=Paper=
{{Paper
|id=Vol-1223/paper1
|storemode=property
|title=PurposeNet: A Knowledgebase Organized Around Purpose
|pdfUrl=https://ceur-ws.org/Vol-1223/paper1.pdf
|volume=Vol-1223
}}
==PurposeNet: A Knowledgebase Organized Around Purpose==
https://ceur-ws.org/Vol-1223/paper1.pdf
PurposeNet: A Knowledgebase Organized Around
Purpose
Rajeev Sangal, Soma Paul, P. Kiran Mayee
Language Technologies Research Centre
International Institute of Information Technology
Hyderabad, India
sangal@iiit.ac.in, soma@iiit.ac.in,
kiranmayee@research.iiit.ac.in
Abstract. We show how purpose can be used as a central guiding principle for
organizing knowledge about artifacts. It allows the actions in which the artifact
participates to be related naturally to other objects. Similarly, the structure or
parts of the artifact can also be related to the actions.
A knowledgebase called PurposeNet has been built using these principles. A
comparison with other knowledebases shows that it is a superior method in
terms of coverage. It also makes it possible for automatic extraction of simple
facts (or information) from text for populating a richly structured knowledge-
base.
An experiment in domain-specific question-answering from a given passage
shows that PurposeNet used alongwith scripts (or knowledge of stereotypical
situations), can lead to substantially higher accuracy in question answering. In
the domain of car racing, individually they produce correct answers to 50% and
37.5% questions respectively, but together they produce 89% correct answers.
Keywords: Ontology, Semantic Knowlegdebase, Information Extraction,
OWL, Question-Answering
1 Introduction
There is a need to represent knowledge for a variety of applications, ranging from
natural language processing to reasoning in sciences, education, business, social sci-
ence and humanities. This requires Knowledge Representation (KR) schemes, as well
as good ways of organizing knowledge.
KR schemes and inference methods have received a great deal of attention. This
has resulted in several effective schemes which are strong as well as have efficient
and powerful inference methods. Notable among them have been Sowa (1984),
(2002), (2005) and Bharati et. al (1987), (1991), (1995).
Besides the KR schemes, there is also a need to work out the organization of
knowledge. The question naturally arises as to what principles should be used to or-
ganize knowledge, namely, what knowledge should be put in, and how would parts of
2 PurposeNet: A Knowledgebase Organized Around Purpose
that knowledge relate with other parts of knowledge? For example, if the domain of
transport needs to be described, how should the different elements starting from car
and trucks and going to repairs and roads, be organized?
The answer lies in recognizing that there are principles underlying the organiza-
tion. Once these are understood, it becomes easier to relate different parts of
knowledge with each other. Such knowledge can then be represented in a suitable KR
scheme.
We have used purpose as an organizing principle in our work. This principle has
been applied primarily to artifacts or manmade objects. It has been developed and
used extensively in the Indian philosophical tradition. Objects are described in terms
of four major types of attributes: rup, gun, svabhav, dharm.
Dharm is that property which is intrinsic (essential) to the objects in the category,
and helps distinguish the category from other categories. Dharm is given by its pur-
pose. For example, for a car, its dharm or purpose would be to transport (a small
number of) people from one place to another on land.
Svabhaav refers to those attributes which the object shares with objects of the same
class and which it does not share with other classes. For example, Car shares attrib-
utes with other machines, but does not share attributes with living beings.
Rup (literally meaning, form) refers to those attributes which can directly be per-
ceived by our sensory organs. For example, rupa of car would be its shape, colour,
weight, etc. Gun refers to properties that are not perceived directly but indirectly such
as load carrying capacity, etc. dharm and gun are performative, where assvabhav and
rup are non-performative (though they are essential for performance).
While building PurposeNet, a knowledgebase, we have used purpose as the prima-
ry principle of organizing knowledge. We note that the dictionary uses the same idea
to give meanings of words. Let us take some examples from popular resources such as
WordNet (Miller et. al, 1990), Wikipedia (Wikipedia, 2004) and Cambridge diction-
ary (http://dictionary.cambridge.org/ dictionary/american-english/).
WordNet defines the artifacts “fork”, “bomb” and “knife” in the following manner:
1. Fork - cutlery used for serving and eating food.
2. Bomb - an explosive device fused to explode under specific conditions.
3. Knife - edge tool used as a cutting instrument; has a pointed blade with a sharp
edge and a handle.
In Wikipedia articles on artifacts, the first sentence generally describes the artifact as
exemplified below:
1. Chair – A chair is a raised surface, commonly for use by one person.
Wall – A wall (from Old English weall) is a vertical structure, usually solid, that de-
fines and sometimes protects an area.
2. Football - A football is an inflated ball used to play one of the various sports
known as football.
