Internally Hive works on a schema which is present in the HDFS, Create a location in HDFS and then move all your data into that HDFS location.
Now create a Hive table which will read/point to that location.
If it is any commercial distribution like Cloudera/HortonWorks then Apache Hue is your saviour.
If it is not a commercial distro please use scp protocol to create a hdfs file and then use the Hive syntax mentioned below:
hive> CREATE TABLE `remote_file`(
> `Temperature` int,
> `Year` string)
> ROW FORMAT DELIMITED
> FIELDS TERMINATED BY ','
> LINES TERMINATED BY '\n'
> LOCATION
> 'hdfs:///tmp/remote_file';