现代信息检索导论作业讲评优质PPT.ppt

1、How to do linear merge of postings with g（d）?Write a C function to present your idea.忽略了did未保持g（d）的顺序第三次作业第二题第四次作业第二题在10000篇文档构成的文档集中，某个查询的相关文档总数为8，下面给出了某系统针对前20个有序结果的相关（R）和不相关（N）情况：RRNNN NNNRN RNNNR NNNNRA.前20篇文档的正确率：P=6/20=30%B.前20篇文档的F1值：F1=2PR/（R+P）其中R=6/8，故F1=0.4286第四次作业第二题RRNNN NNNRN RNNNR NNN

2、NRC.在25%召回率水平上的插值正确率：100%D.在33%召回率水平上的插值正确率：36.4%0.1250.250.3750.50.6250000000000010.7500000000000011009080706050403020100第四次作业第二题RRNNN NNNRN RNNNR NNNNRE.假定该系统所有返回结果的数目就是20，则MAP=（1+2/2+3/9+4/11+5/15+6/20+0+0）/8 =0.4163第四次作业第二题RRNNN NNNRN RNNNR NNNNRF.该系统可能的最大MAP:当第21和22篇文档都是相关文档时，MAP达到最大值。MAP=（1+2/

3、2+3/9+4/11+5/15+6/20+7/21+8/22）/8 =0.5034G.该系统可能的最小MAP:当第9999和10000篇文档是相关文档时，MAP达到最小值。MAP=（1+2/2+3/9+4/11+5/15+6/20+7/9999+8/10000）/8 =0.4165第四次作业第二题RRNNN NNNRN RNNNR NNNNRH.在一系列实验中，只有最靠前的20篇文档通过人工来判定，（E）的结果用于近似从（F）到（G）的MAP取值范围。对于上例来说，通过（E）而不是（F）和（G）来计算MAP所造成的误差有多大（采用绝对值来计算）？|MAP F-MAP G|=0.0869第四次作

4、业第三题Write a C program to highlight the keywords of an input query in the text of an input document,where both the query and document text are input as a character string:const char*q=“word1 word2 word3”;const char*doc_text=“”;（Requirements:first segment the text to sentences,then select them.）要求用C语言

5、首先分句Highlight整个查询出现的地方，而不是查询中某个单词程序应该生成一个HTML文件第五次作业第二题Give three reasons why relevance feedback has been little used in web search.用户不愿意进行显示反馈（延长搜索交互时间）相关反馈会造成长查询，降低系统效率相关反馈主要用于提高召回率，而WEB检索中准确率能提升用户体验很难使普通用户理解并使用第五次作业第三题Why is positive feedback likely to be more useful than negative feedback to an

6、IR system?正反馈返回的相关文档中相似度更高，聚类性质强，容易带来更多的相关文档Why might only using one nonrelevant document be more effective than using several?在实际检索中绝大部分文档都是不相关文档，相关文档的聚类不够强，容易相互抵消第五次作业第四题Omar has implemented a relevance feedback web search system,where he is going to do relevance feedback based only on words in th

7、e title text returned for a page（for efficiency）.The user is going to rank 3 results.The first user,Jinxing,queries for:banana slugand the top three titles returned are:banana slug Ariolimax columbianusSanta Cruz mountains banana slugSanta Cruz Campus MascotJinxing judges the first two documents rel

8、evant,and the third nonrelevant.Assume that Omars search engine uses term frequency but no length normalization nor IDF.Assume that he is using the Rocchio relevance feedback mechanism,with =1.Showthe final revised query that would be run.（Please list the vector elements in alphabetical order.）第五次作业

9、第四题Query:banana slugDocuments:（R）banana slug Ariolimax columbianus（R）Santa Cruz mountains banana slug（N）Santa Cruz Campus MascotAriolimax banana Campus columbianusCruz Mascotmountains Santa slugQ010000001D1110100001D2010010111D3001011010第五次作业第四题把文档写成向量Q =（0,1,0,0,0,0,0,0,1）D1=（1,1,0,1,0,0,0,0,1）D2=（0,1,0,0,1,0,1,1,1）D3=（0,0,1,0,1,1,0,1,0）由公式，其中 =1得Qm=（0.5,2,-1,0.5,-0.5,-1,0.5,-0.5,2）负的weight变为0Qm=（0.5,2,0,0.5,0,0,0.5,0,2）