Of course, I'm talking about romantic love, not one's love for one's children or friends or non-spouse family members. True love exists, though I'd guess it's uncommon. It's not necessary for a good marriage or a happy life. There can be good marriages in which the spouses love one another but one (or both) doesn't "truly love" the other. There are lives which lack true love but are much better than some lives that have it. It can cause trouble in a life, too. So, it's not a cardinal purpose of life. But I digress. The point is to define it.
Here is an analytical definition of true love:
"S truly loves R" =df "S admires many of R's characteristics in the categories of the psychological and moral (character traits, talents, dispositions) and the physical (appearance and behaviors), and including the large and important characteristics and also the small and unimportant ones. Also, S deeply admires some traits of R from each category. Finally, S also has romantic feelings for R (desires to show R affection physically, desires to couple with R, desires R sexually, and so forth.)
To refute an analytical definition, think of a counterexample. I don't know what would count as a case of true love in which S wasn't described by the definiens I've given. Nor can I imagine evidence that would suggest that although S could be described by the definiens, S did not truly love R. That is the argument for this definition. But it could collapse under the weight of a good counterexample.