Docstoc

Concept Measure Supported Aesthetics Ontology Construction for Videos

Document Sample
Concept Measure Supported Aesthetics Ontology Construction for Videos Powered By Docstoc
					                                                         (IJCSIS) International Journal of Computer Science and Information Security,
                                                         Vol. 8, No. 9, December 2010




           Concept Measure Supported Aesthetics Ontology
                     Construction for Videos
                                                        Dr Sunitha Abburu
                                    Professor & Director, Department of Computer Applications
                                               Adhiyamaan College of Engineering
                                               Hosur, pin-635109, Tamilnadu, India.
                                                     ausunithaa@yahoo.co.in



Abstract—Entertainment plays a vital role in human life.                improve the effectiveness and the efficiency of the video
Multimedia conquers top position in the entertainment world.            retrieval system video semantics should be identified,
Video stands first among the multimedia entertainment. The              represented and must be used during video object retrieval. The
rapid growth of the videos has resulted in the emergence of             semantic annotation generation can be manual or automated.
numerous multimedia repositories that require efficient and             The annotation generation system and the retrieval system
effective video storage, semantic annotation, indexing and              performance increases by considering the ontology. Ontology
retrieval systems. The introduction of ontologies in multimedia         plays a vital role in artificial intelligence, semantic web,
retrieval system can improve the precision and recall rate              software engineering, information retrieval, knowledge
effectively. The performance of the annotation and the retrieval
                                                                        representation, knowledge sharing, knowledge integration,
system increases with the support of the domain ontology. Most
of the video annotation, indexing and the retrieval systems focus
                                                                        knowledge reuse, and so on. It is a well known fact that the
on the semantic concepts like objects, people, location, events,        performance of the annotation and the retrieval system
actions etc. But most of the multimedia systems are filled with         increases with the support of the domain ontology. The
human and their emotions. Any multimedia system like cinema,            introduction of ontologies in multimedia retrieval system can
news videos, sports videos, and any domestic functional videos          improve the precision and recall rate effectively. Focusing on
tries to capture the emotions of the human involved in the              the completeness and the effectiveness of the video retrieval
occasion. A video retrieval system will be complete if the system       system raises the need for multiple sub ontologies by
identifies, captures and represents the emotions of the humans.         considering the various aspects of the video semantics. The
This paper focus on identification and representation of human          literature shows that most of the video annotation, indexing and
emotions and the intensity of the emotions are represented using        the retrieval systems focus on the semantic concepts like
the fuzzy logic. The concept of Navarasra has been brought in to        objects, people, location, events, actions etc. But most of the
video retrieval system to classify the human emotions. The              videos are filled with human and their emotions. Any
proposed approach is generic and flexible. It is designed and           multimedia system like cinema, news, sports, and any domestic
constructed for all videos where human beings and their                 functional videos tries to capture the emotions of the human
emotions are been captured. A practical implementation is done          involved in the occasion. A video retrieval system will be
using Protégé as an Ontology developing tool.                           complete if the system identifies, captures and represents the
                                                                        emotions of the humans and the intensity of the emotions.
   Keywords-component; Video Semantics, Concept measures,
                                                                        Intense understanding of the user’s perspective and their
Ontology, Retrieval, Human Emotions.
                                                                        expectations towards the video retrieval system is essential.
                                                                        The user queries can be pertaining to the human emotions and
                      I.    INTRODUCTION                                the intensity of the emotions involved in the video. This paper
    Entertainment plays a vital role in the human life.                 focus on identification and representation of human emotions
Multimedia plays an important role in the entertainment world.          and the intensity of the emotions are represented using the
Video systems stand first in multimedia presentations. The              fuzzy logic.
reason why a lot of research work, in the area of multimedia, is        The rest of the paper is organized as follows. Literature survey
carried out on video data compared with other multimedia data
                                                                        report is in section 2. Section 3 discusses the proposed method
types is twofold. First, video contains audio, visual, text
                                                                        for identification and representation of human emotions and
information etc. It is the most powerful and at the same time
most complex, voluminous, unformatted, and unstructured of              their intensities. In section 4, we present a practical
all media used for conveying information. Hence, representing           implementation and experimental results on aesthetic ontology
video information to enable effective and efficient retrieval is        construction. Finally, we conclude with a summary and future
an interesting problem. Second, news videos, cinema videos              work in section 5.
and the resent trend of capturing any occasion, function, event
                                                                                              II.   RELATED WORK
in videos, raises the demand for the efficient video storage,
semantic annotation, indexing and retrieval systems. Video              What is an ontology, the answer is twofold as given in [1],
retrieval systems which focus on the low level features and             philosophical and computing. In the context of philosophy,
ignore the semantic are less efficient and effective. In order to       Ontology is the philosophical study of the nature of being,



