Re: [foaf-dev] FOAF Relationship

Hi Seth,

I found something in the Apache Stanbol project [1] that could be useful to your problem.
There is a closed ticket (with source code attached) about Entity Disambiguation using FOAF Correlation [2].

Cheers,
— 
Andrea Giovanni Nuzzolese, Ph.D.
Semantic Technology Laboratory (STLab)
Institute for Cognitive Science and Technology (ISTC)
National Research Council (CNR)
Via Nomentana 56, Roma - Italy


[1] https://stanbol.apache.org/ <https://stanbol.apache.org/>
[2] https://issues.apache.org/jira/browse/STANBOL-1161 <https://issues.apache.org/jira/browse/STANBOL-1161>

> On 18 Apr 2016, at 23:01, Seth Grimes <grimes@altaplana.com> wrote:
> 
> Thanks. I'm very aware of NLP capabilities. There really aren't many that will do familial relationships out of the box -- IBM Watson, Open Calais. In principle, a tool that does deep syntactic parsing could be adapted for this sort of use, because syntactic relationships indicate semantic ones, but I'm looking for built-in capabilities
> 
> But I'm sure there are quite a few NLP tools I don't know. I decided to come at the question from a different direction. Since the FOAF Relationship scheme can capture familial relationships (among others), I was hoping someone who works with it might have ideas about tools that could automate extraction that would use it.
> 
>     Seth
> 
> 
> On Mon, 18 Apr 2016, Alvaro Graves wrote:
> 
>> Hi Seth,
>> 
>> Although I'd love to hear of a simpler process, I think you may need to
>> look at NLP software, such as CoreNLP (see http://corenlp.run for a demo),
>> and then programmatically link the entities detected.
>> 
>> Hope it helps
>> 
>> Alvaro Graves-Fuenzalida, PhD
>> Web: http://graves.cl - Twitter: @alvarograves
>> 
>> On Mon, Apr 18, 2016 at 12:08 PM, Seth Grimes <grimes@altaplana.com> wrote:
>> 
>>> Do you have any sample document URLs? That would help anyone seeking to
>>>> respond to your request that produces these kinds of solutions.
>>>> 
>>> 
>>> Here's a text sample. There's actually quite a lot of salient information
>>> to extract: Persons, inter-personal relationships, locations, events
>>> (including sequence). But I'm not looking for comprehensive narrative
>>> mining. For now, I'm looking only for limited relation extraction and
>>> representation, to support queries across the relationship graph.
>>> 
>>>        Sandy was removed from her home by CPS and placed in foster care
>>> after her class teacher observed multiple bruises on Sandy's body. It is
>>> alleged that mother's boyfriend inflicted the injuries. Mother claims she
>>> was unaware of the injuries. Mother has a history of reported drug and
>>> alcohol use and was intoxicated at the time of CPS investigation. Sandy is
>>> currently placed in the agency-operated foster boarding home of Ms. Jones.
>>> Child's mother is allowed to have supervised visits at the foster care
>>> agency until the next planning conference.  Caseworker visited the foster
>>> boarding home of Ms. Jones today to assess Sandy's adjustment to placement
>>> with the Jones' family.  Present in Ms. Jones's home were Ms. Jones, her
>>> two daughters (Erika and Mona), and her foster child, Sandy. Ms. Jones said
>>> her husband was away on business as an insurance evaluator.  The home was
>>> messy with clothes strewn all over the living room floor. The sink was
>>> piled high with dirty dishes and two roaches were observed crawling on the
>>> wall. There were an opened box of cereal, bread, and peanut butter on the
>>> dining table. The apartment is sparsely furnished with a dining table and
>>> chairs, a sofa that is fabric upholstered and heavily soiled, and a
>>> television set. The sleeping arrangement is adequate but there are no
