Report: Data-Driven Techniques in Human Language Technologies (IS)

This report provides a comprehensive analysis of data-driven techniques within Human Language Technologies (HLT), specifically focusing on the differences between data-driven and knowledge-based approaches. It highlights the importance of data-driven methods in modern HLT systems, contrasting them with traditional knowledge-based approaches. The report explores the advantages and disadvantages of both approaches, including aspects such as expertise distribution, ease of update, and error rates for knowledge-based, and information augmentation, platform-friendliness, and investment in open ecosystems for data-driven approaches. The report also discusses the limitations of each approach, such as the abstract nature of knowledge and the potential for incorrect data in data-driven systems. The findings are related to the student's background in Information Systems, with real-world applications and relevant figures to support the analysis. The conclusion emphasizes the benefits of a data-driven approach in HLT, advocating for its adoption due to its efficiency and effectiveness, while acknowledging the challenges of data extraction and information sharing within knowledge-based systems. The report uses at least ten research papers as references.

Running head: HUMAN LANGUAGE TECHNOLOGIES
HUMAN LANGUAGE TECHNOLOGIES
Name of the Student:
Name of the University:
Author Note:

1HUMAN LANGUAGE TECHNOLOGIES
Executive Summary:
The purpose of the report is to showcase the importance of the data driven technique in
Human Lnguage technology. The report briefly describes the differences between the data
driven approach and knowledge-based approach, it also describes about the advantages and
disadvantages of both types of approach. Lastly, the report concludes with importance of the
data driven technique.

2HUMAN LANGUAGE TECHNOLOGIES
Table of Contents
Introduction:...............................................................................................................................3
Knowledge based approach in the development of Information Systems:................................3
Data-Driven approach of Information system:..........................................................................6
Advantages and Disadvantages of Knowledge based approach:...............................................8
Advantages and Disadvantages of Data-Driven approach:........................................................9
Conclusion:..............................................................................................................................10
References:...............................................................................................................................12

3HUMAN LANGUAGE TECHNOLOGIES
Introduction:
Human Language Technology is the study of the computer programs or the electronic
device about how they analyse, modify, produce and respond to the human’s speech and the
texts. Human Language Technology composed of Natural Language Processing (NLP) and
Computational linguistics (CL). It also requires some application for the operation.
Information Systems are formal, organizational, sociotechnical process designed for the
purpose of gathering, distributing, processing and storing data or information. It is an
educational study of systems with a definite reference to information and complimentary
networks of software and hardware that the people and company use to deal with that data.
Data driven technique is defined as the process which is controlled only by data. The actions
are guided by data and it is an efficient process as it gives a zero-error output and takes zero
guesses. Data driven approach can be of many types like Strategy, Design, Automation,
Communication, Marketing, Artificial Intelligence, performance management, Decisions,
Estimates and so on. The report focuses on data driven approach of Information Systems. The
report briefly describes the difference between the knowledge-based approach and data-
driven approach, advantage of knowledge based and data driven approach of the information
system,
Knowledge based approach in the development of Information Systems:
The main problem of the software engineering is in the region of Information
Systems. The finding of the sufficient methodology, methods as well as the tools is the
lacking part of the region in the Software Engineering (Meyer and Nordio, 2016). The life
cycle of the information system is studied by the knowledge-based approach. An expert
system is created to analyse the life cycle of the information system. The integrated CASE
system is such type of system where the life cycle, tools and modelling process of the

4HUMAN LANGUAGE TECHNOLOGIES
software can be configured (Tzafestas, 2013). The Configuration is based on the Knowledge-
extended project database (KBextPDb), which composed of the database of the software
technologies (Db TSwP), knowledge-based methods and the project database (PDb) which
allow to manually select the appropriate methodology, tools and methods for the
development of software system along with IS properties (Griffin and Care, 2014). The
configured CASE system is used to implement the IS. The software methodologies changes
in a regular way so it requires to use specific suitable CASE device but it is not an efficient
and effective way of implementation because it requires too busy the CASE tool and also
requires to re-educate the staff (Martin and Moodysson, 2013). The solution to this issue is to
utilize the KBinCASE because the software technology is re-usable. The specification of
Software Technology is utilized to configure the in CASE configuration which is then used to
apply on the IS development.
Figure 1: Knowledge based integrated configurable CASE system

5HUMAN LANGUAGE TECHNOLOGIES
(Source: uni-obuda, 2019)
Knowledge-based Approach regarding the Maintenance and usage of Information
Systems
The most important factor in Software Engineering regarding the Information System
is its insufficient flexibility and efficiency of the tools and process that are used to modify the
great software system according to the requirements of the user. Knowledge based approach
deals the problem by creating a software layer that controls, modify and maintain the
information system in the development level. Knowledge based Information System (KB_IS)
deals with the three layers of the architecture of IS (Becerra-Fernandez and Sabherwal,
2014). The knowledge layer has a great impact on the Information System architecture. The
Information system identifies itself by the help of the knowledge layer. Knowledge layer
helps the information system to modify and maintain.
Figure 2: Knowledge based information system architecture
(Source: uni-obuda, 2019)