Cambridge dictionary has the following entries:
Rajeev Sangal, Soma Paul, P. Kiran Mayee 3
1. Telephone – A device for speaking to someone in another place by means of elec-
trical signals
Brush – Any of various utensils consisting of hairs or fibers arranged in rows or
grouped together, attached to a handle, and used for smoothing the hair, cleaning
things, painting, etc.
Rack – A frame, often with bars or hooks, for holding or hanging things.
Open_Pen_Cap Upturn_Pen Press_Pen_Nib_On_Paper Move_Pen _On_Paper
subaction
precondition
Write_On_Paper semantic roles doer Human
outcome purpose instrument topic
location
Some_Topic
birth actions Pen Paper
Nib
Assemble_Pen_Parts
parts Cap
destruction
Barrel
Fig.1.Illustration of importance of Purpose as a basis for knowledge representation.
All the nine entries cited above are defined in the form “an X”, where, X ={Fork, Bomb, Knife, Chair, Wall, Football, Telephone, Brush, Rack},
Y = {cutlery, explosive device, edge tool, raised surface, vertical structure, ball, de-
vice, utensil, frame}, Z={to sit on, that protects an area, to play, for throwing, for
holding, …}. Thus, purpose is very significant information about artifacts. An artifact
is made in order to serve one particular purpose. The various characteristics and activ-
ities associated with an artifact depend upon the purpose for which it is created.
As one would have noticed, the purpose of an object is given in terms of an action
that the object helps accomplish. The object also has a structure, i.e., is made up of
parts which are put together in well-defined way. The structure is related to the pur-
pose of the object, namely, the structure helps accomplish the purpose.
4 PurposeNet: A Knowledgebase Organized Around Purpose
In the case of a pen, for example, the purpose is to write on paper. The Pen has a
thin and cylindrical shape for a comfortable gripping while used for writing. It has
many sub-parts, such as Barrel, Nib, Feed and Cap, which together help carry out the
action of putting marks on paper. The action can be broken into sub-actions which
relate to the parts, where each part helps in carrying out some sub-action(s). Barrel
holds Ink, Nib allows Ink to pass through and Cap prevents the Ink from drying.
Therefore, when Ink-Pen is made, it is an assemblage of the aforementioned compo-
nents and we know why the components are in the way stated. Each of them helps in
fulfilling the central purpose of Pen, which is, writing.
INSTANCE
NAME ALIAS DESCRIPTION PROPERTIES PROPERTY-VALUE PAIRS
NACCESSORY
CORE COMPONENT
SUBTYPE
DESCRIPTOR FEATURES
ARTIFACT
FEATURES
ACTION FEATURES
PURPOSE
MAKE / BIRTH LIFE DESTRUCTION
ACTION ONTOLOGY
NAME PRECONDITIONS OUTCOME SUBACTIONS THETA_ROLES
RESULTS SIDEEFFECTS WEAR & TEAR
MAINTENANCE
Fig. 2. Architecture of PurposeNet
Rajeev Sangal, Soma Paul, P. Kiran Mayee 5
If we look at the life cycle of an entity, we find that it has three major phases: crea-
tion, life and destruction. The purpose of an artifact is fulfilled at the second phase of
life cycle, namely, when it has life. Therefore, at this phase, the artifact gets associat-
ed to other entities without which the purpose cannot be fulfilled. For example, a
human being is an 'agent' who uses 'Pen' as an 'instrument' of writing. The writing is
done on a smooth surface, for example, Paper. Ink is a requisite for writing. Thus, the
artifact Pen is now related to the artifacts Ink and Paper as well as a 'human agent'
without whom the action of 'writing' will not take place. There might be a change of
state, for example, a Pen-Barrel can break; Ink gets over after a period of time. Final-
ly, in the third phase of the life cycle of the artifact, it undergoes destruction. For
example, ‘Pen’ undergoes destruction and gets converted to another entity, such as the
reuse of metal parts for making of some other entity, such as ‘Staple Pin’. It is there-
fore possible to engineer a knowledgebase of entities based on the characteristics
activities and states of entities. Whereas object-oriented paradigm suggests that ob-
jects should be the central focus for engineering knowledgebase, our observations on
entities suggest that entity-based knowledge cannot be complete unless it is focused
on the purpose of entities and the actions that the artifacts are involved in.
We formally define PurposeNet in the following terms:
PurposeNet is a knowledgebase of artifacts with its properties, relationships and
actions in which it participates with purpose as the underlying design principle.
2 Architecture of PurposeNet
PurposeNet has the artifact as its primary focus for organizing knowledge. Arti-
facts are fully described by its features and relationships with other artifacts. Two
kinds of features have been postulated for the task: descriptor features and action
features. The details of these features are given in section 2.1. Artifacts can also be
described by the company it keeps, i.e. its relation with other artifacts as illustrated
insection 2.2. The architecture of the PurposeNet is shown in the figure 2.
