Overview

The pos­i­tion is forseen to do research and devel­op­ment in the BMBF-fun­ded pro­ject OKS. This pro­ject aims at the optim­iz­a­tion of human-machine inter­ac­tion with learn­ing con­ver­sa­tional speech inter­faces. Machine learn­ing is already suc­cess­fully used in research for the imple­ment­a­tion of intel­li­gent and nat­ural speech-based per­sonal assist­ants. However, even large ICT enter­prises do not over auto­matic meth­ods for the gen­er­a­tion of dia­log sys­tems.
The aim of the pro­ject is to fill this imple­ment­a­tion gap with tried and tested tools and work pro­cesses for the Ger­man eco­nomy in order to enable the cost-effi­cient devel­op­ment of such speech-based assist­ants. To this end, adapt­ive dia­logue sys­tems shall be cre­ated auto­mat­ic­ally from know­ledge bases and optim­ized by using machine learn­ing, espe­cially rein­force­ment learn­ing. On the basis of this tech­no­logy, sys­tems will be cre­ated that enable speech access to com­plex data struc­tures in every­day pro­fes­sional life and speech-based sup­port in every­day private life. The solu­tions avail­able in basic research have to be val­id­ated with regard to their usage for dia­log sys­tem archi­tec­tures and eval­u­ated with users.
If you have the wish and the right suit­ab­il­ity, you can pur­sue a doc­tor­ate at our chair.

  • The con­crete tasks include, depend­ing on pre­vi­ous know­ledge:
    Deriv­a­tion of dia­log domains and dia­log man­agers from onto­lo­gies and data­base sys­tems
  • Con­cep­tion of dia­logue strategies and dia­logue sys­tems
  • Imple­ment­a­tion of a rein­force­ment learn­ing envir­on­ment for stat­ist­ical dia­logue man­agers (e.g. based on Q-Learn­ing or neural net­works)
  • Con­cep­tion of suit­able Reward func­tions
  • Imple­ment­a­tion of heur­istic and data-driven user sim­u­la­tions
  • Integ­ra­tion of the dia­log man­ager in spoken dia­log sys­tems
  • Eval­u­ation of spoken dia­log sys­tems
  • The oppor­tun­ity for work­ing on a doc­toral dis­ser­ta­tion is given.

Require­ments:

  • Suc­cess­fully com­pleted uni­versity degree (Mas­ter, Dip­lom or equi­val­ent) in com­puter sci­ence, com­pu­ta­tional lin­guist­ics or a com­par­able field of study
  • Good pro­gram­ming skills in Python and secure hand­ling of GIT as well as unit tests, ideally exper­i­ence with Con­tinu­ous Integ­ra­tion
  • Good know­ledge in deal­ing with onto­lo­gies, SQL or No-SQL data­bases
  • Interest in con­duct­ing empir­ical stud­ies with test par­ti­cipants or expert eval­u­ations for the eval­u­ation of spoken dia­logue sys­tems
  • Lan­guage skills: Ger­man at the level of a nat­ive speaker, Eng­lish flu­ent in spoken and writ­ten form
  • Good skills in the field of machine learn­ing (train­ing and eval­u­ation), ideally exper­i­ence with rein­force­ment learn­ing
  • Ideally prior exper­i­ence in the devel­op­ment of spoken dia­log sys­tems or nat­ural lan­guage pro­cessing
  • A desire to work in an inter­na­tional and inter­dis­cip­lin­ary envir­on­ment

How to ap­ply:

Please send your writ­ten applic­a­tion with the ref­er­ence num­ber and the usual doc­u­ments (in par­tic­u­lar letter of application, curriculum vitae, copies of certificates, job references) to Tech­nis­che Uni­versität Ber­lin – Der Präsid­ent – Fakultät IV, Institut für Softwaretechnik und Theoretische Informatik, Quality and Usability Lab, Prof. Dr. Möller, Sekr. TEL 18, Ernst-Reuter-Platz 7, 10587 Berlin or by e-mail to qu-jobs@lists.tu-berlin.de.

To ensure equal oppor­tu­nit­ies bet­ween women and men, app­li­ca­ti­ons by women with the requi­red qua­li­fi­ca­ti­ons are expli­citly desi­red.
Qua­li­fied indi­vi­du­als with disa­bi­li­ties will be favo­red. The TU Berlin values the diversity of its members and is committed to the goals of equal opportunities.

Please send cop­ies only. Ori­gi­nal docu­ments will not be retur­ned.

The offer is under the reserve that funds are gran­ted – part-time employment may be pos­sible.

Tagged as: , , , , , , , ,

You can apply to this job and others using your online resume. Click the link below to submit your online resume and email your application to this employer.

About Technische Universität Berlin, Quality and Usability Lab

The major­ity of sys­tems and ser­vices that are provided by com­puter sci­ence, elec­trical engin­eer­ing and inform­a­tion tech­no­logy finally are ori­ented towards the needs of their human users. To suc­cess­fully build such sys­tems and ser­vices it is essen­tial to invest­ig­ate and under­stand users and their beha­vior when inter­act­ing with tech­no­logy. On the hand, this under­stand­ing allows to derive design prin­ciples for human-machine-inter­faces and to define require­ments con­cern­ing the sys­tem and its under­ly­ing tech­no­lo­gies. On the other hand, new avail­able tech­no­lo­gies do offer new pos­sib­il­it­ies regard­ing the design of such inter­faces, and the devel­op­ment of new kinds of inter­ac­tion.
The Qual­ity and Usab­il­ity Lab is part of TU Ber­lin’s Fac­ulty IV for Elec­trical Engin­eer­ing and Com­puter Sci­ence and deals with the design and eval­u­ation of human-machine inter­ac­tion. Here, the main sub­jects of our research are human per­cep­tion, aspects of tech­nical sys­tem that are related to the inter­ac­tion, as well as inter­ac­tion design.