Вы находитесь на странице: 1из 2

Search System Requirements of Patent Analysts

Leif Azzopardi, Wim Vanderbauwhede


Department of Computing Science
University of Glasgow, United Kingdom
{leif,wim}@dcs.gla.ac.uk
Hideo Joho
Graduate School of Library,
Information and Media Studies
University of Tsukuba, Japan
hideo@slis.tsukuba.ac.jp
ABSTRACT
Patent search tasks are dicult and challenging, often re-
quiring expert patent analysts to spend hours, even days,
sourcing relevant information. To aid them in this pro-
cess, analysts use Information Retrieval systems and tools
to cope with their retrieval tasks. With the growing interest
in patent search, it is important to determine their require-
ments and expectations of the tools and systems that they
employ. In this poster, we report a subset of the ndings of
a survey of patent analysts conducted to elicit their search
requirements.
Categories and Subject Descriptors: H.3.3 Information
Storage and Retrieval: Information Search and Retrieval
General Terms: Human Factors
Keywords: User Study, Patent Engineers, Patent Analysts
1. INTRODUCTION
With new initiatives providing patent test collections and
a spate of workshops and symposiums on patent retrieval
1
,
there has been renewed interest in researching and develop-
ing Information Retrieval (IR) tools, techniques and theory
for patent search. Patent analysts perform a number of di-
cult and challenging search tasks (such as Novelty search or
Infringement search) [2] and rely upon sophisticated search
functionality, tools, and specialised products [1]. These search
tasks are often performed under stringent conditions (esp.
regulatory and legal requirements) [2], and they also require
dierent search strategies to achieve the end goal (which
in some cases means not actually nding documents, i.e. no
killdocument, for instance) [3]. Whilst there has been sub-
stantial research on patent search and the tasks and tools
involved, little work has been performed investigating the
requirements of patent searchers, and what they want. It is
vital that users are consulted and their needs understood.
This is to ensure that the ongoing research and develop-
ment meets their current requirements, identies new re-
quirements, highlights areas of potential opportunity (i.e.
what advancements are needed to improve patent search
techniques) and identies constraints. To this aim, we sur-
veyed over eighty patent analysts in order to obtain a better
picture of their search habits, preferences, and the types of
1
New initiatives include CLEF-IP, TREC-CHEM, MAREC, PaIR,
etc. while past initiatives have been run at NTCIR and TREC.
Copyright is held by the author/owner(s).
SIGIR10, July 1923, 2010, Geneva, Switzerland.
ACM 978-1-60558-896-4/10/07.
functionality that they want/need, along with how they go
about accomplishing their search tasks. In this poster paper,
we present a subset of the ndings from this survey, where
we focus on reporting the results about the search tasks and
the functionality that patent searchers desire and require.
2. METHOD AND RESULTS
The survey instrument
2
that we used consisted of an on-
line questionnaire with 86 questions consisting of four parts:
(i) demographics and expertise (ii) search tasks (iii) search
functionality and (iv) open ended questions on search system
requirements. The survey was designed to be completed in
approximately 30 minutes.
To obtain a large and representative sample we sent the
survey instrument out to two patent user group mailing lists:
(i) the Confederacy of European Patent Information User
Groups (CEPIUG) and (ii) the international Patent Infor-
mation Users Group (PIUG). In total, these lists have over
700 members from over 27 dierent countries, and of these
members, approximately 300 are patent information profes-
sional who regularly perform patent searching. We received
81 responses in total to the survey a response rate of just
over 10% overall. Of the 81 respondents, 58% were male,
while 42% were female. Respondents were from 14 dier-
ent countries including France, Netherlands, UK, US, other
european countries and China. The majority were aged be-
tween 39-52 (44%) and had about 10 years experience in
patent searching. Almost all respondents primarily searched
in English (98%), with a few also using German, French and
Dutch (less than 15% in total).
In this poster, we report only on a subset of questions
regarding search task frequency and querying functionality
requirements. These represent the current demands and re-
quirements that analysts have when using an IR system. It
is important to appreciate these needs to ensure that re-
search is relevant and targeted to meet them. Firstly, it
should be noted that respondents spend approximately 17
hours searching (i.e. searching takes place over several days),
and they submit about 27 queries per search task on aver-
age. The amount of time they spent formulating queries
was about 18 minutes per query. This shows that analysts
are willing to spend a lot of time and eort searching and
formulating their needs. Table 1 presents the results from
a series of questions regarding how often an analyst per-
forms the dierent types of search tasks (a description of
the dierent search tasks is provided in the table, see [2] for
2
Ethics approval from the University of Glasgow was obtained to
conduct this survey (Reference code: ETHICS-FIMS00638).
Search Task Description Often Sometimes Rarely