                                                                    4                               http://sites.google.com/site/ijcsis/
                                                                                                    ISSN 1947-5500
                                                        (IJCSIS) International Journal of Computer Science and Information Security,
                                                        Vol. 8, No. 9, December 2010




existence or reality, as well as the basic categories of being         the emotions of the humans. All these applications need to
and their relations. Traditionally listed as a part of the major       retrieve the video objects based on the human emotions. The
branch of philosophy known as metaphysics, ontology deals              proposed approach is generic flexible and not specific to any
with questions concerning whether entities exist or can be said        video application. It is designed and constructed for all videos
to exist, and how such entities can be grouped, related within a       where human beings and their emotions are been captured. In
hierarchy, and subdivided according to similarities and                sports video sports players, sponsor etc would like to see all the
differences. In the context of computer and information                video objects where the audiences are happy, overwhelm, sad
sciences [2], an ontology is an explicit specification of a            etc (score, out, amazing shots). In cinema the user would like to
conceptualization. An ontology defines a set of                        watch the video objects with emotions like of comedy, sad,
                                                                       compassion, pathetic, furious, anger, heroic, energy, terror,
representational primitives with which to model a domain of
                                                                       horror, astonishment, surprise, tranquility etc. In news domain,
knowledge or discourse. The representational primitives are            the news reader would like to display the videos of the news
typically classes (or sets), attributes (or properties), and           clippings pertaining to the emotions of the human as
relationships (or relations among class members). The                  mentioned. Sports video ontology [4], the concept based video
definitions of the representational primitives include                 retrieval system for sports video explore the method of
information about their meaning and constraints on their               constructing concept ontology for sports video by identifying
logically consistent application. This set of objects, and the         the concepts, concept hierarchy and the relations ships. The
describable relationships among them, are reflected in the             concepts like events, actions, players etc are identified and
representational vocabulary with which a knowledge-based               represented in an ontology. The current research on
program represents knowledge. Ontology is a kind of concept            construction of ontologies is focusing on identification of
model that could describe system at the level of semantic              concepts like events, actions, objects, locations, and people.
knowledge. It aims to access knowledge in a domain in a                But most of the video retrieval requirements are pertaining to
general way and provides a common understanding for                    the human and their emotions involved in the video. Semantic
concepts in the domain so as to realize knowledge sharing and          video retrieval efficiency can be increased by considering the
reuse among different application programs and organizations.          emotions of the humans involved in the video. The semantic
As a new kind of knowledge organization tool and an                    video retrieval system will be more effective, if the retrieval
ontological commitment is an agreement to use a defined                system supports the retrieval of video objects or images based
vocabulary by a group of people agreed upon in a coherent              on the human emotions.
and consistent manner. N. F. Noy, and D. L. McGuiness in [3]
describe the need for ontology as:                                     A. Human Emotions - Aesthetics
                                                                       The ancient scriptures describe nine fundamental emotions
        To share common understanding of the structure of             from which all complex emotions may be produced. Just as
         information among people or software agents.                  all shade of colors are produced from basic RGB-three
        To enable reuse of domain knowledge.                          primary colors. In the same way all emotions are said to be
        To make domain assumptions explicit.                          derived from principal emotions known as Navarasa (in
        To separate domain knowledge from the operational             Sanskrit). Sanskrit, an ancient language of India, is also one of
         knowledge.                                                    the oldest languages in the world. Sanskrit is a member of the
        To analyze domain knowledge.                                  Indo-European language family and is a classical language of
                                                                       India. The word Sanskrit is derived from 'sam' which means
At present, the main methods of ontology construction are:             'together' and 'krtam' which means 'created'. 'Sanskrit' together
TOVE, Skeletal, METHONTOLOGY, KACTUS, SENSUS,                          means completed, refined and perfected. Nava means 'Nine'
IDEF5 and Seven Steps method. Ontology development                     and Rasa signifies 'mood,' 'emotion,' 'expression' or
process is an iterative process that will continue in the entire       'sentiment.' The Navarasa - aesthetics in the scriptures refer to
life cycle of the Ontology. An ontology is typically built in          the nine expressions that humans often show. The long
more-or-less the following manner. The basic steps for                 standing concept of Navarasa is a way to express the emotions
building Ontology are [3]:                                             of human that is exceptionally original. The individuality of
                                                                       the characters is the element that each character is the
        Determine the domain and scope of the ontology.               personification of one rasa-emotion. Their nature, the intensity
                                                                       of their reactions, their strengths, their failings - all guided by
        Consider reusing existing ontology.
                                                                       the rasa they represent, which in turn plays an important role
        Enumerate important terms in the ontology.
                                                                       in Video semantic retrieval. Video objects which are retrieved
        Define the classes and the class hierarchy.                   based on the Navarasa makes the retrieval system highly
        Define the properties of classes—slots.                       efficient and the only one of its kind ever made. Navarasa is
        Define the facets of the slots.                               accepted worldwide and been used in all art forms. Navarasa
        Create instances.                                             are the emotions that human show according to the situations.
 III.   HUMAN EMOTION CONCEPTS AND CONCEPT MEASURES
                                                                       The Nine Moods - Aesthetics (Nava Rasa) are:
    Various multimedia applications like sports, news, cinema               Shringar – Love, Attractiveness, Amour
or any video captured at a function/occasion tries to captures



