Skip to content

Latest commit

 

History

History

RhizomerEye

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 

BESDUI Results for RhizomerEye

This document reports the results for the BESDUI benchmark for the RhizomerEye tool, a semantic data exploration and visualization frontend build on top of the RhizomerAPI backend. RhizomerEye automatically generates, after analyzing the semantic dataset, some user interface components to explore it. First, they facilitate gaining an overview of the dataset structure through a network overview or wordmap. Then, zooming and filtering into details through a faceted view.

Summary

RhizomerEye Capability K (0.2s) P (1.1s) H (0.4s) Operator Count Time Task Efficiency
Task 1 100% 10 7 2 19 10.5 5.7
Task 2 100% 11 8 2 21 11.8 5.1
Task 3 100% 10 3 2 15 6.1 9.8
Task 4 100% 13 11 2 26 15.5 3.9
Task 5
Task 6 100% 25 22 2 49 30 2.0
Task 7 100% 7 3 1 11 5.1 11.8
Task 8
Task 9
Task 10 100% 15 4 1 20 7.8 7.7
Task 11
Task 12
SUM 700% 91 58 12 161 86.8 46.0
AVERAGE 58% 13.0 8.3 1.7 23.0 12.4 2.8

Results per Task

Task 1. Find products for a given set of features combined

"Look for products of type sheeny with product features stroboscopes and gadgeteers, and a productPropertyNumeric1 greater than 450".

Interaction Steps K P H
1. On the network overview, click on “sheeny” green bubble 1 1
2. Click "+" for facet “productFeature” 1 1
3. Click facet value “stroboscopes” 1 1
4. Type in input “Search facet values” “gad...” and then RETURN key 5 1 1
6. Expand facet "productPropertyNumeric1" clicking "+" 1 1 1
6. Set facet's slider left end to aprox. "450" 1 2
Total KLM Operations 10 7 2
Time 10.5

Task 2. Find products for a given set of alternative features

"List products of type sheeny with product features stroboscopes OR gadgeteers, and a productPropertyNumeric1 greater than 450".

Interaction Steps K P H
1. On the network overview, click on “sheeny” green bubble 1 1
2. Click "+" for facet “productFeature” 1 1
3. Click facet value “stroboscopes” 1 1
4. Type in input “Search facet values” “gad...” and then RETURN key 5 1 1
5. Switch facet operator from AND to OR mode 1 1 1
6. Expand facet "productPropertyNumeric1" clicking "+" 1 1
7. Set facet's slider left end to aprox. "450" 1 2
Total KLM Operations 11 8 2
Time 11.8

Task 3. Retrieve basic information about a specific product for display purposes

"Get details about product boozed".

Interaction Steps K P H
1. Type "boozed" in "Search all..." input and then RETURN key 8 1 1
2. Click "Product" in "Type" facet 1 1 1
3. Click result "boozed" from right column products list 1 1
Total KLM Operations 10 3 2
Time 6.1

Task 4. Find products having some specific features and not having one feature

"Look for products of type sheeny with product features stroboscopes but NOT gadgeteers, and productPropertyNumeric1 value greater than 300 and productPropertyNumeric3 smaller than 400".

Interaction Steps K P H
1. On the network overview, click on “sheeny” green bubble 1 1
2. Click "+" for facet “productFeature” 1 1
3. Click facet value “stroboscopes” 1 1
4. Type in input “Search facet values” “gad...” and then RETURN key 5 1 1
5. Click selected value “gadgeteers” marked in yellow to negate it 1 1 1
6. Expand facet "productPropertyNumeric1" clicking "+" 1 1
7. Set facet's slider left end to aprox. "300" 1 2
8. Expand facet "productPropertyNumeric3" clicking "+" 1 1
9. Set facet's slider right end to aprox. "400" 1 2
Total KLM Operations 13 11 2
Time 15.5

Task 5. Find products matching two different sets of features

"Look for products of type sheeny with product features stroboscopes and gadgeteers and a productPropertyNumeric1 value greater than 300 plus those of the same product type with product features stroboscopes and rotifers and a productPropertyNumeric2 greater than 400".