2.1 Features
The various distinct properties of an artifact are called its features. These features may
be morphological such as the physical state of the artifact, its size, shape, magnitude
and so on. The features may also be physiological like make, wear and tear, activities
it performs, and so on. Based on whether the feature is morphological or physiologi-
cal, we subcategorize features into the descriptor features and action features.
Descriptor Features.
The descriptor features of PurposeNet have three constituents that are found in
WordNet as well, viz., Name, Alias and Description. SUMO has one attribute Internal
that contains some properties which are similar to PurposeNet descriptor features.
However SUMO properties are limited to olfactory, visual, texture and taste, with no
6 PurposeNet: A Knowledgebase Organized Around Purpose
further refinement. The descriptor features of PurposeNet have been prepared after a
study of texts of Nyaya-Vaisheshikadarshana(Prasastipada(1977),
Singh(2001),Kulkarni(1994)) and others (Isvarkrsna(2007), Nagaraj(2003), Cow-
ell(2001)).
Descriptor Definition Value
Feature
Color The property possessed by an object of producing Red, Blue, Green,
different sensations on the eye as a result of the Cyan,Indigo, Orange,
way it reflects or emits light Pink, Black, White,
Any
Constitution The material with which an artifact is made of Metal, rubber, wood,
foam, plastic, glass
Shape The external appearance of an artifact Cubical, Oval, Trian-
gular, Circular,
Spherical, Aero, any
Size The amount of space occupied by the artifact Microscopic, very
small, small, medi-
um, large, any
State The physical state in which the artifact usually Solid, liquid, gas
exists
Table 1. Descriptor Features and their description
From the complex set of properties, we have selected twenty five based on the ones
most suitable for all artifact types. Also, we have added properties of significance
such as Standard Capacity, Standard Weight, and, Physical State to enable a more
comprehensive representation of information about artifacts. The possible values that
can be taken by these properties (qualitative) have been extracted from various
sources, including Wikipedia, Alani and Brewster(2006), Helmhotz (1970), Sunder
Rao (2003), and Gayatri Devi (2007). A brief description of some properties in de-
scriptor features is given in table 1. Comprehensive Descriptor feature list is given in
Appendix.
The value of some descriptor properties with respect to the artifact Car is given in
table 2:
SNo Decriptor_Feature Value
1 Name Car
2 Alias Automobile
3 Description A type of motor vehicle used to transport peo-
ple.
4 State Solid
5 Shape Aerodynamic
Rajeev Sangal, Soma Paul, P. Kiran Mayee 7
6 Color Any
7 Constitution Metal
8 Size Moderate_Size
Table 2.Values of Descriptor properties for the artifact Car
Action Features
Since the very need for an artifact is to serve some purpose in the human environ-
ment, it is understood that every artifact is associated with some actions. The various
activities associated with an artifact constitute its Action Features. This categoriza-
tion has been developed based on the various stages in the Lifecycle of an artifact.
The first stage of an artifact is its Make or Birth. It is then prepared for the first-time
use, after which it reaches the purpose-serving stage, i.e., Life. Here it may be pre-
pared again for reuse or may be in the general or repair-related maintenance stage.
From here, the artifact again goes back to the purpose-serving stage. After one ormore
iterations of the purpose-serving stage, the artifact becomes no longer usable, which is
when it is in the Destruction stage, and is therefore a last stage activity. Its individual
parts are recycled and it becomes the basis for the birth of the same or another catego-
ry of artifact. The various action features are accordingly classified primarily as –
make actions, purpose-serving actions, and, actions after destruction. The secondary
actions are first-time preparation-before-use actions that makes an artifact usable and
the trio of subsequent preparations before use actions, general maintenance and repair
maintenance actions that allow for subsequent usage of an artifact. Table 3 shows
these actions for Transport_using_Car artifact.
SNo Action Feature Value
1 Make/Birth 1. Integrate(Car_Interior_Parts)
2. Integrate(Chassis and Car_Body)
First-time-Preparation 1. Fill(Car_Fuel)
1a. before use 2. Test(Car_Pedals)
3. Test-Drive(Car) ….
2 Life - Purpose Transport things
Subsequent preparation 1. Check(Fuel)
2a. before use 2. Test(Car_Pedals)
3. Check(Rear_View_Mirror) …...
1. Repair(Car_Engine)
2b. Repair Maintenance 2. Repair(Car_Ignition_system)
3. Repair(Car_Pedals)
4. Repair(Car_Door) ….
1. Wash(Car)
2c. General Maintenance 2. Oil(Car_Engine)
8 PurposeNet: A Knowledgebase Organized Around Purpose
3. Oil(Ignition_System)
4. Fill(Car_Tyre)
1. Car_Engine - Recycled-to-metal
3 Destruction 2. Car_Tyre - Recycled-to-fuel-and-oil
3. Car_Chassis - Recycled-to-another-Car
4. Car_Seat - Reused-in-another-Car
Table 3. Table showing all the Action-features of a Car
Every non-primitive action can be fully described using a quadruple consisting of its
preconditions, outcomes, subactions and semantic roles. We call this Quadruple as the
action frame. Every primitive action can be described using the same frame as above,
minus the subactions. This description remains unchanged irrespective of the broad
category into which the action belongs – i.e., whether it is birth or make action, or
action related to life. The action frame places a formal structure on the Action features
(Kiranmayee et al., 2011).The action frame for a sample action, namely 'transport
thing', which is the purpose of the artifact Car is given in table 4.
