-- all -- People Communities [Browse] [Tag cloud]
<ul><li>present: Michaela, David, Timo</li><li>Allgemeines:</li><ul><li>übliche Anwesenheit in Golm</li><ul><li>es ist generell ganz nett, abends bevor man geht noch kurz zu sagen, wann man das nächstemal kommt </li></ul><ul><li>David: Mo, (Di), Fr</li><li>Timo: Meeting-Montage, Di, (Mi), Do</li><li>Michaela: mal abwarten wie schön ihre Wohnung wird</li></ul><li>ansonsten nicht scheuen, auch per IM/E-Mail zu kommunizieren (IM-Status richtig einstellen)</li><li>zur besseren Kommunikation im Institut gibt es realistisch drei Möglichkeiten</li><ul><li>Coffeetalks/Vormittagstalks</li><li>unregelmäßige Postersessions</li><li>häufigeres Kolloquium mit internen Vorträgen</li><li>dabei geht es nicht nur um den wissenschaflichen Austausch, sondern auch um allgemein organisatorische Dinge. <br /></li><li>David schreibt ne Mail an die Professoren um zu erfahren, ob ansonsten auch Interesse an mehr Austausch besteht</li></ul><li>Urlaubsplanung zu Weihnachten:</li><ul><li>Michaela: weiß nicht genau, zwischen den Jahren in Augsburg, danach wohl Umzug</li><li>David: 21.-26. Dezember, ansonten strebsam</li><li>Timo: faul vom 21. Dezember bis 4. Januar in Hamburg/Kiel</li></ul><ul><li>also müssen unsere Ergebnisse so weit wie möglich schon vor Weihnachten stehen</li></ul></ul><li>Paper</li><ul><li>Selling point: Syntax (esp. incrementally)</li><li>What's a turn? <br /></li><ul><li>we should probably distinguish more between turn yield and turn hold instead of just EOT</li><li>we want to be able to try different experiments, thus our feature vectors will be "verdongelt" afterwards and must be able to be identified in different settings</li><ul><li>Syntax: dialogID, wordID</li><li>Prosody: dialogID, channel, time</li></ul><li>currently, the mapping in *words.csv uses information from MSaligned/ and from /projekte/korpora/orig-korpora/pennTreebank/dysfl/mgd/swbd</li><li>Timo will review the scripts that create the mapping until next tuesday</li></ul><li>class feature</li><ul><li><em>n</em> words still missing<br /></li><li>time to EOT (from end of this word)<br /></li><li>make it easier for the parser:</li><ul><li>possible completion points (PCP?)<br /></li></ul><li>I'm still not sure we completely acknowledge the fact that we are in some areas dealing with EOT and sometimes with EOU. </li></ul><li>features</li><ul><li>additional features<br /></li><ul><li>expected POS after current word </li></ul></ul><ul><ul><li>flags: seen a verb? how many NPs? <br /></li></ul><li>better documentation of our features </li></ul><li>Gold standard</li><ul><li>how do we compare to the Penn treebank data directly <br /></li><ul><li>how would that data have to be "incrementalized"</li></ul><li>difference between inproPitch and GoldPitch</li></ul><li>infrastructure</li><ul><li>Verdengelung of data independent of turns (thus we can still change the notion of turn)<br /></li><li>identification via dialogID/wordID for syntax and dialogID/channel/time for acoustics</li><li>master table has mapping between the two<br /></li></ul></ul></ul>
Public (everyone has full access) Moderated (other people can read, but not edit) Private (other people can not read or edit) You must first log in to change permissions.
Home Page