State of the Art identify patents for the purposes of a general review 22.2 33.3 44.5
(aka landscaping)
Novelty identify patents and non-patents which may aect the 37.0 40.7 22.3
patentability of an idea/invention
(performed before writing a patent application)
Patentability given a patent application, ensure novelty 34.6 39.5 25.9
Infringement identify patents or applications which cover the 21.0 23.5 55.5
proposed product or process and are still in force
Opposition identify literature available to the public to show lack 13.6 17.3 69.1
of novelty or inventive step of a granted patent
Freedom to Operate like infringement, but also includes non-patent literature 46.9 25.7 28.4
Due Diligence analyze strengths, weaknesses and scope of IP rights. 16.1 19.8 64.2
Other 27.2 27.2 45.7
Table 1: How often search tasks are undertaken by respondents (values shown are percentages).
more details). The category headings are: often (daily or
weekly), sometimes (monthly), and rarely (once or twice a
year, or never) and the value is the percentage of respon-
dents. The results indicate that Novelty, Patentability and
Freedom To Operate searches are the most frequent types
of search tasks which are routinely performed, given our re-
spondents. It is notable that these tasks require the analyst
to search both patents and non-patents collections. By con-
trast, Infringement, Opposition and Due Diligence search
tasks were rarely or never performed by 55-69% of the re-
spondents. These ndings suggest that a lot of eort is
focused on ensuring that infringements or challenges are not
made by ensuring that patent applications are novel and
patentable while products are appropriately licensed.
For any given search task they perform, analysts were
asked about the importance of a variety of operations they
use when querying the retrieval system. Table 2 presents
a summary of these results. Participants were asked to rate
statements such as Boolean operators are important to for-
mulate eective querieson a ve point scale (Strongly Agree
to Strongly Disagree). Here, we have compressed the scale to
Disagree, Neutral and Agree. The results show that Boolean
operators were very important to almost all respondents.
Proximity, Truncation, Wildcards and Field Operators were
important to most respondents (80-91%), while Expansion
and Translation were seen as important by around half the
respondents with a large proportion of respondents impar-
tial to these features. This may be because they have not
used such functionality frequently enough to form an opin-
ion, or because such functionality is not required as most
respondents searched primarily in English.
Finally, the Weighting of terms in the query obtained a
very mixed response; where most respondents are impar-
tial, while the rest were split between important and not
important. This may be because of the diculties associ-
ated with weighting query terms manually. It appears that
the features that introduce some uncertainty in the process
(i.e. Weighting, Expansion and Translation) are not consid-
ered to be as important as the other very precise operators,
which can be controlled and have a clear interpretation. This
is perhaps due to the fact that analysts are often required
to full strict and stringent practices given the legal and
regulatory requirements [2].
These ndings suggest that research and development of
models, methods and systems for patent search need to
consider the user requirements identied as important. It
would appear that patent searchers prefer search function-
ality which provides a high degree of control and precision
for accomplishing their search tasks, and they are willing to
spend a lot of time and eort in constructing requests and
examining documents.
Question Disagr. Neut. Agree
Boolean operators 2.5 1.2 96.3
Proximity, Adjacency 4.9 11.1 84.0
or Distance operators
Weighting 22.2 48.2 29.6
Truncation (left/right) 1.2 7.4 91.4
Wildcards 6.2 13.6 80.3
Field Operators 3.7 11.1 85.2
Query Expansion 2.5 42.0 55.6
Query Translation 11.1 43.2 45.7
Table 2: Survey Questions: is the feature important
to formulate eective queries, and responses.
3. SUMMARY AND FUTURE WORK
In this poster, we have provided empirical evidence which
suggests that the requirements of patent analysts are quite
dierent from those assumed in standard retrieval tasks (like
web search, etc). In particular, our results indicate a strong
preference towards functionality that gives ne grained con-
trol over the search process using operators with a clear se-
mantic and precise interpretation. These ndings motivate
further research into understanding more deeply the impor-
tance of each type of functionality given the dierent patent
search tasks along with ascertaining a better understanding
of the context in which patent search is undertaken, in order
to meet searcher needs and requirements.
Acknowledgments: We would like to thank the CEPIUG
and PIUG user groups and the participants who took part
in the survey. This work was supported by the Information
Retrieval Facility (http://www.ir-facility.org).
4. REFERENCES
[1] K. H. Atkinson. Toward a more rational patent search
paradigm. In Proceeding of PaIR 08:, pages 3740.
ACM CIKM, 2008.
[2] D. Hunt, L. Nguyen, and M. Rodgers. Patent Searching:
Tools and Techniques. John Wiley and Sons, 2007.
[3] Y.-H. Tseng and Y.-J. Wu. A study of search tactics for
patentability search: a case study on patent engineers.
In Proceeding of PaIR 08:, pages 3336. ACM CIKM,
2008.

Вам также может понравиться