Artifact: Car:: Purpose – Transport_Thing
No Action Frame Element Value(s)
1 Precondition 1) Exists_Car_at_Source
2) Exists_Thing_Near_Car
Result 1) Change_Position (Thing)
2 Out- Side Effect 1) Change_Position (Car)
come 2) Change_Position (Driver)
Wear-and- 1) Wornout(Engine)
tear 2) Wornout(Tyre)
1) Load(Thing)
3 Subactions 2) Drive(Car)
3) Unload(Thing)
1) Theme – Thing
4 Theta Roles 2) Source – Place
3) Destination – Place
4) All other Roles – Null
Table 4.The Action Frame for the action transport_thing_from_Source_To_Target
2.2 Relations
An artifact can also be described in terms of its association with other objects in
the world. For example, objects that come to our mind when we think of the artifact
Car might be the following: engine, wheel, steering, gear, seat, petrol, diesel, road,
Rajeev Sangal, Soma Paul, P. Kiran Mayee 9
petrol pump, car window, music system, rear view window, car body and so on. The
relations of these artifacts with car exist at different planes in terms of purpose that
the Car is used for. The primary purpose of Car as shown in table 3 is ‘transport
things from one place (X) to another place (Y)’. In order to fulfill the action of trans-
porting, a Car needs to move from X to Y and we call the action ‘drive’. For ‘drive’
action to take place, following parts of Car which claim to have a purpose of their
own, is essential: Engine, Wheel, Steering, Gear. Such components are called Core
Component. Rear view window is also part of Car but it is useful for some specific
movement of car (i.e., when the car moves back). Such components are called pur-
pose-serving-accessory in contrast to non-purpose-serving-accessory such as AC,
music system which are parts of Cars but are not directly related to Car driving. Oth-
er kind of artifacts such as petrol, diesel, road are directly related to driving even
though they are not part of Car. Such artifacts are related to Car with in terms of a
relation called Naccessory. Apart from these relations, there exist the usual subtype
relations between an artifact and its specific types. The following figure demonstrates
various relations and example cases for the artifact Vehicle:
Relation
Subtype Component Naccessory
{Car, Bus, ...} {Engine, Wheel, ...} {Petrol, Road, ...}
Core-Component Accessory
Purpose-Serving Non-Purpose-Serving
{Rear-view-window, …} {Music System, AC, ….}
Fig.3. Relations describing an artifact
10 PurposeNet: A Knowledgebase Organized Around Purpose
3 Implementation
The best possible design to represent our architecture of PurposeNet is object-
oriented and top-down methodology. The PurposeNet knowledge base has been im-
plemented using the concept of Ontology. Ontology is a formal explicit description of
concepts in a domain of discourse, properties of each concept describing various fea-
tures and attributes of the concept, and restrictions on slots. Ontology together with a
set of individual instances of classes constitutes a knowledge base (Noy, 2001). On-
tology helps us develop the Semantic Web, which is a vision for the future in which
information is given explicit meaning on the web, making it easier for machines to
automatically process and integrate information. We have chosen OWL to implement
our knowledgebase.
3.1 Statistics of PurposeNet Implementation
The active ontology for purposenet in Transport domain has an Artifact count of
3678 (Car_Door, Car, Car_Hinge ...), general property count of 87 (Color, Shape, …,
Birth, Processrel, ...), data property count of 8 (capacity, number, ...), Instances count
of 264 (Audi_A4, BMW_6_Series, Chevrolet_Tahoe, Daewoo_Matix, ...), and Sub-
Classes count of 8045 (Car_Rear_Seat, Car_Passenger_Seat, ...). The same is devel-
oped Semi-automatically by Domain Experts. The statistical data is given in table 5.
Rajeev Sangal, Soma Paul, P. Kiran Mayee 11
Metric Count (Transport Domain)
Class count 3678
Object Property count 87
Data Property count 8
SubClassOf axioms count 8045
SubObjectPropertyOf axioms count 76
Individual count 264
Annotation Assertion axioms count 1918
Class Assertion axioms count 258
Number of Assertions 1000 000
Table 5. Statistics of implementation of PurposeNet in OWL
4 Comparison with Other Ontologies
We evaluate PurposeNet from two perspectives:
1. Quality Evaluation in terms of various metrics as tabulated in table 6;
2. Estimation of how well the ontology represents the given search terms in the con-
text of ontology search engine.
