Readme File for Operating the Neo4J Database Constructed in the Paper: Evidence-Based Practice in Sustainable Travel Behavior Intervention: A Knowledge Graph-Based Systematic Review

0. Related paper:

The paper Evidence-Based Practice in Sustainable Travel Behavior Intervention is available here or here with open access.

Supplementary data is available here.

1. The address of the database:

The Neo4J database constructed in the paper: Evidence-Based Practice in Sustainable Travel Behavior Intervention: A Knowledge Graph-Based Systematic Review is available at http://47.100.202.56:7474/browser/ without authentication.

2. A glance of the structure of the database:

Entities (or nodes) and relations (or relationships) of the database, two main concepts of knowledge graph, are available by pressing the "Database" button.

The construction of the database is the same as Table 1 in the paper.

	Type of Entities	Name of Entity	Name of Relations	Examples of Entities
Main entity	Number of evidence	Intervention	/	E0020
Reliability of evidence	Evidence hierarchy	Category	Inter_Category	Random controlled trials, Semi-experiment, Experiment with no comparison
	Risk of bias	Bias	Inter_Bias	Random sequence generation bias, Allocation concealment bias
	Evidence assessment	GRADE_level	Inter_GRADE	High, Moderate, Low, Very low
Intervention practice	Intervention types	Measure	Inter_BelongTo	Goals and planning, Feedback and monitoring
	Intervention targets	Target	Inter_Target	Car proportion, Carbon emission
	Intervention outcomes	Outcome	Inter_Outcome	Change in household travel patterns; total carbon emissions decreased by 10%.
Design of intervention studies	Region	N/A (In the form of descriptions)	Inter_Region	London, Xuzhou, Ghent
	Kinds of participants	Participant	Inter_Participant	People, Households, Schools
	Scenario	Scenario	Inter_Scenario	Workplace interventions, School interventions, Unspecified interventions

3. Guide to search the evidence stored in the database:

Evidence stored in the database can be searched by entering CQL (Cypher Query Language) commands in the search box.

As the main and central entity is number of evidence, the searching strategy should be based on number of evidence ("E00**"). For example, if we are searching evidence number 20, we can enter the CQL command as "match r=(n:Intervention)-[]-() where n.ID="E0020" return *", where the key of number of evidence (or Intervention) is "ID".

And the search yields everything regarding evidence number 20, while some details are available by activating the relationship arrow.

4. Advanced search strategy:

The database enables advanced search strategy mainly for decision-making. For example, if decision supporters are searching for evidence. Searching criteria includes: (1) scenario is school intervention, (2) target is to lower car usage proportion, (3) decision supporters value only high or moderate level of evidence as reliable. In this case, the searching strategy can be expressed as "match (m:Intervention)-[]-(n:SubTarget), (m)-[]-(l:Scenario), (m)-[]-(p:GRADE_Level) where n.ID="Car Proportion" and l.ID="School Interventions" and (p.ID="high" or p.ID="moderate") with m as mm match (mm)-[]-(q:Outcome), (mm)-[]-(r:DetailedMeasure), (mm)-[]-(s:SubMeasure)-[]-(t:MajorMeasure), (mm)-[]-(u:Region) return *".

And the search yields two pieces of evidence that satisfy the requirements of decision supporters.

5. Instructions of CQL (Cypher Query Language) commands:

Basic knowledge regarding CQL commands can be found at https://neo4j.com/docs/cypher-manual/current/introduction/.