                                                                   5                               http://sites.google.com/site/ijcsis/
                                                                                                   ISSN 1947-5500
                                                        (IJCSIS) International Journal of Computer Science and Information Security,
                                                        Vol. 8, No. 9, December 2010




        Hasya – Comic, Laughter, Mirth, Comedy                        represented using the fuzzy logic see figure 1 and 2. Concept
        Karuna – Sadness, Kind-heartedness or Compassion,             measures are ranked on a scale of      0 to 1. The descriptions
         Mercy, Pathetic                                               (d) can be described using concept measure (cm) as:
        Raudra – Furious, Anger, Fury
        Veera – Heroic, Courage, Energy
        Bhayanak – Terrible, Fear, Terror, Horror
        Bibhatsam - Vibhats , Disgusting, Odious
        Adbhuta – Wonderment, Amazement, Astonishment,
         Surprise
                                                                                              Figure 1. Crisp Sets
        Shanta – Peace, tranquility.
      In addition to the nine Rasas, two more appeared later in
literature are, Vātsalya - Parental Love and Bhakti - Spiritual
Devotion.
NavaRasa plays a significant influence on Indian cinema. The
Rasa method of performance is one of the fundamental                                          Figure 2. Fuzzy Sets
features that differentiate Indian cinema from that of the
Western world. In the Rasa method, empathetic "emotions are            d1 = {(excellent, 1), (good, 0.75), (average, 0.5), (bad, 0.25),
conveyed by the performer and thus felt by the audience," in           (worst, 0.0)}. d2 = {(very high, 1), (high, 0.8), (medium, 0.6),
contrast to the Western Stanislavski method where the actor            (low, 0.4), (very low, 0.2), (ground level, 0.0)}. The queries
must become "a living, breathing embodiment of a character"            from different application domains that have human emotion
rather than "simply conveying emotion."                                and the intensity of the emotions like:
                                                                       Sports: Show all excellent shots of Sachin.
All videos which involve human beings implicitly convey the            Cinema: Show all, incredible comedy clippings Mr. Bean.
emotions. In order to build a more effective retrieval system,         News: Show most terrible scenes of xxx disaster, etc.
human aesthetics-emotions and the intensity of the emotions
should be identified, represented, stored and should be used                       IV.    A PRACTICAL IMPLEMENTATION
during the video object retrieval.                                     [5][6][7] [8] describes ontology tools. We have used Protégé
The queries from different application domains that involve            as an Ontology developing tool to represent the human
human emotions like:                                                   emotion concepts. Protégé [9] was developed by
Sports: Show all incredible shots of Sachin.                           (http://protoge.stanford.edu) Mark Musen‘s group at Stanford
Cinema: Show all comedy clippings of Mr. Bean.                         University. The human emotions are represented in the
News: Show all terrible scenes of xxx disaster, etc.                   Aesthetics ontology using Protégé. Aesthetics ontology and
                                                                       onto graph plug-in representation is shown in Fig.3, Fig.4, and
B. Concept Measure                                                     Fig.5. We selected OWL, as the ontology language, which is
                                                                       standard ontology language recommended by W3C.
All human emotions inhabit different levels of intensity or
extent. Or the concept like beauty can be measured as more
beautiful, most beautiful, less beautiful, least beautiful. The
intensity of the emotion courage can be measured as less,