4.1 Metric based Comparison
Three popular ontologies were selected for a metric-based comparison with Pur-
poseNet to evaluate its quality. The three ontologies selected are – the general Se-
mantic Web Technology Evaluation Ontology (SWETO) (Meza et. al, 2007), Gly-
comics Ontology (GlycO) (Satya et. Al, 2005), and, TAP (Guha and McCool,
2003).The results shows that PurposeNet scores much higher than all the other ontol-
ogy in terms of Class Importance (which determines the importance of a class by the
ratio of number of instances connected to the subtree attached to a class Ci in compar-
ison to the total number of instances (I) in the ontology, showing how many classes
play a central role compared to other classes). The completeness check (for populat-
ing relations, showing the percentage of relation slots filled in by values, thereby de-
termining how well the ontology can be utilized) yielded incompleteness for 7 of the
443 classes defined in the Car subtree.
12 PurposeNet: A Knowledgebase Organized Around Purpose
SN Metric SWETO TAP GlycO PurposeNet
o
1 Classes 44 6,959 361 3678
2 Relations 101 25 85,63 95
7
3 Instances 813,217 85,63 660 264
7
4 Schema Relationship Rich- NA NA NA 0.185
ness
5 Schema Inheritance Richness NA NA NA 1.68
6 Schema Attribute Richness NA NA NA 44
7 Class Richness 59.1 0.2 48.1 0.029
4
8 Class Connectivity 8 6 10 4
9 Class Importance (max. val- 59 31 18 100
ue)
Cohesion NA NA NA 881
10
11 Class Relationship Richness NA NA NA 100
(max. value)
Table 6. Comparative representation of various ontology metrics
4.2 Comparative rank scores of PurposeNet and akt ontology for browser-
retrieval
The efficiency of an ontology can also be determined based on the rank search en-
gines on the web gives. Browser-wise, ontologies are usually ranked based on three
criteria – user popularity, evaluation tests and structural criteria (Gangemi, 2006). An
ontology may be ranked structurally based on CMM, DEM and SSM (HarithAlani
and Christopher Brewster, 2006). We have used the reference ontology of the akt
(Advanced Knowledge Technologies) project on extraction and use of knowledge
(Motta, 2001). The observations with respect to the various ranking measures in Pur-
poseNet in comparison to the best ontology (ranked 1) outcomes obtained by Alaniet.
al. (2005) with respect to the akt reference ontology is tabulated in table 7 below. It is
observed that the akt ontology performed better with respect to the CMM (Class
Match Measurement, the number of concepts in the ontology that either match (M) or
contain the search term (C), that determines how many search terms exactly match
with terms in our ontology, that presents the certain degree of detail in the representa-
tion of the knowledge concerning that concept) as well as DEM scores(Density, the
number of superclasses (U), subclasses (S), attributes (A) and siblings (I) associated
with the individual concepts in the ontology), whereas, PurposeNet had a better SSM
score(Semantic Similarity, how close related terms are placed in the ontology, where,
ontologies that position concepts further away from each other are less likely to repre-
sent the knowledge in a coherent and compact manner. It is measured by the path
Rajeev Sangal, Soma Paul, P. Kiran Mayee 13
distance between the two different concepts in question), favoring its faster represen-
tation on Swoogle.
Ontology CM DEM SSM Total
M Score
PurposeNet 0.786 0.589 0.413 0.596
akt reference ontology 0.833 0.632 0.250 0.571
Table 7. Comparative rank Scores of PurposeNet ontology and akt ontology
5 Purpose Detection and Extraction
The method of knowledge discovery by manual extraction of data and manual
building of PurposeNet ontology is quite exhaustive as several experts are required to
put in hours of browsing to find the data corresponding to the concerned features and
to incorporate it. This also leads to a slow progress in the creation of a knowledge
base that was supposed to finally have a size of a million artifacts. We follow a two-
step process for the extraction of data from the web. The first task is to find an appro-
priate method to detect the presence or absence of a relation. The second step would
be to extract the relation from the text that is known to contain the semantic relation.
This methodology has been applied on the purpose relation as a case study for gener-
alization across all other relations in PurposeNet.
5.1 Purpose Detection
Sentences containing particular relations have specific structure(s) in terms of a
key word or words in a particular order. We select WordNet as the corpus for our
work. The principle behind the selection of the WordNet as the corpus is the observa-
tion that 70% of the WordNet corpus contains purpose data. We perform automatic
detection by transforming the problem of relation detection to a binary classification
problem. There are many supervised as well as unsupervised methods of classification
that have been graded equally well in other domains. Some of these are the Typed
Dependency Parse (Catherine et. Al, 2006), Decision tree forest (
http://www.dtreg.com/treeforest.htm ), the Naïve Bayes method (Bayes et. Al, 1763),
the kernel based Neural Network approach and the more popular Support Vector Ma-
chine (Vapnik et. al, 1995) based approach. A comparative study of these various
methods of detection of purpose data in table 8 shows that the typed dependency and
simple decision trees method of detection gives maximum precision over others. A
comparison of the various recall values shows that the typed dependency method has
the highest recall. Hence, we suggest the typed dependency method as the most favor-
able among all methods of purpose detection.
