Follow-up on Claims Defaults for Predictive Check-In Systems Bill Benson 05 Dec 1995 22:23 UTC

                                        Date:     05-Dec-1995 02:18pm EDT
                                        From:     Willis O. Benson
                                        Dept:     WL/DORT (Serials Dept.)
                                        Tel No:   (513) 255-6750

Subject: Follow-up on Claims Defaults for Predictive Check-In Systems

To Serials folks who answered my query on predictive check-in systems -

I will "tweak" where necessary, but below is what I finally decided on.  I
have also attached some of the messages I received, and my original message.
There are lots of differences out there, but we are generally all in the same
ball park.

Thanks again,

Bill Benson                  Internet:
Serials Librarian
Wright Lab Tech Library      Fax: (513) 476-4826
2690 C ST  STE 4
Wright Patterson AFB OH 45433-7411

                --- CLAIMS TABLE (in number of days) ---


YA      A        Annual       365        90

MB      BM       Bi-monthly    60        30

WB      BW       Bi-weekly     14        15

YB      --       Biennial     730       180

D       D        Daily        1-3         7

--      IR       Irregular    unk        90

M       M        Monthly       30        30

Q       Q        Quarterly     90        30

YS      SA       Semi-annual  180        90

MS      SM       Semi-monthly  15        15

WS      SW       Semi-weekly  3-4        15

--      TA       Tri-annual   120        60

WT      --       Tri-weekly   2-3        15

MT      --       Tri-monthly   10        15

YT      --       Triennial   1095       180

W       W        Weekly         7        15


--      --       On-order       -        30

--      --       Not Claimed   varies     0



Here at the University of Washington, we are using the serials control
system from Innovative and have very similar "claim after" periods to the
ones you listed.  During conversion, the receipt pattern for each title
was reviewed and the most appropriate standard period default was chosen
for that particular title.  These defaults have helped a great deal to
standardize receipt patterns but we have learned a few things since
conversion as we actually put our claiming module into use.  The
unpredictable nature of serials receipt has been the most frustrating
problem.  But since we can't really control what publishers and mail
services do, we've decided to take a more laid-back attitude about it all.
We soon realized that the defaults were merely guides - accurate for
titles with regular receipt patterns but highly inaccurate for irregular
or problem titles.

About 200 of our check-in records come up for claiming review each week
and we have titles coming in from anywhere in the world.  What the "claim
after" period actually does is to allow us to review the receipt pattern
for titles that are actually late or that don't fit the pattern we
predicted.  This doesn't mean we automatically claim each issue that is
late.  Our first question is: does this issue really need to be claimed?
Oftentimes it does not because the publication schedule has become out of
sync with what we predicted.  At that point we can change the default or
the next expected date to reflect the new reality.  We are now in this
process of fine-tuning our prediction dates and accounts for some number
of the overall so-called "late" issues we must review.  This process will
go on indefinately as publication schedules for serials is constantly
changing.  But it doesn't mean we have to re-evaluate and change default
patterns for every title - only the ones that we didn't predict well in
the first place, or that have changed since we assigned its current
default "claim after" period.  Out of those 200 I look at each week, I
probably now have 15-20%, or about 30-40 records, which need this
adjustment.  Since we have a little over 135,000 check-in records in
total, I see this as a small number.

I look at the "claim after" default periods as actually a chance to
monitor the "health" of our serial receipts and to update or correct our
prediction patterns.  Perhaps this is most apparent for titles which are
highly irregular, like volumes issued in monographic series.  Since they
are published out of order and we can't predict when the next volume will
be received, or even which volume it will be, we actually set a pattern
based on the question, "When do we want to review this record again to see
if there is a problem with our order?", etc.

So again, the default periods are helpful but you will need to fine-tune
as you go along.  And in reality, they are really just a way for you to
monitor serial receipts, rather than an accurate prediction for when the
next issue needs to be claimed.

I hope this helps and if you want more information or discussion, please
don't hesitate to contact me.  Good luck!

Carol Christiansen
Serials Division
University of Washington Libraries
Seattle, Washington  98195

(206) 543-1853
fax: (206) 685-8743


>From the University of Cincinnati:

Our system runs on a 360 day year.  Using that, I have set up the
following defaults for serials claiming:
        Daily - 7 days
        Weekly - 7 days
        Semi-Monthly - 15 days
        Monthly - 30 days
        Bi-Monthly - 30 days
        Quarterly - 45 days

Any other time period is 1/2 the arrival (ie: yearly is 180, twice
yearly is 90).

Some journal records have been adjusted according to vendor
information or past history.  Issues are claimed immediately if the next
issue comes in.

Hope this helps.

Carol Feustel                           email: carol.feustel@UC.EDU
Serials Specialist                      phone: 513-558-0179
Medical Center Libraries                fax:   513-558-1709
University of Cincinnati
PO Box 67057
4231 Bethesda Ave.
Cincinnati, OH 45267-0574

Hello Bill, here at Wheelock we are using the DRA serials subsystem
(since 1992)

The claims were set up for all frequency as  1st claims = 30dys.

