Pular para o conteúdo principal

WD support for trust - feed back Pitts

 Thanks so much for sharing this, Daniel.  I found the discussion of how more rigorous use of qualifiers could aid trust very interesting. I also thought the Israel/Palestine and Crimea examples of disputed cases were illustrative and clarifying, as was the table displaying the frequency of disputed subjects (interesting it was so common for the creation of fictional characters). 

In broad strokes, we are using wikidata to collect information about ethnic  groups (and to a lesser extent, political parties and armed groups) in Sub-Saharan Africa.  As we were familiarizing ourselves on the way group data is stored on wikidata, we noticed some challenges with the different entries for Irish Republican Army with respect to start date and end date (and in theory, members of IRA splinter groups which are still active or have been active since 2005 might reasonably challenge any of these end dates).  The q nodes which are relevant for this include:  Q504628, Q175284, Q3015384, Q3084968.  None of these start date/end date qualifiers are necessarily wrong, but they do fail to help the researcher understand the continuity between these identities and a user without background information might find one of these pages and arrive at an incomplete time frame.  This becomes tricky if one hopes to use WD for an area in which one lacks contextual understanding!

"Disputed" qualifiers for ethnic groups may be of particular interest as ethnic group origin (or interrelationship between groups) can often become contested, especially in the context of territorial claims or irredentism-- but many ethnic groups may lack community members with the resources or inclination to initiate WD disputes. 

Here is a sample of some of the Qnodes of interest to us, a list of ethnic groups in Nigeria:  'Q190168', 'Q244157', 'Q572665', 'Q528981', 'Q7079875', 'Q1235705', 'Q239577', 'Q33454', 'Q1287326'..  We basically arrive at our Qnodes of interest by identifying proper nouns which appear in US State Department human rights reports and finding the corresponding Qnode.   For instance, I arrived at the Qnodes listed in this paragraph through the following python script.  The list of search terms comes from the State dept human rights reports. 

import requests
import json

search_terms = ["Yoruba people", "Igbo people", "Kanuri people", "Ijaw people", "Ogbe tribe", "Tiv people", "Ibibio people", "Fulani", "Edo people"]
group_ids = []

for search_term in search_terms:
    response = requests.get("https://www.wikidata.org/w/api.php", params={
        "action": "wbsearchentities",
        "format": "json",
        "language": "en",
        "type": "item",
        "limit": "1",
        "search": search_term
    })
    data = response.json()
   
    if len(data["search"]) > 0:
        group_id = data["search"][0]["id"]
        group_ids.append(group_id)
    else:
        print(f"No results found for search term: {search_term}")

The fields on those pages which are of interest include:  alsoKnownAs, instanceOf, nativeLanguage, country, locatedIn, religionOrWorldview, hasParts. 
  
I haven't arrived at a way to identify the useful Pnodes for our project yet, but if you have any suggestions, I am all ears. 

===========================================================================

Sobre os QNodes do Ira

No QNode Irish Republican Army (Q175284) temos a indicação de uma violação de constraint UNIQUE em relação ao QNode Irish Republican Army (Q3084968). Neste último também temos a violação em relação a constraint INVERSE ao relação ao QNode Category:Irish Republican Army (Q8554003).




Entre os QNodes Irish Republican Army (Q3015384) e Irish Republican Army (Q3084968) existe uma relação INVERSA representadas pelos PNodes replaced by (P1366) e replaces (P1365). Estes PNodes possuem a constraint INVERSE definida e a relação traz a semântica de ordem cronológica (contexto temporal).



O QNode Provisional Irish Republican Army (Q504628) possui um statement com a data de fundação (inception (P571) ) mas não com a dissolução. 


Enquanto o QNode Irish Republican Army (Q3084968) possui tanto a data de fundação quanto a de dissolução (dissolved, abolished or demolished date (P576))



Já o QNode Irish Republican Army (Q3015384) possui a data de fundação e a data de fim (utilizando outro PNode - end time (P582)

O QNode Provisional Irish Republican Army (Q504628) é o único que o contexto temporal está associado aos qualificadores data de início e data de fim para diferenciar os valores múltiplos para os statements Instance of (P31)

Selecionei com o kgtk os nós (sujeito e objeto) que envolvem estes QNodes

kgtk filter -i /app/kgtk/data/wikidata/claims.tsv.gz -p ' ; ; Q504628, Q175284, Q3015384, Q3084968 ' -o /app/kgtk/data/IRA-object.tsv

kgtk filter -i /app/kgtk/data/wikidata/claims.tsv.gz -p ' Q504628, Q175284, Q3015384, Q3084968 ; ; ' -o /app/kgtk/data/IRA-subject.tsv

E também os QNodes dos grupos étcnicos

kgtk filter -i /app/kgtk/data/wikidata/claims.tsv.gz -p ' ; ; Q190168, Q244157, Q572665, Q528981, Q7079875, Q1235705, Q239577, Q33454, Q1287326 ' -o /app/kgtk/data/Nigeria-object.tsv

kgtk filter -i /app/kgtk/data/wikidata/claims.tsv.gz -p ' Q190168, Q244157, Q572665, Q528981, Q7079875, Q1235705, Q239577, Q33454, Q1287326 ; ; ' -o /app/kgtk/data/Nigeria-subject.tsv

Depois para ambos os datasets eu adicionei os labels do node1, node2 e do PNode e ao final gerei o html de visualização com o kgtk. Mas o resultado da visualização não ficou bom. 