14 PurposeNet: A Knowledgebase Organized Around Purpose
Sno Method Precision Recall F-Measure
1 Typed dependency 0.84 0.68 0.751
2 Simple Decision 0.83 .67 .74
Tree
3 Decision Tree Forest 0.679 0.644 .661
4 Bagging .755 .619 .68
5 Naïve Bayes .7 .638 .668
6 Bayes Net .699 .639 .668
7 RBF Neural Net- .679 .595 .634
work
8 SVM .694 .639 .665
Table 8. Comparison of efficiencies of various automatic purpose detection methods
5.2 Purpose Extraction
Our target is to extract the artifact whose purpose is known to be available in text.
This section explains the three methods used for extraction of purpose from text: a.
Clue Based Extraction, b. Extraction using Typed Dependency Parse and c. Extrac-
tion using Surface Text Pattern.
Method Precision for extraction of (artifact, action) pair
given purpose-containing sentences
Purpose clues 69
Surface Text Patterns 88
Typed dependency Parse 98.1
Table 9. Comparativeperformancemeasuresof various purpose extraction methods
Table 9 shows a comparison of the performance of the three methods. The results
show that Typed dependency method performs well in extraction of (artifact, purpose)
pair. Surface Text Patterns perform well too, considering that the entire web is its
corpus, vis-a-vis the other two methods which used offline corpora.
Rajeev Sangal, Soma Paul, P. Kiran Mayee 15
6 Applications
PurposeNet has a number of applications in various reasoning tasks, including Ques-
tion Answering (QA), provision of online help in web pages, aiding expert systems
and broadly in Natural Language Understanding. We describe an application that we
have built to evaluate our ontology.
6.1 Domain Specific Question Answering
In this application, a passage is given as input to the automated QA system and the
output to a set of questions is obtained. The same task is given to an average car user
and the two outputs are compared.
Design.
We have built four alternative modules and each module uses a different resource for
producing the answer. Module 1 uses only the passage from where the answer is to be
retrieved. Module 2 uses passage and script; Module 3 uses passage and PurposeNet
and Module 4 uses passage, script and PurposeNet. We have used a racing car text to
test the modules. We have developed a script for racing car. A script (Schank, 1974)
is a structure that prescribes a set of circumstances which could be expected to follow
on from one another. PurposeNet contains information which is true for an artifact in
all circumstances and a script is a structure that prescribes a set of circumstances
which could be expected to follow on from one another. It is similar to a thought se-
quence or a chain of situations which could be anticipated. The components of the
script for the text are:
1. Entry Conditions – the conditions that must be satisfied before events in the script
can occur.
2. Results – Conditions that will be true after events in script occur.
3. Props – Slots representing objects involved in events.
4. Roles – Persons involved in the events.
5. Track – Variations on the script. Different tracks may share components of the
same script.
6. Scenes – The sequence of events that occur. Events are represented in conceptual
dependency form.
The theme of car racing can be segmented into 5 scenes: 1. Arranging track; 2.
Prepare for the race; 3. The race; 4. The finish; 5. The victory lap.
Script:Car Race Track: American Car Race – A Win
Props: Roles:
R = Race Car D = Car Driver
T = Race Track S = Spectator
F = Checkered Flag Q = Pit team
G = Shot gun O = Organizer
16 PurposeNet: A Knowledgebase Organized Around Purpose
P = Petrol
L = Finish Line
Entry Conditions: Results:
• T exists • D has more money.
• R exists • D has won the race.
• D exists • R has less P.
Scene1: Arranging the track Scene 2 : Prepare for Race
• O sprinkles T • O checks T
• O grinds T • O signals R line-up
• (go to scene 2) • D line-up R
• D test-drive R
• O signals start race with G
• (go to scene 3)
Scene 3: Race Scene 4: Finish Race
• D accelerates R • D crosses L.
• D steers R • (go to scene 5)
• (go to scene 4)
Scene 5: Victory Lap
• D gets F.
• D waves F.
• D drives on T.
Table 10. A simplified racing script
The complete Script could be described in Figure above.
Result.
Experiments were conducted on answering questions where both the passage and
the questions were given as input to each of the 4 modules and compared with theout-
put of human users. The results show that the comprehension passage alone yielded
6% of the answers. These were Queries that were directly related to the story in the
Comprehension passage, such as –Did the drivers test-drive? 10% of the queries re-
lated to Car race are answered by PurposeNet alone. These were technical Queries
related to Cars such as – How did the pit Team repair Clint’s car tyre? 27 % of the
queries are answered using Scripts alone. These pertained to the sequence of events in
a stereotypical Car race, such as – What is the connection between waving the check-
ered-flag and the victory-lap?
