Creation of Software Focusing on Patent Analysis

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 03 | Mar -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1641
Creation of Software focusing on Patent Analysis
Kandare Dipak Prabhu, Bhosale Sayali Ravindra, Dhande Priyanka Umesh, Shaikh Nida Zahid
Kandare Dipak Prabhu BE Student, PUNE UNIVERSITY, GESCOE, Nashik, Maharashtra, India.
Bhosale Sayali Ravindra BE Student, PUNE UNIVERSITY, GESCOE , Nashik, Maharashtra, India.
Dhande Priyanka Umesh BE Student, PUNE UNIVERSITY, GESCOE, Nashik, Maharashtra, India.
Shaikh Nida Zahid BE Student, PUNE UNIVERSITY, GESCOE, Nashik, Maharashtra, India.
Prof. C.B.Patil Internal guide GESCOE, Nashik, Maharashtra India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Within the early phases of technology
management processes, patents are often used as a source of
inspiration for new ideas. Patents contain detailed technical
information about a technical problem and the preferred
technical solution. This informationcanbe usedforexampleto
assess the state of the art or as a basis to identify possiblegaps
in a technology field. But often it is a very time consuming
process to analyse the information provided by patents,
because huge amounts of patents have to be considered.
Therefore special text-mining and data mining concept are
used to help extracting the desired information in short time.
Classification is used to classify the problem and its solution.
Our approach to make an effective Pre-Processing steps to
save both space and time requirements by using improved
Stemming Algorithm. Stemming algorithms are used to
transform the words in texts into theirgrammaticalrootform.
Key Words: Extraction, Stemming, StopWordRemoval.
1.INTRODUCTION
Patent documents contain important research results that
are valuable to the industry, business, law, and
policy-making communities. If carefully analysed, they can
show technological details and relations, reveal business
trends, inspire novel industrial solutions, or help make
investment policy (Campbell, 1983;Jung, 2003)[2].Inrecent
years, patent analysis had been recognized as an important
task at the government level in some Asian countries.
1.1 A typical patent analysis scenario
1. Task identification: define the scope, concepts, and
purposes for the analysis task
2. Searching: iteratively search, filter, and download
related patents
3. Segmentation: segment, clean, and normalize structured
and unstructured parts
4. Abstracting: analyse the patent content to summarize
their claims, topics, functions, or technologies
5. Clustering: group or classify analysed patents based on
some extracted attributes
6. Visualization: create technology-effect matrices or topic
maps
7. Interpretation: predict technology or business trends
and relations[4].
2. PROBLEM STATEMENT
To analyse and register patent through software by using
stemming and classification algorithm which were earlier
register after checking problem statement and solution
manually.
2.1 A GENERAL METHODOLOGY
Patent analyses based on structured information such as
filing dates, assignees, or citations have been the
major approaches. These structured data canbeanalysed by
bibliometric methods, data mining
techniques, or well-established database management tools
such as OLAP (On-Line Analytical Processing)
modules[1].
Therefore, based on the patent analysis scenario introduced
above, a text mining methodology specialized for full-text
patent analysis is proposed.
This may involve a repeated process of devising a set of
query terms (query formulation), searching a couple of
patent databases (collection selection), filtering undesired
patents (relevance judgment), and downloading patents for
local analysis (data crawling). Depending on the analysis
purpose, the step can be as easy as, for example, fetching all
the patents under some IPC (International Patent
Classification) categories[2] .
The general text mining methodology for patent analysis.
o Document Pre-processing
- Collection Creation
- Document Parsing and Segmentation
- Text Summarization
- Document Surrogate Selection
o Indexing
- Keyword/Phrase Extraction
- Morphological Analysis
- Stopword Filtering
- Term Association and Clustering
o Topic Clustering