kgtk cat -i /app/kgtk/data/Nigeria-object.tsv /app/kgtk/data/Nigeria-subject.tsv \
     -o /app/kgtk/data/Nigeria-all.tsv
kgtk join --left-join \
     --left-file /app/kgtk/data/Nigeria-all.tsv \
     --right-file /app/kgtk/data/wikidata/labels.en.tsv.gz \
     -o /app/kgtk/data/Nigeria-label-node1.tsv
kgtk join --left-join \
     --left-file /app/kgtk/data/Nigeria-label-node1.tsv \
     --right-file /app/kgtk/data/wikidata/labels.en.tsv.gz \
     --left-file-join-columns node2 --right-file-join-columns node1 \
     -o /app/kgtk/data/Nigeria-label-node1-node2.tsv
kgtk join --left-join \
     --left-file /app/kgtk/data/Nigeria-label-node1-node2.tsv \
     --right-file /app/kgtk/data/wikidata/labels.en.tsv.gz \
     --left-file-join-columns label --right-file-join-columns node1 \
     -o /app/kgtk/data/Nigeria-all-labels.tsv        
kgtk lift -i /app/kgtk/data/Nigeria-all-labels.tsv -o /app/kgtk/data/Nigeria-final.tsv
               
kgtk visualize-graph -i /app/kgtk/data/Nigeria-final.tsv --show-edge-label --show-text above -o /app/kgtk/data/Nigeria.html

kgtk cat -i /app/kgtk/data/IRA-object.tsv /app/kgtk/data/IRA-subject.tsv \
     -o /app/kgtk/data/IRA-all.tsv
kgtk join --left-join \
     --left-file /app/kgtk/data/IRA-all.tsv \
     --right-file /app/kgtk/data/wikidata/labels.en.tsv.gz \
     -o /app/kgtk/data/IRA-label-node1.tsv
kgtk join --left-join \
     --left-file /app/kgtk/data/IRA-label-node1.tsv \
     --right-file /app/kgtk/data/wikidata/labels.en.tsv.gz \
     --left-file-join-columns node2 --right-file-join-columns node1 \
     -o /app/kgtk/data/IRA-label-node1-node2.tsv
kgtk join --left-join \
     --left-file /app/kgtk/data/IRA-label-node1-node2.tsv \
     --right-file /app/kgtk/data/wikidata/labels.en.tsv.gz \
     --left-file-join-columns label --right-file-join-columns node1 \
     -o /app/kgtk/data/IRA-all-labels.tsv        
kgtk lift -i /app/kgtk/data/IRA-all-labels.tsv --show-text above -o /app/kgtk/data/IRA-final.tsv

kgtk visualize-graph -i /app/kgtk/data/IRA-final.tsv --show-edge-label --show-text above -o /app/kgtk/data/IRA.html

Carreguei o dataset do IRA no Neo4J, a visualização é interessante mas não consigo exportar como html ou uma imagem interativa. 

Não agreguei os qualificadores pq nestas opções de visualização não iram ser carregados.

Comentários

Postagens mais visitadas deste blog

Connected Papers: Uma abordagem alternativa para revisão da literatura

Durante um projeto de pesquisa podemos encontrar um artigo que nos identificamos em termos de problema de pesquisa e também de solução. Então surge a vontade de saber como essa área de pesquisa se desenvolveu até chegar a esse ponto ou quais desdobramentos ocorreram a partir dessa solução proposta para identificar o estado da arte nesse tema. Podemos seguir duas abordagens:  realizar uma revisão sistemática usando palavras chaves que melhor caracterizam o tema em bibliotecas digitais de referência para encontrar artigos relacionados ou realizar snowballing ancorado nesse artigo que identificamos previamente, explorando os artigos citados (backward) ou os artigos que o citam (forward)  Mas a ferramenta Connected Papers propõe uma abordagem alternativa para essa busca. O problema inicial é dado um artigo de interesse, precisamos encontrar outros artigos relacionados de "certa forma". Find different methods and approaches to the same subject Track down the state of the art rese...

Knowledge Graph Embedding with Triple Context - Leitura de Abstract

  Jun Shi, Huan Gao, Guilin Qi, and Zhangquan Zhou. 2017. Knowledge Graph Embedding with Triple Context. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management (CIKM '17). Association for Computing Machinery, New York, NY, USA, 2299–2302. https://doi.org/10.1145/3132847.3133119 ABSTRACT Knowledge graph embedding, which aims to represent entities and relations in vector spaces, has shown outstanding performance on a few knowledge graph completion tasks. Most existing methods are based on the assumption that a knowledge graph is a set of separate triples, ignoring rich graph features, i.e., structural information in the graph. In this paper, we take advantages of structures in knowledge graphs, especially local structures around a triple, which we refer to as triple context. We then propose a Triple-Context-based knowledge Embedding model (TCE). For each triple, two kinds of structure information are considered as its context in the graph; one is the out...

KnOD 2021

Beyond Facts: Online Discourse and Knowledge Graphs A preface to the proceedings of the 1st International Workshop on Knowledge Graphs for Online Discourse Analysis (KnOD 2021, co-located with TheWebConf’21) https://ceur-ws.org/Vol-2877/preface.pdf https://knod2021.wordpress.com/   ABSTRACT Expressing opinions and interacting with others on the Web has led to the production of an abundance of online discourse data, such as claims and viewpoints on controversial topics, their sources and contexts . This data constitutes a valuable source of insights for studies into misinformation spread, bias reinforcement, echo chambers or political agenda setting. While knowledge graphs promise to provide the key to a Web of structured information, they are mainly focused on facts without keeping track of the diversity, connection or temporal evolution of online discourse data. As opposed to facts, claims are inherently more complex. Their interpretation strongly depends on the context and a vari...