[Solved]1 Using Recall R Precision P Evaluate Search Engine Determine Training Data 100 Documents Q37201118
1. We are using recall R and precision P to evaluate our searchengine. We determine in the training data that 100 documents shouldmatch our query “Kimmer Informatics”. Our search engine returns 150documents, 60 of which are valid results for the query. How manyfalse positives does our search engine return? How many falsenegatives? What are the precision, recall, and F value in thiscase?
2. We have the following document:
It is a period of civil war. Rebel spaceships, striking from ahidden base, have won their first victory against the evil GalacticEmpire. During the battle, rebel spies managed to steal secretplans to the Empire’s
ultimate weapon, the DEATH STAR, an armored space station withenough power to destroy an entire planet. Pursued by theEmpire’s
sinister agents, Princess Leia races home aboard her starship,custodian of the stolen plans that can save her people andrestore
freedom to the galaxy….
For the query with terms Empire and plans, how many covers doesthis document have? What are the covers (indicated by circling orboxing them on a copy/paste of the text into your assignment)?Assume our stopword list is empty (i.e., we’re not using anystopwords).
Using the scoring from slide 9 of covers and phrases, calculatethe positional score of this document for that query. Show yourwork as much as possible.
3. For the document from part 2, assume we have a cosinesimilarity score with the query of 0.72 and that the document has apageRank of 0.15. We are weighting cosine similarity as 60% of thefinal score, pageRank as 15%, and the positional score as 25%. Whatis the score of the document?
4. For the network below, compute one round of PageRank with asmoothing value of 8/10 (i.e. 0.8). Use the initial PageRank valuesgiven rather than initializing each node to the same PageRank.
1/4 1/8 1/8 108 1/8 1/4 Show transcribed image text 1/4 1/8 1/8 108 1/8 1/4
Expert Answer
Answer to 1. We are using recall R and precision P to evaluate our search engine. We determine in the training data that 100 docum… . . .
OR