good, very good, incredibly courage. The intensity of the
emotion is defined as Concept Measure. All degree adverbs
like good, bad, high, low, very, most, more, less, incredibly,
worst, bad, excellent, quickly, loudly, awfully, fairly, quite,
really, extremely, pretty, fully, almost, little, too, partial,
completely, adequately, immensely, perfectly, strongly etc are
concept measures. To describe the human emotions, events
and actions more effectively concept measures are attached to
the concepts. This can be represented as, humans involved in
the video, emotions and the intensity of the emotions. Actions
and the intensity. Concept measures like easy, good, bad,
worst, high, low etc are purely human judgment. And always
there is a chance that two individuals may judge it differently.
A particular scene may be too comedy to one individual which
is just comedy for other individual. The degree of judgment
varies from one individual to other. The fuzziness involved in
human judgment pertaining to the concept measures are                                     Figure 3. Aesthetics Ontology




                                                                   6                               http://sites.google.com/site/ijcsis/
                                                                                                   ISSN 1947-5500
                                                                (IJCSIS) International Journal of Computer Science and Information Security,
                                                                Vol. 8, No. 9, December 2010




                                                                              Ontology is a kind of concept model that could describe
                                                                              system at the level of semantic knowledge. It aims to access
                                                                              knowledge in a domain in a general way and provides a
                                                                              common understanding for concepts in the domain so as to
                                                                              realize knowledge sharing and reuse among different
                                                                              application programs and organizations. The paper identifies,
                                                                              classifies and represents the human emotions in aesthetics
                                                                              ontology. The intensities of the emotions are drawn using the
                                                                              fuzzy logic. The current paper constructs the aesthetics
                                                                              ontology. Further research could be conducted on semi or
                                                                              automatic identification, extraction, annotation generation of
                                                                              human emotions and the retrieval of video objects using
                                                                              aesthetics ontology.


         Figure 4. Onto Graph notation of Aesthetics Ontology                                           ACKNOWLEDGMENT
                                                                              This work has been partly done in the labs of Adhiyamaan
                                                                              College of Engineering where the author is currently working as a
                                                                              Professor& Director in the department of Master of Computer
                                                                              Applications. The author would like to express her sincere thanks
                                                                              to Adhiyamaan College of Engineering for their support rendered
                                                                              during implementation of this module.

