Page 1 of 1

Who owns the data output by a program?

Posted: Mon Feb 21, 2011 4:47 pm
by rich.d.berry
I'm wanting to re-use data that is output by another copyright holder's program as input for my program.

Is there any possibility that I will be infringing copyright law by doing so?

The owner of the software lays no specific claim to the data in the terms and conditions.

The data is produced by calculating values based on measurements and counts gathered and input by the software user and all the input data is reproduced in the output data file, along with the results of some calculations.

My concern stems from the following definition:

Literary works include computer programs and databases. Work must be original, though the interpretation is very wide, for example railway timetables are regarded as literary works under the Act. A database is a compilation in any form, print or non-print, but its selection or arrangement must be the author’s own intellectual creation and it is protected for 15 years from the latest substantial revision.

The data is in XML format, and XML can be used to represent a database. I personally would not class the data as a database because there is no row and column type structure typical of a database, but it could be argued that the data represents a single entry (row) within a database table as its arrangement is rigidly defined. But most XML data would qualify using this logic.

I am hesitant to approach the program copyright holders directly, because being a government office, I've discovered that the default position is always no. If I ask the question and alert them to my intentions, I would make myself an easy target.

I intend to offer an upload facility on my website to save users who have already input their data from having to do so again. I will use the uploaded data to perform additional calculations and forecasts. Users can currently enter the data manually.

I am not in competition with the government office. They use the data to produce a certificate (and the XML file in question). I use much of the same data to allow people to maximise their savings by manipulating various "what if" scenarios.

Thanks in anticipation.

PS. The users of the software are licensed agents. These agents produce the certificates on behalf of members of the public. Once a certificate has been produced then anyone in possession of the certificate number can go online to the government agency and download the XML file without using the licensed software. Does this change anything?

Posted: Mon Feb 21, 2011 7:44 pm
by AndyJ
As you are probably aware, the original 1988 Act on Copyright could not be expected to foresee the enormous changes in technology which have occurred in the last 20 years. Extending the law to include databases happened in 1997 and the law continues to evolve, mainly as a result of court cases. Technically speaking, copyright and database right are two separate things, although apart from their respective durations, the rights are pretty much the same.
However the main principles behind copyright remain largely the same. In this case, to be eligible for copyright, something has to be substantially the product of the human brain. Thus the code which makes up a computer program is written by humans so is copyrightable, and in certain circumstances the output from a program may also be considered to be largely the product of a human brain. For instance a graphics program like Photoshop or a wordprocessor produce little without the human input. Hence their output would virtually always be subject to copyright. But data which is merely the result of a pre-determined set of algorithms applied to a set of input data would not normally pass this test. And the fact that the data could be output into the format of a database (or a spreadsheet) again does not necessarily mean the output is subject to the database right.

A recent test case may help to show what is meant by the "selection and arrangement" of data in a database. In Football Dataco Ltd v Brittens Pools Ltd [2010] EWHC 841 Ch Mr Justice Floyd proposed the following test
It seems to me that the task for the court is as follows:

i) Identify the data which is collected and arranged in the database;

ii) Analyse the work which goes into the creation of the database by collecting and arranging the data so identified, to isolate that work which is properly regarded as selection and arrangement;

iii) Ask whether the work of selection and arrangement was the author's own intellectual creation and in particular whether it involved the author's judgment, taste or discretion;

iv) Finally one should ask whether the work is quantitatively sufficient to attract copyright protection.*
This was in respect of the compilation of football match fixtures, which involved both computer sorting and human 'tweaking' to arrive at the desired result. In this case the fixtures tables had sufficient human involvement in the selection and arrangement of the data to qualify for copyright. Clearly the mere output from a computer program would entirely fail that test and so would not be subject to copyright.

In an interesting parallel to your project, the designer of an app which used the output from the NationalRail train arrival and departures database was denied a licence to use their data, not for reason of copyright but because of the app's alleged commercial nature. More details here: ... times-app/ On that precedent, I suspect that if the government department concerned objected to your use of their data, they would simply revoke your certificate to access the data, which their terms and conditions may define as misuse, rather than resort to a claim of copyright infringement which they probably couldn't win..

* Source: Para 91


Posted: Mon Feb 21, 2011 11:13 pm
by rich.d.berry
Thanks Andy, that has clarified the entire issue for me. I appreciate the trouble you went to and the examples you gave were very helpful.