6HUMAN LANGUAGE TECHNOLOGIES
Separation of Business rules in knowledge layer:
Business rules that are made for the information system remain implicitly in the
system behind the other elements of the system. They are present either in the program code
or they are in the habits or in the methods. The finding of those element is impossible so
many companies are losing those elements. This problem can only be solved by some
Business Rules Approach (Kaula, 2015). This helps to separate the business rules from the
program code and helps to change the rules without requiring any modification in the
program code.
Open design architecture:
The IS design and the maintenance of the system can be simplified by the Open
Design Architecture. The main motive of this type of architecture is to mix the IS design and
IS together. Layers which contain information regarding the structure, relationships,
architecture and other information regarding the project are closely attached with the IS (Ku,
Lu and Gerla, 2014). The information is a part of the CASE system database which is used to
develop the IS system and exist in various forms and can be used in different types of
projects. The architecture defines the knowledge layer with respect to the project information
and requirements of the user that gives the knowledge required for the higher-level
automation of the modification of IS and activates regarding the IS maintenance.
Data-Driven approach of Information system:
The data driven research utilizes the exploratory approaches which analyse the
information system to get insights information. Data driven approach involves some tasks
like: identification of the research question, creating the data sources, Extracting annotating
the information for the preparation of analysis, integrating, representing and aggregating the
data to achieve the insights information and so on (Baak et al., 2013).

7HUMAN LANGUAGE TECHNOLOGIES
Information System framework in two perspectives:
The differences between the theory driven and data driven approach is quite
problematic. An analysis of information analytics without the domain theory cannot give an
appropriate and sufficient information regarding the system and can lead the system to
identify correlations, patterns and trends in some questions, again, an exclusive emphasis of
information system analytics can fail without some practical dat. It is important to approach
any sy6stem with both the approaches. Information system to the domain theory starts the
data driven research that focuses to identify the patterns which represent the concepts. The
patterns can be analyzed in two main ways, first the data that are extracted from the
information system can be utilized to derive the insights regarding the domain theories.
Secondly, it can be used in attempting to interpret the patterns with respect to existing theory
(Ferguson, 2017). Data driven process includes particular tasks like exploratory analytics,
model selection, choice of variables and data preparation. These operations include various
techniques and algorithms. The path from the theory driven to the data driven method
displays that the domain theory can help to search the patterns by searching possible
relationships and constructs that are used in the analysis.

8HUMAN LANGUAGE TECHNOLOGIES
Figure 3: Framework for IS research
(Source: uni-saarland, 2019)
Advantages and Disadvantages of Knowledge based approach:
Some of the advantages of the knowledge-based approach towards the information
system are (Maher, Balachandran and Zhang, 2014):
1. Allow Expertise distribution
2. Easy to update
3. Consistent
4. Expertise preservation
5. Capable of operating with less amount of data
6. Capable to explain the answers
7. Encourages the organizations to verify the logic of decision making
8. Protect intellectual capital
9. Makes the experience of problem-solving reusable

9HUMAN LANGUAGE TECHNOLOGIES
10. Enables the organization to leverage the size
11. Error rate is nearly zero
12. Stimulating innovation and the development
13. Focus on the human capital
Some disadvantages of Knowledge based approach in information system are:
1. High technology image of AI field
2. Abstract nature of knowledge
3. Limitations of Cognitive scientific theory and methods (Vaishnavi and Kuechler,
2015)
4. Availability
5. Difficult to extract data from humans
6. Users have normal cognitive limits
7. Problem in sharing the information by the organization.
8. Extracting data is too difficult
9. Organization failed to keep the respective technologies.
Some countermeasure to be taken to overcome the disadvantages are:
1. Organization should share sufficient amount of information as it is the main part of
knowledge management.
1. Extracting data should be difficult
2. Idea implementation is more valuable than idea generation (Salter et al., 2015)
3. Data should be efficiently re-used, collected and analyzed.
Advantages and Disadvantages of Data-Driven approach:
The data driven approach in Information System helps to (Moro, Cortez and Rita, 2014):