                                                                                                             REFERENCES
                                                                              [1]   El-Ghalayini, H., (2007). Reverse Engineering Domain Ontologies To
                                                                                    Conceptual Data Models, PhD thesis, University of the west of England,
                                                                                    UK.
                                                                              [2]   T.R.Gruber, Toward principles for the design of ontologies used for
                                                                                    knowledge sharing, Human Computer Studies, 43 (5-6), 1995, 907-928.
                                                                              [3]   N. F. Noy, and D. L. McGuinness, ―Ontology Development 101: A
                                                                                    Guide to Creating Your First Ontology, Stanford Knowledge Systems
                                                                                    Laboratory Technical Report, KSL-01-05, 2001.
                                                                              [4]   Sunitha Abburu, et al, ”Concept ontology construction for sports video”
                                                                                    Proceedings of the 1st Amrita ACM-W Celebration on Women in
                                                                                    Computing in India 2010, Coimbatore, India September 16 - 17, 2010.
                                                                              [5]   F. L. Mariano and G. P. Asunción, ―Overview and analysis of
                                                                                    methodologies for building ontologies, The Knowledge Engineering
                                                                                    Review, vol. 17(2), pp. 129–156, 2002.
                                                                              [6]   Cimiano P., Volker J., and Studer R., ―Ontologies on Demand? – A
        Figure 5. OWL/XML rendering of Aesthetics Ontology                          Description of the State-of-the-Art, Applications, Challenges and Trends
                                                                                    for Ontology Learning from Text Information, Wissenschaft und Praxis,
                                                                                    Vol. 57, No. 6-7. (October 2006), pp. 315-320.
             V.    CONCLUSION AND FUTURE WORK                                 [7]   Duineveld A. et al. ―Wonder Tools‘ A Comparative Study of
                                                                                    Ontological Engineering Tools. Intl. Journal of Himian-Computer
Due to the rapid growth of the voluminous video repositories,                       Studies. Vol. 52 No. 6, pp. 11 11-1 133. 2000.
the demand for efficient and effective video storage, semantic                [8]   Michael Denny, Ontology Building, A Survey of Editing
annotation, indexing and retrieval systems is growing day by                        Tools,http://www.xml.com/ pub/a/2002/11 / 06/ ontologies.html.
day. Domain ontology plays an important role in information                   [9]   Matthew Horridge, Simon Jupp, Georgina Moulton, Alan Rector, Robert
retrieval, knowledge representation, knowledge sharing,                             Stevens, Chris Wroe. OWL Ontologies using protégé 4 and CO-ODE
                                                                                    Tools Edition 1.1. The University of Manchester , October 16,2007.
knowledge integration, knowledge reuse and so on. Ontologies
focusing only on the concepts like persons, things, events,
                                                                                                          AUTHORS PROFILE
actions, places are not complete. Since most of the videos are
                                                                               Dr. Sunitha Abburu: Working as a Professor and Director, in the Department
captured to capture the human and their emotions. The video                   of Computer Applications, Adiyamaan College of Engineering, Tamilnadu,
storage, annotation and the retrieval system would be                         India. She received BSc and MCA from Osmania University, A.P, India.
complete by considering the human emotions and intensity of                   M.phil and Ph.D from Sri Venkateswara University, A.P, India. She is having
the human emotions. This raises the need for identification,                  13 years of teaching experience and 3 years of industrial experience.
classification and representation of human emotions. As




                                                                          7                                    http://sites.google.com/site/ijcsis/
                                                                                                               ISSN 1947-5500

				
DOCUMENT INFO
Description: The International Journal of Computer Science and Information Security (IJCSIS) is a well-established publication venue on novel research in computer science and information security. The year 2010 has been very eventful and encouraging for all IJCSIS authors/researchers and IJCSIS technical committee, as we see more and more interest in IJCSIS research publications. IJCSIS is now empowered by over thousands of academics, researchers, authors/reviewers/students and research organizations. Reaching this milestone would not have been possible without the support, feedback, and continuous engagement of our authors and reviewers. Field coverage includes: security infrastructures, network security: Internet security, content protection, cryptography, steganography and formal methods in information security; multimedia systems, software, information systems, intelligent systems, web services, data mining, wireless communication, networking and technologies, innovation technology and management. ( See monthly Call for Papers) We are grateful to our reviewers for providing valuable comments. IJCSIS December 2010 issue (Vol. 8, No. 9) has paper acceptance rate of nearly 35%. We wish everyone a successful scientific research year on 2011. Available at http://sites.google.com/site/ijcsis/ IJCSIS Vol. 8, No. 9, December 2010 Edition ISSN 1947-5500 � IJCSIS, USA.