Bibliographic record
Abstract
Interview with Sean Reese Q: Can you tell us something about Ocean Spray? A: Ocean Spray Cranberries is a growerowned farming cooperative with $1.5 billion in annual sales, headquartered in Lakeville, MA. The ownership is comprised of roughly 700 cranberry and 200 grapefruit growers. The company manufactures a variety of juices and fruit-based products in four primary facilities located in: Bordentown, NJ; Kenosha, WI; Henderson, NV; and Sulphur Springs, TX. Q: How large is the Ocean Spray forecasting group? A: There are seven people who work in the forecasting group. The official name of the group is Demand Planning, and it represents Ocean Spray's commitment to and investment in the forecasting field/practice. Q: What are the forecasting organization's titles and positions? A: Ocean Spray's group is composed of a Manager of Demand Planning, five Demand Planners, and one System Administrator. Q: How do they relate to the planning function? A: Demand Planning directly supports Operations. The emphasis of our work is demand-oriented, by product and by geographic area. Its primary concern is to support short-term production planning and short- to mid-term management decision-making. Q: Whom does the group report to? A: The Demand Planning Group reports to the Director of Logistics and Planning, who reports to the Vice President of Operations. Q: What type of forecasts do you make? A: We forecast at both the SKU (UPC) level and the Category/Size level. For example, 64-oz. CranApple(R) cranberry apple juice drink would be a SKU, and would be a subset of the larger 64-oz. Cranberry Drinks category/size grouping. We forecast on a disaggregated level, using cases as our base unit. As an example of the case unit, eight 64-oz. bottles of Cranberry Juice Cocktail represent one case. These case units are linked to retail accounts that correlate with Ocean Spray's sales organization. Each account is in turn linked to one of our four distribution centers around the country. We forecast the current month and the following six months, with the greatest attention on the next three months. Q: What software do you use? A: We primarily utilize Manugistics in our forecasting work. Q: Do you rely on any other MIS-type systems or support? A: On the front end, we get data feeds from SAP (an Enterprise Resource Planning system) and a proprietary intermediate system, which we call BIS. On the back end, we use Oracle, Microsoft Access, and Microsoft Excel for analysis and presentation. Q: How much data and which models do you use in forecasting? A: The forecasting models that we use in our Manugistics system are primarily time series models. We keep three years of history in that system, on which to base the statistical forecasts. We, of course, have the latitude to override the time series with event-based inputs such as promotional plans, advertising and the effects of product reformulations. We use time series forecasts as a starting point from which we query the field sales personnel as well as our Marketing team for additional market intelligence. As such, our modeling process is a part of an intensive collaborative process with the Sales and Marketing team, a process that I suppose you could call a judgmental approach. Causal and Regression-based models are deployed sparingly, mainly in support of upper-- management's strategic planning needs. Q: Why are Causal models used sparingly? A: At the disaggregated level, for which we are responsible, it is difficult to find qualified explanatory variables. Q: Do you use scanning data? If so, what types of data do you use and how do you use it? A: Yes, we do. For our domestic accounts we use IRI, and for Canada we use Neilsen data. This scanning data is used heavily by our Marketing department as they try to ascertain changes in our base volume and consumer reaction to promotions and advertising. …
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".