Interests divorced from benefactor

[psychemedia]

Many of the interests listed are divorced from the party that provided the benfit example query

For example:

item	date	person_id	name
On 30 January 2013 I was appointed as a non-executive director of the Social Investment Business Group, Address 1st Floor, Derbyshire House, St Chad's Street, London WC1H 8AG.	2013-11-25	uk.org.publicwhip/person/10051	Directorships
27 February 2013, received £364.11. Hours: 12 hrs. (Registered 13 March 2013)	2013-11-25	uk.org.publicwhip/person/10051	Directorships
27 March 2013, received £333.34. Hours: 12 hours (estimated).  (Registered 3 June 2013)	2013-11-25	uk.org.publicwhip/person/10051	Directorships
26 April 2013 received £333.34. Hours: 12 hours (estimated).  (Registered 3 June 2013)
--	--	--	--

There are multiple payments, presumably from the first stated organisation (assuming we can trust the sort order to re-present the original grouped items in the correct sequence? Or is there a better sort strategy?)

I also note:

• there may be duplicate entries?
• it would be useful to pull out the (Registered DATE) element;
• entity extraction (eg spacy) could be used to extract companies, addresses, amounts, etc.