I once used the erlang-json-eep-parser, and tried it on your data.
7> json_eep:json_to_term("({ id1 : [\"str1\", \"str2\", \"str3\"], id2 : [\"str4\", \"str5\"]})").
** exception error: no match of right hand side value
{error,{1,json_lex2,{illegal,"("}},1}
in function json_eep:json_to_term/1
Right, it doesn't like the parentheses.
8> json_eep:json_to_term("{ id1 : [\"str1\", \"str2\", \"str3\"], id2 : [\"str4\", \"str5\"]}").
** exception error: no match of right hand side value
{error,{1,json_lex2,{illegal,"i"}},1}
in function json_eep:json_to_term/1
And it doesn't like the unquoted keys:
18> json_eep:json_to_term("{ \"id1\" : [\"str1\", \"str2\", \"str3\"], \"id2\" : [\"str4\", \"str5\"]}").
{[{<<"id1">>,[<<"str1">>,<<"str2">>,<<"str3">>]},
{<<"id2">>,[<<"str4">>,<<"str5">>]}]}
That looks better.
So it seems that your data is almost JSON, at least as far as this parser is concerned.