- Term Selection
- Document Clustering/Categorization
- Cluster Title Generation
- Category Mapping
o Topic Mapping
- Trend Map
- Query Map
- Aggregation Map
- Zooming Map
3. TECNICH DETAILS
3.1 Extraction
This method is used to tokenize the file content into
individual word.
3.2 Stemming
This method is used to find out the root/stem of a word. For
example, the words user, users, used, using all can be
stemmed to the word “USE”. The purpose of this method is
to remove various suffixes, to reduce number of words, to
have exactly matching stems, to save memory space and
time. The stemming process is done using various
algorithms. Most popularly used algorithm is “M.F. Porters
Algorithm[5].
3.3 Stop word removal
Most frequently used words in English are useless in Text
mining. Such words are called Stop words. Stop words are
language specific functional words which carry no
information. It may be of the following types such as
pronouns, prepositions, conjunctions. Our system uses the
SMART stop word list[5].
4. PROPOSED SYSTEM
In this paper we aim to analyse and register patent through
software by using stemming and classification algorithm
which were earlier register after checking problem
statement and solution manually. The patent databases
world-wide grow continuously,
there is a growing need for software solutions assisting the
user to handle the patent analyses, because the analysis of
hundreds of patents is very complexandtimeconsuming. To
deploy methods, we have proposed a new architecture for
identifying problem and solution of particular patent.
Our system involves following steps:
During our work with patents it occurred that there are
many phrases like e.g. What is claimed is: or A method
comprising . that are very frequently used in patents[1]. In
addition, patent documents of various countries are
generally structured in a similar way: they all provide an
abstract, the claim section, a description of the invention as
well as figures 1. This similar structure makes it easier to
quickly identify the elements that are of interest for the
various patent analysis reasons.
Fig. 1: Core element of Analysis: extraction and analysis of
problems and solutions
To extract problems and solutions because in the
majority of cases not only the solutions (the invention) are
described in this part but also the problems (why the
invention was made)[6]. Generally a patent provides more
than one problem as well as more than one solution. But the
description of problems is not always very detailed. In
addition the relation between problems and solutions must
not always be apparent, in some cases there is no close
relationship or even no problem mentioned. This makes an
extraction of any sort of relationships quite difficult.
Objective of the problem and solution extraction
first of all is to retrieve the main claim of a
patent and then trying to identify at least one problem that
refers to the first claim. Some patents actually provide a
short summary of the main problem referring to the main
claim, but most of them don’t[7]. In those cases it could be
possible that the extracted problem refers not to the first
solution (main claim) but to further sub-solutions. Thus it is
mandatory to check the problem retrieving results of
afterwards.

Fig. 2: Analysis of extraction and analysis of problems and
solutions
In the following, problem or solution indicating phrases are
parts of sentences that are surrounding a problem or
solution, i.e. directly before, in between or after those
phrases a problem or solution is described. For example
within the sentence[1].
The Patent Skill Cartridge is able to extract
problems and solutions by searching for those indicating
phrases in the text and then displaying the sentence parts
before or after the phrases in a pre-defined length. So within
the Patent Skill Cartridge not only the problem and solution
indicating phrases are defined, but also where the problem
or solution text can be found, that means before, after or in
between the indicating phrases.
For the development of the prototype of the Patent
Skill Cartridge a first set of 57 patentsfrom randomlychosen
technology fields like electric vehicles and women hygiene
articles was selected. Based on these patents problem and
solution indicating key phrases were identified. The result
was a list of over 100 phrases that were implemented in the
Patent Skill Cartridge. A short sectionofthecompletephrase
list is shown in the following Table 1.
For the development of the prototype of the Patent
Skill Cartridge a first set of 57 patentsfrom randomlychosen
technology fields like electric vehicles and women hygiene
articles was selected. Based on these patents problem and
solution indicating key phrases were identified. The result
was a list of over 100 phrases that were implemented in the
Patent Skill Cartridge. A short sectionofthecompletephrase
list is shown in the following Table 1. We found that some
phrases, especially solution indicating phrases, occurred
several times in more than one patent even
though the wording of some phrases was slightlydifferentin
the selected set of patents. For example the most frequent
phrase indicating a solution was it is an object of the
invention / general object of the (see Fig. 2). The challenge
during the implementation of phrases in the Patent Skill
Cartridge was to describe the phrases as universal as
possible in order to also cover slightly different phrases. For
example it was not sufficient to only implementthephraseA
in Fig. 2 object of (the invention) . Because also common
phrases like goal of the invention and aim of the invention
appeared in the patent data set and had to be considered in
order to retrieve the corresponding solutions[2]. In
comparison to the solution key phrases the problem
indicating phrasesweremorediverseandthereforecomplex
to implement in the Patent Skill Cartridge. In only few cases
very clear phrases like The present invention was made to
solve were found. Instead a wide variety of phrases like a
disadvantage of / a drawback associated with or therefore
there is a need for or None of the prior attempts... were
found and had to be implemented.