2nd claims = 45dys. (this is 15 days after the 1st)

and 3rd claims = 60 dys(15 days after 2nd  and 30 dys after 1st.

We run our claims report only once each month and it usually is around the 25th
of the month.  In doing so only items that should have been rec'd before the
25th of the last month will be picked up on a 1st claim. 2nd claims are items
that were in the 1st catergory on the 25th of the previous month that still
hasn't been rec'd to date. 3rd claims are item that haven't been rec'd since
the last 2 reports were run.  This may not seem like enough time to get results
but this is better in the case where claims must be received within 3 months to
be honored. You will have proof that 3 claims were sent within 2 months and no
results(wheen there isn't any results) and the publishers have to honor your
claim because you did in fact process claims with their time constraints.
Good luck, and I hope this will be of some help to you.
bye for now, Carla.


------------------------------------ BOTTOM ------------------------------

Date: Fri, 01 Dec 1995 14:02:50 -0500
From: Judy Cerqua <jcerqua@SLONET.OHIO.GOV>
Subject: Re: Predictive Claim Intervals (Bill Benson)


At the State Library of Ohio we use the following claim parameters in our
INNOPAC system:

    Daily -      claim after 3 days
    Weekly -     claim after 7 days
    bi-monthly - claim after 7 days
    monthly -    claim after 15 days
    quarterly -  calim after 30 days
    semi-annuals/annual - claim after 60 days past due
    all other frequencies - claim after 90 days past due.

Hope this helps.

Judith Cerqua
Head, Acquisitions/Processing
State Library of Ohio
# 5
I have found that if the claim cycles and the default claim cycles
are too short, there is a problem with past dues appearing on the
Claims Review too frequently--that is before issue could possible
be received due to mail delivery slow downs around holidays, weekends
or after claim is made, before claims could possible be processed by
publisher or vendor.  Through trial and error, I have found that these
claim cycles work well.

Regular Claim cycle:

Daily        7 days
Weekly      14 days
Monthly     30 days  (with slow monthly 60)
Biweekly    30 days
Bimonthly   60 days  (with slow bimonthly 90)
Quarterly   90 days
Semiannual  180 days
Annual      365 days  (or 2 year for slow)

Default Claim cycle:

Daily       30 days
Weekly      30 days
Monthly     60 days
BiWeekly    60 days
Quarterly   180 days
Semiannual  365 days

Mary Jane Wepasnick                   Phone: 716-375-2345
Periodicals Assistant                 Email:
Friedsam Memorial Library
St. Bonaventure, NY l4778

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     29-Nov-1995 11:48am EDT
                                        From:     Willis O. Benson
                                        Dept:     WL/DORT (Serials Dept.)
                                        Tel No:   (513) 255-6750

Subject: Predictive CLAIMS period defaults

To Serialsts using predictive check-in systems,


I am in the process of trying to create serials check-in records, and would
appreciate your input.  Specifically, I would like to know what "Claim After"
defaults you are using successfully.  Perhaps you will also be interested in
the claim default values listed below.

As you know, the defaults we use for our "manual" systems, are not necessarily
the best for a predictive system.  However, I have been unable to find much in
the way of guidance on this important issue.  It is important because I do not
want to have to readjust all the records, if I find out that I have chosen the
wrong core default numbers.

The table below lists the default values (days) that the University of Kansas
uses with their UKASE System, an internally developed, PREDICTIVE serials
system.  It is the best that I have found so far, and I would like to know how
it compares to what you use.  I am thinking about using the "Foreign"
subscription defaults for all my records.  Opinions?

I will be glad to share your thoughts, and the results of this informal survey
with the List membership.  (We have Horizon, but any predictive system should
have the same problems and fixes - I hope!)

Please e-mail or fax your responses directly to me.  Thank you,

Bill Benson                  Internet:
Serials Librarian
Wright Lab Tech Library      Fax: (513) 476-4826
2690 C ST  STE 4
Wright Patterson AFB OH 45433-7411

                --- CLAIMS TABLE (in number of days) ---

        OUR                             CLAIM DEFAULTS
HORIZ.  LOCAL    PERIODICAL   DUE       U.S.   Foreign
CODE    CODE     FREQUENCY   EVERY      sub.*    sub.*

YA      A        Annual       365        60      90

MB      BM       Bi-monthly    60        15      30

WB      BW       Bi-weekly     14         7      14

YB      --       Biennial     730       120     180

D       D        Daily        1-3         7      14

--      IR       Irregular    unk        90     120

M       M        Monthly       30        21      42

Q       Q        Quarterly     90        30      60

YS      SA       Semi-annual  180        60      90

MS      SM       Semi-monthly  15         7      14

WS      SW       Semi-weekly  3-4         7      14

--      TA       Tri-annual   120        30      60

WT      --       Tri-weekly   2-3         7      14

MT      --       Tri-monthly   10         7      14

YT      --       Triennial   1095       120     180

W       W        Weekly         7         7      14


--      --       On-order       -        30      30

--      --       Not Claimed   varies     0       0

* University of Kansas defaults