Rajeev Sangal, Soma Paul, P. Kiran Mayee 17
Resource used to No. of Que- Efficiency % of answers
SNo obtain Answer ries correctly (/30) in % given using this
replied (/30) resource
1 Comprehension 2 6 12.5
Passage
2 Script 8 27 50
3 PurposeNet 6 20 37.5
4 PurposeNet + 14 + 3 57 89
Script
Total 19
Table 11. Comparative results of Queries answered by AOM Script Applier using various
resources
7 Conclusion and Future Work
The paper presents the conceptual base, architecture and implementation of a se-
mantic knowledgebase called PurposeNet with an evaluation performed on it compar-
ing it with some other available knowledgebase. Building an exhaustive knowledge-
base is a laborious and intense task, it needs human expertise and it needs good web
data processing tools so that information from the web can be easily extracted in order
to build the knowledgebase semi-automatically. In order to maintain the quality of the
resource, we have, till now, manually created the knowledgebase. Nevertheless, we
understand that creating such huge resource completely in manual mode would be a
time-consuming work. We have noticed that artifact related information which is
useful for our knowledgebase is available in various resources such as WordNet, Wik-
ipedia and other web corpora. We have conducted a few experiments on detecting and
extracting purpose of artifacts from web corpus and reported the result in this paper.
Experimental results in domain-specific question-answering have produced promising
results.
References
1. Alani, H., Brewster, C.: Ontology ranking based on the analysis of concept structures. In:
Proceedings of the 3rd international conference on Knowledge capture (K-CAP '05).
ACM, New York, NY, USA, 51-58 (2005)
2. Alani, H., Brewster, C.:Metrics for ranking ontologies. In: WWW2006, May 22–26, 2006,
Edinburgh, UK.Harry D Patton. Physiology of Smell and Taste: Annual Review of Physi-
ology 12. pp 469–484 (2006)
3. Aleman-Meza, B., Halaschek, C., Sheth A., Arpinar, I. B., Sannapareddy, G.: SWETO:
Large-Scale Semantic Web Test-bed. In: Proceedings of the 16th SEKE 2004: Workshop
on Ontology in Action, Banff, Canada, June 21-24, 2004, pp.490-493 (2004)
4. Bharati, A., Chaitanya, V., Sangal, R.: Natural Language Processing: A Paninian Perspec-
tive", Prentice-Hall of India, New Delhi (1995) (Down-
load:http://ltrc.iiit.ac.in/downloads/nlpbook/nlp-panini.pdf).
18 PurposeNet: A Knowledgebase Organized Around Purpose
5. Bharati, A., Nawathe, S.A.,Chaitanya, V., Sangal, R.: A New Inference Procedure for
Conceptual Graphs. In: Proc. of 4th University of New Brunswick Artificial Intelligence
Symposium (1991)
6. Cowell, E. B., Gough, A. E.: The Sarva-Darsana-Samgraha or Review of the Different
Systems of Hindu Philosophy: Trubner's Oriental Series, Taylor & Francis (2001)
7. Lenat, D. B.: CYC: a large-scale investment in knowledge infrastructure, Communications
of the ACM, v.38 n.11, p.33-38 (1995)
8. Gangemi, A., Catenacci, C., Ciaramita, M., Lehmann, J.: Modelling Ontology Evaluation
and Validation. In: Proceedings of the 2006 European Semantic Web Conference (2006)
9. Devi, G.: Padartha Vijnana made easy. Chaukhamba Sanskrit Pratishthan, Delhi (2007)
10. Miller, G., Beckwith, R., Fellbaum, C., Gross, D., Miller, K.J.: WordNet: An online lexical
database. International Journal of Lexicography, 3(4) (1990)
11. Iśvarakṛṣṇa, Sāṁkhyakārikā with SankaraMisra’s commentary Sāṁkhyatattvakaumudi ,
Edited and translated into Hindi by Nigam Sharma. Varanasi: Parimal Prakashan (2007)
12. Mayee, P.K., Sangal, R., Paul, S.: Action Semantics in PurposeNet. In: Proceedings of
2011 World Congress on Information and Communication Technologies. IEEE WICT'11.
pp. 1299-1304 (2011)
13. Kulkarni, Amba P, Navya-Nyaya and Logic, MTech Thesis, IIT Kanpur (1994)
14. Liu, H., Singh, P.: ConceptNet: A Practical Commonsense Reasoning Toolkit. BT
Technology Journal, Volume 22. Kluwer Academic Publishers (2004)
