Ambitions for FrameNet

The title of the presentwork comes from Fillmore’s impromptu talkduring aninternationalworkshopaboutFrameNet,held in Berkeleyon April 19, 2013.2The eventbrought together developers of FrameNet styleresources(FONTENELLE, 2003;RUPPENHOFER ET AL.,2010) and constructicons (FILLMORE ET AL.,2012)forBrazilian Portuguese, English, Japanese, and Swedishas well ascomputerscientistsandsoftware developerswhoexploitFrameNet data for natural language processing applications.This paper, initiallyonlyintended to introduce the collection of articlesinthe first part ofthisspecial issue ofVeredasdedicated to Frame Semantics, also provides the opportunity to shareFillmore’s wisdomand recommendationsabout future directions for FrameNet, the researchproject that he founded in 1997.In so doing, thepresentpaperalsoseeks todemonstrate thataspects of each article in this special issuestartto fulfill some of Fillmore’sAmbitions forFrameNet.3The rest of this introductoryinvitedessayisorganized as follows. Section 1presentsFillmore’s ambitions for FrameNet, quotingdirectly from the talkwherepossible, orsummarizing the essence of Fillmore’s contributions, and section 2describes the articles inthefirst part of thecurrentspecial issue ofVeredas, also indicating(where appropriate)howeachoneaddressesthese ambitions. Section 3summarizes another of Fillmore’s ambitionsforFrameNet,offeringa suggestion about how FrameNet might handle the concept of TIME.

Veredas, Vol. 17, No. 1, pp. 5-10

AI