TABLE 1: EXCERPT OF LIST OF PROBLEM AND SOLUTION
INDICATING PHRASES
Conclusion
Analysis of patents can be done using stemming
algorithm andrequirementofspecial patentanalysismethod
like white spot analysis can be met. With the developed
patent skill , It is possible to automatically identify text
element. Like problems or solution in patents and retrieve
them. There is a need for more problem and solution
indicating phrases. As for the development of the cartridge
only very few patents were considered (in comparison with
the patent data that is already available world-wide) this
result was expected. In addition the Patent Skill Cartidge
should provide a systematic approach on how to
differentiate between main problems, solutions, sub-
problems, and sub-solutions. Concluding, the results of
problems and solutions are currently not clustered or
classified in a technology specific way. Especially if a lot of
patents are analysed it is often advantageous to cluster the
results e.g. by the use of a technology specific ontology and
therefore minimize the patent map of problems and
solutions in order to support the expert to work more
efficiently.
REFERENCES
[1] Yvonne Siwczyk, Joachim Warschat, Dieter Spath
Software-based Patent Analysis: How to Leverage a
Text-mining Tool December 2014.
[2] Yuen-Hsien Tseng , Chi-Jen Lin , Yu-I Lin Text mining
techniques for patent analysis 26 January 2007
[3] Nizar Ghoula, Khaled Khelif and Rose Dieng-Kuntz
Supporting Patent Mining by using Ontology-based
Semantic Annotations 2007.
[4] Khaled Khelif, Aroua Hedhili and Martine Collard
Semantic Patent Clustering forBiomedical Communities
2008.
[5] C.Ramasubramanian1, R.Ramya Effective Pre-
Processing Activities in Text Mining using Improved
Porter’s Stemming Algorithm 2013.
[6] Peter Anick, Marc Verhagen and James Pustejovsky
Extracting Aspects and Polarity from Patents 2014.
[7] Hei Chia Wang Patent Threat Analysis Search Engine
2015.
Patents Problems
indicating
phrases
Solutions
indicating
phrases
US2006042846
Backgroundtothe
Invention
Main challenges
for
Summery for the
Invention
The present
invention
overcomes the
aforentioned
drawbacks for
provindings
US2001568676
Discussion of the
prior art
While this idea is
known in the
prior art… do not
utilize the full
potential of
Field of invention
The present
inventions
relates generally
to… and specially
EP1067876
Backgroundtothe
invention
In general… is
limited by there
arises a problems
that…
Summery of
invention
It is therefore an
object of the
presentinvention
to provide
US2008316755
Backgroundtothe
invention
Thus it may be
difficult to
Summery of the
invention
The present
invention has
been made in
consideration of
the foregoing.
US2002725671
Backgroundtothe
invention
Common to all…
therefore need …
In order to
maximize the
efficiency
therefore there is
a need for
Summery of the
invention
In one
exemeplary
embodiment … is
disclosed

Creation of Software Focusing on Patent Analysis

More Related Content

What's hot (10)

Similar to Creation of Software Focusing on Patent Analysis (20)

More from IRJET Journal (20)

Recently uploaded (20)

Creation of Software Focusing on Patent Analysis