This task cannot be performed using RhizomerEye because it does not support modelling the union of different query patterns.

Task 6. Find products that are similar to a given product

"Look for products similar to boozed, with at least one shared feature, and a productPropertyNumeric1 value between 427 and 627 (100 more or less than its value for boozed, 527) and a productPropertyNumeric2 value between 545 and 945 (200 more or less than its value for boozed, 745)".

The user can look for a specific product. As a result, the corresponding faceted view for it is generated and a query where the restriction is that the product is the one selected. The faceted view can be then used to select all the product features of the selected product. The user can also see the values for the numeric properties and manually set the ranges in the corresponding facet sliders. Finally, the user can remove the restriction for the specific product while keeping the rest of restrictions to retrieve similar products.

Interaction Steps K P H
1. On the network overview, click on “Product” blue bubble 1 1
2. Expand facet "label" by clicking "+" 1 1
3. Type "booz" in facets's "Search facets value" input and RETURN key 6 1 1
4. Expand facet "productFeature" by clicking "+" 1 1 1
5. Select all 11 values for "productFeature" facet 11 11
6. Expand facets "productPropertyNumeric1" and "productPropertyNumeric2" 1 1
7. Notice that values for "boozed" are "527" and "745"
8. Click value "boozed" in facet "label" to negate it 1 1
9. Swith operator mode for facet "productFeature" from AND to OR 1 1
10. Set facet "productPropertyNumeric1" slider to range aprox. "427" and "627" 1 2
10. Set facet "productPropertyNumeric2" slider to range aprox. "545" and "945" 1 2
Total KLM Operations 25 22 2
Time 30

Task 7. Find products having a name that contains some text

"Search products whose name contains ales".

Interaction Steps K P H
1. On the network overview, click on “Product” blue bubble 1 1
2. Expand facet "label" by clicking "+" 1 1
3. Type "ales" in facets's "Search facets value" input 5 1 1
4. The results appear as autocomplete recommendations
Total KLM Operations 7 3 1
Time 5.1

Task 8. Retrieve in-depth information about a specific product including offers and reviews

"For the product waterskiing sharpness horseshoes list details for all its offers by Chinese vendors and still valid by 2008-05-28 plus details for all reviews for this product having either rating1 or rating2".

This task requires defining facet restrictions for different related classes and it is not yet possible to perform it using RhizomerEye. The development of the pivoting support is currently future work.

Task 9. Give me recent reviews in English for a specific product

"For the product waterskiing sharpness horseshoes list the 20 more recent reviews in English".

RhizomerEye does not currently provide a mechanism to filter literals by language.

Task 10. Get information about a reviewer

"Get all available information about the author of Review5481".

Interaction Steps K P H
1. On the network overview, click on “Review” blue bubble 1 1
2. Type "Review5481" in "Search all facet values" input and RETURN key 12 1 1
4. Click displayed "Review5481" for details 1 1
4. Click reviewer link "Reviewer268" 1 1
Total Operators 15 4 1
Time 7.8

Task 11. Get offers for a given product which fulfill specific requirements

"Look for the cheapest and still valid by 2008-06-15 offer for the product waterskiing sharpness horseshoes by a US vendor that is able to deliver within 3 days".

This task requires defining facet restrictions for different related classes and it is not yet possible to perform it using RhizomerEye. The development of the pivoting support is currently future work.

Task 12. Export the chosen offer into another information system which uses a different schema

"Save in the local computer the information about the vendor for Offer3499, this is half the task. To complete it, restrict the output to just label, homepage and country and map them to schema.org terms name, url and nationality".

Currently, Rhizomer does not provide a way to map data and export it using a different schema.