Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
LIS618 lecture 2 Dialog by example Thomas Krichel 2011-04-21 task • find out about published literature mentioning Thomas Krichel • Known: – economist – information scientist – founder of RePEc • Limit search to Dialog. Known source Journal of Economic Literature. find the JEL file • • • • • ? b 415 File 415:DIALOG Bluesheets(TM) ? s economic(N)literature S1 1 ECONOMIC(N)LITERATURE ? t s1/ti/all ECONLIT - File 139 • ? b 139 start with econlit • ? krichel • S1 0 KRICHEL • We should have a look at the bluesheet http://library.dialog.com/bluesheets/html/bl0 139.html • We find authors are not in the basic index. we can find out about papers written • s au=thomas krichel S2 0 AU=THOMAS KRICHEL • ? s au=krichel S3 0 AU=KRICHEL • ? s au=krichel,thomas S4 0 AU=KRICHEL,THOMAS • ? s au=krichel, thomas S5 8 AU=KRICHEL, THOMAS • ? s au=krichel, t. S6 0 AU=KRICHEL, T. turning to RePEc • • • • • • • • • ? b 139 ? s repec S1 38 REPEC ? t s1/k/all ? s (repec/ti) or (repec/ab) or (repec/de) 3 REPEC/TI 10 REPEC/AB 30 REPEC/DE S2 38 (REPEC/TI) OR (REPEC/AB) OR (REPEC/DE) evaluating • The ones that have RePEc in the title and abstract are about RePEc or use it extensively. • The ones that have it in the descriptors sometimes have a URL of the form ../wider/…/repec/… meaning the information has been made available as part of RePEc. • The entries for the copyright are on all EconLit entries that come from RePEc. They are not searchable. descriptor records • ? s (repec/de) not (repec/ti) not (repec/ab) 30 REPEC/DE • 27 (REPEC/DE) NOT (REPEC/TI) NOT (REPEC/AB) • ? s (repec/de) not (repec/ti) not (repec/ab) not (wider/de) ? s (repec/de) not (repec/ti) not (repec/ab) not (wider/de) • S4 0 (REPEC/DE) NOT (REPEC/TI) NOT (REPEC/AB) NOT (WIDER/DE) first set • The 3 papers that have RePEc in the title talk about RePEc. • The 8 that have RePEc in the abstract but not in the title use RePEc. We put those into a latter part of our written report to the patron. • There are probably more papers that use RePEc but the use has not been reported. look at infosci • • • • • • • • ? b infosci File 1: ERIC File 2: INSPEC File 7: Social SciSearch(R) File 47: Gale Group Magazine DB File 121: Brit.Education Index File 148: Gale Group Trade & Industry File 438: Library Lit. & Info. Science finding the right krichel • • • • ? s krichel S1 15 KRICHEL ? t s1/k/all Manual inspection reveals that all the records that relate to the patron either are – krichel(N)thomas – krichel(N)t – tkrichel(N)openlib • ? s krichel(N)thomas or krichel(N)t or tkrichel(N)openlib • We get 13 records that deal with the patron in different ways, it is prudent to add all to the report. move on to RePEc • ? s repec not (krichel(N)thomas or krichel(N)t or krichel(N)openlib) • 111 REPEC • 107 REPEC NOT (KRICHEL(N)THOMAS OR KRICHEL(N)T OR KRICHEL(N)OPENLIB) • How to filter this? filtering URLs • First, kwic inspection reveals that a lot of records have URLs of the form http://ideas.repec.org/…. • ? s s1 not (http(N)ideas(N)repec(n)org) • 107 S1 • 4218576 HTTP • 644548 IDEAS • 111 REPEC 551055 ORG • 46 HTTP(N)IDEAS(N)REPEC(N)ORG • S3 61 S1 NOT (HTTP(N)IDEAS(N)REPEC(N)ORG) better filtering • There are some URLs that have a RePEc in the URL as part of the path, so • s s9 not (http(5N)repec(N)org or edu(2N)repec not au=krichel t) • When looking at this, there are about half the papers that appear to be relevant, but I have not been able to find a clever filter for those. http://openlib.org/home/krichel Please shutdown the computers when you are done. Thank you for your attention!