>>> closets or chests of drawers to store the children's clothing. Mary has her
>>> own bedroom and Johnny and Jonathan share a bedroom with two small beds.
>>> Ms. Brown has her own bedroom. The home has safety devices installed
>>> (window guards, fire and carbon monoxide alarms).  Ms. White, Sandy's
>>> mother came to the office today for a supervised visit. Mother looked
>>> somewhat disheveled; her hair was uncombed, her clothes were heavily soiled
>>> and she smelled of alcohol. Caseworker asked mother if she had been
>>> drinking. Mother said she had just had a beer. Sandy took a while to warm
>>> up to her mother. Mother brought a packet of potato chips for Sandy. Sandy
>>> took the chips after much prodding from Ms. Jones but returned to snuggle
>>> up to Ms. Jones.  Sandy disclosed to caseworker that she was fearful of
>>> returning home in case mother's boyfriend hurt her again. Sandy said
>>> mother's boyfriend often hit her for no reason.
>>> 
>>> Thanks,
>>> 
>>>                                        Seth
>>> 
>>> 
>>> 
>>> On Mon, 18 Apr 2016, Kingsley Idehen wrote:
>>> 
>>> On 4/18/16 2:47 PM, Seth Grimes wrote:
>>>> 
>>>>> Yes, that's pretty much it, thanks. Anaphora (pronoun) resolution
>>>>> would be a big plus.
>>>>> 
>>>>>                    Seth
>>>>> 
>>>> 
>>>> Do you have any sample document URLs? That would help anyone seeking to
>>>> respond to your request that produces these kinds of solutions.
>>>> 
>>>> 
>>>> Kingsley
>>>> 
>>>>> 
>>>>> 
>>>>> On Mon, 18 Apr 2016, Kingsley Idehen wrote:
>>>>> 
>>>>> On 4/18/16 9:44 AM, Seth Grimes wrote:
>>>>>> 
>>>>>>> Hello,
>>>>>>> 
>>>>>>>    Are there any available information-extraction systems that
>>>>>>> implement the FOAF Relationship vocabulary
>>>>>>> (http://vocab.org/relationship/)? By available, I mean commercial or
>>>>>>> open source and currently maintained. My particular interest at this
>>>>>>> moment is identification and extraction of familial relationships from
>>>>>>> documents, and preferably also of attributes associated with the
>>>>>>> relationships, and representation of extracted relationships.
>>>>>>> 
>>>>>>>    Thanks,
>>>>>>> 
>>>>>>>                    Seth
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> --
>>>>>>> Seth Grimes    grimes@altaplana.com   +1 301-270-0795    @sethgrimes
>>>>>>> Alta Plana Corp, analytics strategy consulting, http://altaplana.com
>>>>>>> Sentiment Analysis Symposium, July 12 in NYC, SentimentSymposium.com
>>>>>>> 
>>>>>> 
>>>>>> Hi Seth,
>>>>>> 
>>>>>> Are you looking for technology that would perform the following tasks?
>>>>>> 
>>>>>> [1] analyze a document containing a collection of English sentences
>>>>>> representing familial relationships
>>>>>> [2] extract said relationships and then convert to RDF sentences using
>>>>>> terms from the FOAF vocabulary
>>>>>> [3] return an RDF document and the final output artifact.
>>>>>> 
>>>>>> I've copied the LOD list in on this response to broaden audience for
>>>>>> this exchange.
>>>>>> 
>>>>>> 
>>>>>> 
>>>>> --
>>>>> Seth Grimes    grimes@altaplana.com   +1 301-270-0795    @sethgrimes
>>>>> Alta Plana Corp, analytics strategy consulting, http://altaplana.com
>>>>> Sentiment Analysis Symposium, July 12 in NYC, SentimentSymposium.com
>>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>> --
>>> Seth Grimes    grimes@altaplana.com   +1 301-270-0795    @sethgrimes
>>> Alta Plana Corp, analytics strategy consulting, http://altaplana.com
>>> Sentiment Analysis Symposium, July 12 in NYC, SentimentSymposium.com
>>> 
>>> 
>> 
> 
> -- 
> Seth Grimes    grimes@altaplana.com   +1 301-270-0795    @sethgrimes
> Alta Plana Corp, analytics strategy consulting, http://altaplana.com
> Sentiment Analysis Symposium, July 12 in NYC, SentimentSymposium.com
>