I suggest you ask a more specific question. For example, you could ask "I have 2 mysql database tables. One contains basic contact data for the people. The other table has the properties for the people such as 'smoker','loves dogs','college degree', etc. I want to build an alogirthm for matching people by the compatibility of their properties. I want to write the code using PHP, and I have basic PHP skills. I need help building the matching logic."

okay..well these are some my database tables :
-basic info table(name,age.....etc)
-appearance info table(height,hair color,...etc)
-interests info table(hiking,music,....etc)

i need help building the matching logic..

Matching people for compatibility is a very complex process. I imagine the algorithms that companies like eHarmony have developed took years of psychology and study as well as programming to develop.

I would imagine that you will have to develop a unique algorithm--a set of logic--for each factor. For example, let's look at "Age". Let's say you've determined that for men:

most compatible woman's age = man's age - (man's age / 10)

most compatible woman's age = 40 - (40/10) = 40 - 4 = 36

And let's say you determine to exclude women:

if woman's age < Round(man's age/1.5)
or if woman's age > Round(man's age * 1.2)

So, for a 40 year old man, you would exclude women if the age is less than 27 or greater than 48.

Then assume you assign a ranking of 5 points per year of age in the compatible range.....or whatever you decide.

Let's say you have 7 women candidates

Woman A: age 23
Woman B: age 27
Woman C: age 30
Woman D: age 36
Woman E: age 38
Woman F: age 48
Woman G: age 51

Women A & B would be excluded because they fall outside the range of 27 to 48.

Woman D would have an age compatibility ranking of 100 (most compatible).

Woman B would have a ranking of 100 - (5 * (36-27)) = 55
Woman C would have a ranking of 100 - (5 * (36-30)) = 70
Woman E would have a ranking of 100 - (5 * (38 - 36) = 80
Woman F would have a ranking of 100 - (5 * (48 - 36) = 40

Armed with this algorithm, you can loop through all the women and determine each one's age compatibility ranking.

Do this for every factor then add up all the factors for each woman to give you an "Overall Compatibility Ranking".

I'd imagine the process of tweaking and refining each factor's algorithm would be a never-ending process.

alogirthm for matching people
- It is necessary to execute search of conformity in the second table,
Including favourite properties, excepting unloved properties