15. Nagaraj, A, Manav Vyavhar Darshan, Jivan Vidya Prakashan, Amarkantak, 2003.
16. Praśastapāda, Padārthadharmasamgraha with Sridhara's commentary Nyāyakandali, edited
and translated into Hindi by Srī Durgadhara JhaSharma. Vārānasī: Sampurnananda
Sanskrita University (1997)
17. Ram Sunder Rao . M. Ayurveda Padardha Vijnana (2003)
18. Sangal, R., Chaitanya, V.: An Intermediate Language for Machine Translation: An Ap-
proach based on Sanskrit Using Conceptual Graph Notation, Computer Science and In-
formatics Journal, Computer Society of India, 17, 1, pp. 9-21 (1987)
19. Singh, N.: Comprehensive Schema of Entities: Vaiśeṣika Category System,” Science Phi-
losophy Interface, Vol. 5, No. 2, pp 1-54 (2001)
20. Sowa, J. F.: Conceptual Structures: Information Processing in Mind and Machine, Addi-
son-Wesley, Reading, MA (1984)
21. Sowa, J. F.: The Challenge of Knowledge Soup. In: J. Ramadas & S. Chunawala, Research
Trends in Science, Technology, and Mathematics Education, Homi Bhabha Centre, Mum-
bai, pp. 55-90 (2005)
22. Sowa, John F.: Relating diagrams to logic. In: G. Mineau, B. Moulin, & J. F. Sowa eds.,
Conceptual Graphs for Knowledge Representation, Springer-Verlag, New York (1993)
23. Tartir, S., Arpinar, I. B., Moore, M., Sheth, A. P., Aleman-Meza, B. OntoQA: Metric-
based ontology quality analysis. In Proceedings of the IEEE Workshop on Knowledge Ac-
quisition from Distributed, Autonomous, Semantically Heterogeneous Data and
Knowledge Sources (ICDM'05), Boston, MA. (2005)
24. Varma, V.: Building Large Scale Ontology Networks. Language Engineering Conference
(LEC'02) December 13 - 15, 2002. India, p. 121(2002)
Appendix
SNo Description Definition Values
Feature
Rajeev Sangal, Soma Paul, P. Kiran Mayee 19
1 Color The property possessed by an Red, Blue, Green, Yel-
object of producing different low, Cyan, Indigo, Orange,
sensations on the eye as a Pink, Black, White, Any.
result of the way it reflects or
emits light
2 Constitution The material with which an Metal, Rubber, Wood,
artifact is made up of Foam, Plastic, Glass, etc.
3 Fluidity The physical property of a Fluid, Nonfluid
substance that enables it to
flow
4 Heaviness The comparative weight of an Heavy, Light, Moderate
artifact Weight
5 Inertness The reactivity of an artifact Inert, Alkaline, Acidic
with the substances around it
6 Mobility The movement of an artifact Mobile, Immobile
during the performance of its
target task
7 Oiliness The presence of oil on the Oily, NonOily
surface of the artifact
8 Position The position of an artifact vis- Above, Below, Inside,
à-vis the artifact it is embed- Left_Of, Right_Of,
ded in In_Front_Of, Behind
9 Shape The external appearance of an Cubical, Spherical, Circu-
artifact lar, Oval, Triangular,
Aero, any
10 Size The amount of space occupied Microscopic, very
by the artifact small, small, medium,
large, any
11 Sliminess The sticky, slippery property Slimy, Nonslimy
of an artifact
12 Smell The property of an artifact that No odour, Weak, Very
is sensed by the nose Weak, Strong, Intolerable
13 Smoothness The property of having a sur- Smooth, Rough, Sharp,
face free from projections or etc.
irregularities
14 Softness The property wherein the arti- Soft, hard
fact gets deformed on applica-
tion of pressure
15 Sound Mechanical vibrations emitted Silent, whisper, beara-
by artifacts when they func- ble_sound, unbeara-
tion ble_sound
16 Stability Indicates whether the given Stable, Unstable
artifact remains as it is or dis-
integrates into the environ-
20 PurposeNet: A Knowledgebase Organized Around Purpose
ment
17 State The physical state in which the Solid, Liquid, Gas
artifact usually exists
18 Subtleness Indicates whether an artifact is Subtle, Nonsubtle
so slight that it is difficult to
perceive
19 Taste Indicates the property of an Sweet, Sour, Bitter,
artifact that is perceived by the Umami, Salty
tongue
20 Temperature Indicates the temperature at Hot, Cold, Warm, Nor-
which the artifact usually ex- mal,, Cool
ists
21 Transparency The property of the surface Transparent, Opaque,
of an artifact that allows a Semi-transparent
human to see through it
22 Std. Capacity Maximum weight that this ….kgs, ….lbs,…ltrs
artifact can hold
23 Std. Magnitude Standard dimensions of the ….metres
artifact
24 Std. Weight Weight of this artifact ….kgs, ….lbs