Optimize Time and Space

IMDB-style services aggregate millions of user ratings into per-title averages for display and ranking. The Go code below produces correct numbers but does not scale to a large catalog of titles and ratings.

package main

import "fmt"
import "strconv"

type Movie struct {
    ID    string
    Title string
}

type Rating struct {
    MovieID string
    Score   float64
}

func buildReport(movies []Movie, ratings []Rating) string {
    report := ""
    for _, m := range movies {
        var total float64
        var count int
        for _, r := range ratings {
            if r.MovieID == m.ID {
                total += r.Score
                count++
            }
        }
        var avg float64
        if count > 0 {
            avg = total / float64(count)
        }
        report = report + m.Title + ": " + strconv.FormatFloat(avg, 'f', 2, 64) + "
"
    }
    return report
}

func topRated(movies []Movie, ratings []Rating, minCount int) []string {
    var result []string
    for _, m := range movies {
        count := 0
        for _, r := range ratings {
            if r.MovieID == m.ID {
                count++
            }
        }
        if count >= minCount {
            result = append(result, m.Title)
        }
    }
    return result
}

func main() {
    movies := []Movie{{"tt001", "Arrival Point"}, {"tt002", "Silent Harbor"}}
    ratings := []Rating{{"tt001", 8.0}, {"tt001", 9.0}, {"tt002", 7.0}}
    fmt.Println(buildReport(movies, ratings))
    fmt.Println(topRated(movies, ratings, 2))
}

Explain the time and space complexity of buildReport and topRated as written, and how you would bring them down for a catalog with millions of titles and ratings.