept the proposed lightweight adaptation even when its accuracy is relatively low.

    KW - human-computer interaction

    KW - context adaptation

    KW - multimodal fusion

    M3 - Dissertation

    SN - 978-951-38-8410-9

    T3 - VTT Science

    PB - VTT Technical Research Centre of Finland

    CY - Espoo

    ER -