10HUMAN LANGUAGE TECHNOLOGIES
1. Focus on the Information Augmentation (IA) to manage the data: Information system is
a huge asset for any type of industry, the organization gets reliable and relevant speed
and efficiency from the system. The augmentation of the information is the logical step
to organize the data to reconcile the data. Organizing data with respect to data type with
collaboration and contribution on limitless variables describes the continuous IA.
2. Platform friendly: Data capital needs new computing infrastructure and clear
understanding of generating applications which analyse the information and utilize the
data effectively. Data management allow clean and reliable data which is connected and
develop trusted foundation. The total Cost of Ownership calculates savings realized
from the switching to cloud versus the on-premises are evident.
3. Invest in an open ecosystem of technologies: The data driven approach of information
system becomes an indispensable feature in recent times. Investment in Information
System remain competitive and profitable for every organization. The open ecosystem
allows to choose the technology and domain expertise. The flexibility of the information
system should be maintained.
4. Data driven applications should be built with embedded analytics: Enterprise data
driven applications are made to infuse information from multiple sources of the
enterprise and relate all the related transactions associated with the product of the
information system. The different departments of an organization like marketing, sales,
customer care operates the Information system to take decision faster.
5. Make the algorithms, Data and Operations continuous: Self-learning information
organizations make the data first reliable, and connect with the rest of the enterprise, so
that it can be trusted for any purpose in critical business operations. ML play an
important role in the development of any type of business which helps to make sense

11HUMAN LANGUAGE TECHNOLOGIES
the importance of data and helps to take decisions faster which helps the organization to
act as a self-learning organization.
Some disadvantages of data driven approach of Information system are (Todorov et al.,
2013):
1. Without any type of skepticism, data are trusted blindly.
1. Data are incorrect and often messy. The quality of the data is low.
2. The proliferation of data helps the business users to analyses the data.
3. The quality of the data is low so the table, chart, algorithm, made based on that error
data makes the organization to take wrong decision which greatly impacts the
organization and can affect the development in the organization.
Conclusion:
Therefore, from the report it can be conclude that data driven techniques with respect
to Human Language Technology is a big priority in this world. The knowledge-based
approach and the data driven approach both are clearly described in the report and it is
suggested to evolve the information system with a data driven approach for every
organization as its error rate is nearly zero and is user friend and both efficient and effective
for the business. In knowledge-based approach extracting data is difficult and the approach
involves quite difficulties in sharing the data around any organization. The data driven
approach helps to share data efficiently and effectively.

12HUMAN LANGUAGE TECHNOLOGIES
References:
Baak, A., Müller, M., Bharaj, G., Seidel, H. P., & Theobalt, C. (2013). A data-driven
approach for real-time full body pose reconstruction from a depth camera. In
Consumer Depth Cameras for Computer Vision (pp. 71-98). Springer, London.
Becerra-Fernandez, I., & Sabherwal, R. (2014). Knowledge management: Systems and
processes. Routledge.
Ferguson, T. S. (2017). A course in large sample theory. Routledge.
Griffin, P., & Care, E. (Eds.). (2014). Assessment and teaching of 21st century skills:
Methods and approach. Springer.
Kaula, R. (2015). Business intelligence rationalization: a business rules approach.
International Journal of Information, Business and Management, 7(1), 129.
Ku, I., Lu, Y., & Gerla, M. (2014, August). Software-defined mobile cloud: Architecture,
services and use cases. In 2014 international wireless communications and mobile
computing conference (IWCMC) (pp. 1-6). IEEE.
Maher, M. L., Balachandran, M. B., & Zhang, D. M. (2014). Case-based reasoning in design.
Psychology Press.
Martin, R., & Moodysson, J. (2013). Comparing knowledge bases: on the geography and
organization of knowledge sourcing in the regional innovation system of Scania,
Sweden. European Urban and Regional Studies, 20(2), 170-187.
Meyer, B., & Nordio, M. (Eds.). (2016). Software Engineering: International Summer
Schools, LASER 2013-2014, Elba, Italy, Revised Tutorial Lectures (Vol. 8987).
Springer.

13HUMAN LANGUAGE TECHNOLOGIES
Moro, S., Cortez, P., & Rita, P. (2014). A data-driven approach to predict the success of bank
telemarketing. Decision Support Systems, 62, 22-31.
Salter, A., Ter Wal, A. L., Criscuolo, P., & Alexy, O. (2015). Open for ideation: Individual‐
level openness and idea generation in R&D. Journal of Product Innovation
Management, 32(4), 488-504.
Todorov, A., Dotsch, R., Porter, J. M., Oosterhof, N. N., & Falvello, V. B. (2013). Validation
of data-driven computational models of social perception of faces. Emotion, 13(4),
724.
Tzafestas, S. G. (Ed.). (2013). Knowledge-based system diagnosis, supervision, and control.
Springer Science & Business Media.
uni-obuda. (2019). Microsoft Word - 49_Plocica.rtf [Ebook] (p. 6).
uni-saarland. (2019). JAIS Author Template [Ebook] (p. 21). uni-saarland.
Vaishnavi, V. K., & Kuechler, W. (2015). Design science research methods and patterns:
innovating information and communication technology. Crc Press.