Odkrivanje koreferenčnosti je ena izmed treh ključnih nalog ekstrakcije informacij iz besedil, kamor spadata še prepoznavanje imenskih entitet in ekstrakcija povezav. Namen odkrivanja koreferenčnosti ...je prek celotnega besedila ustrezno združiti vse omenitve entitet v skupine, v katerih vsaka skupina predstavlja svojo entiteto. Metode za reševanje te naloge se za nekatere jezike z več govorci razvijajo že dalj časa, medtem ko za slovenski jezik še niso bile izdelane. V prispevku predstavljamo nov, ročno označen korpus za odkrivanje koreferenčnosti v slovenskem jeziku – korpus coref149. Za avtomatsko odkrivanje koreferenčnosti smo prilagodili sistem SkipCor, ki smo ga izdelali za angleški jezik. Sistem SkipCor je na slovenskem gradivu dosegel 76 % ocene CoNLL 2012. Ob tem smo analizirali še vplive posameznih tipov značilk in preverili, katere so pogoste napake. Pri analiziranju besedil smo razvili tudi programsko knjižnico s spletnim vmesnikom, prek katere je možno izvesti vse opisane analize in neposredno primerjati njihovo uspešnost. Rezultati analiz so obetavni in primerljivi z rezultati pri drugih, bolj razširjenih jezikih. S tem smo dokazali, da je avtomatsko odkrivanje koreferenčnosti v slovenskem jeziku lahko uspešno, v prihodnosti pa bi bilo potrebno izdelati še večji in kvalitetnejši korpus, v katerem bodo koreferenčno naslovljene vse posebnosti slovenskega jezika, kar bi omogočilo izgradnjo učinkovitih metod za avtomatsko reševanje koreferenčnih problemov.
Razreševanje koreferenc je pomemben del jezikovnih tehnologij, vendar za slovenščino ta tehnologija še ni bila razvita. Obstajajo različne vrste koreferenc, članek se osredotoča predvsem na anafore ...pri osebnih zaimkih. Uporabljenih je bilo sedem metod razreševanja, ki se med seboj dopolnjujejo, najpomembnejša temelji na metodah na osnovi aktivacije. Prvi rezultati so obetavni, za podrobnejšo analizo delovanja pa bo potreben korpus z označenimi primeri. Razreševanje koreferenc je bilo uporabljeno tudi v sistemu za odgovarjanje na vprašanja Piflar, ki zna s tem odgovoriti na več vprašanj, ker mu uspe nadomestiti osebne zaimke, hkrati pa je bil Piflar dopolnjen še z drugimi dodatki, npr. z odgovarjanjem na posamične stavčne člene in na trdilne povedi, izboljšano pa je bilo tudi tvorjenje dolgih odgovorov pri odločevalnih vprašanjih. Razreševanje koreferenc je izboljšalo tudi delovanje strojnega prevajalnika Presis, in sicer pri določanju spola osebnih zaimkov in pri razdvoumljanju prilastkovih odvisnikov.
Overdiagnosis is the diagnosis of deviations, abnormalities, risk factors, and pathologies that in themselves would never cause symptoms (this applies only to risk factors and pathology), would never ...lead to morbidity, and would never be the cause of death. Therefore, treating an overdiagnosed condition (deviation, abnormality, risk factor, pathology) cannot, by definition, improve the patient’s prognosis, and can therefore only be harmful.
Overdiagnosis is an extremely harmful and big problem all over the world, and the problem is increasing. This is especially the case in high-income countries, where more sensitive tests, more testing, more screening and earlier diagnosis is in focus, and more of the same will be implemented in the future. Moreover, disease definitions have been and are still being widened, plus thresholds for treating, e.g. risk factors, have been and are still being lowered. Finally, disease mongering is growing, because it is cheaper and faster to invent new “diseases” than new pharmaceutical drugs.
From the definition of overdiagnosis it can be reasoned that a patient who has been correctly diagnosed and a person who has been overdiagnosed can have the same kind of pathologies. Therefore, at the level of the individual person or patient it can never be verified whether he or she has in fact been correctly diagnosed or overdiagnosed. Therefore, the complexity, dilemmas and pitfalls in understanding what overdiagnosis really is so succinctly captured by this quote from the Danish philosopher S⊘ren Kirkegaard (1813-55): ‘Life can only be understood backwards; but it must be lived forwards’.
In 1996 the World Health Organization declared intimate partner violence (IPV) the most important public health problem. Meta-analyses in 2013 showed every third female globally had been a victim of ...violence. Experts find screening controversial; family medicine is the preferred environment for identifying victims of violence, but barriers on both sides prevent patients from discussing it with doctors.
In July 2014, a qualitative study was performed through semi-structured interviews with ten family doctors of different ages and gender, working in rural or urban environments. Sound recordings of the interviews were transcribed, and the record verified. The data were interpreted using content analysis. A coding scheme was developed and later verified and analysed by two independent researchers. The text of the interviews was analysed according to the coding scheme.
Two coding schemes were developed: one for screening, and the other for the active detection of IPV. The main themes emerging as barriers to screening were lack of time, staff turnover, inadequate finance, ignorance of a clear definition, poor commitment to screening, obligatory follow-up, risk of deterioration of the doctor-patient relationship, and insincerity on the part of the patient. Additionally, cultural aspects of violence, uncertainty/ helplessness, fear, lack of competence and qualifications, autonomy/negative experience, and passive role/stigma/ fear on the part of the patients were barriers to active detection.
All the participating doctors had had previous experience with active detection of IPV and were aware of its importance. Due to several barriers to screening for violence they preferred active detection.
PDF
