File size: 12,408 Bytes
7022e9f
 
56e9ca8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8e92d06
 
 
 
56e9ca8
 
 
 
 
 
 
 
 
 
 
 
7022e9f
56e9ca8
8e92d06
 
 
 
 
 
 
 
 
 
 
 
 
 
 
56e9ca8
8e92d06
 
 
 
 
56e9ca8
8e92d06
56e9ca8
8e92d06
 
 
 
 
 
 
 
 
 
 
 
56e9ca8
8e92d06
56e9ca8
8e92d06
56e9ca8
8e92d06
56e9ca8
8e92d06
 
 
 
 
 
56e9ca8
8e92d06
56e9ca8
8e92d06
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
56e9ca8
 
 
 
 
 
8e92d06
 
56e9ca8
8e92d06
 
 
 
 
56e9ca8
8e92d06
56e9ca8
 
8e92d06
 
 
 
 
56e9ca8
 
8e92d06
56e9ca8
8e92d06
56e9ca8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8e92d06
56e9ca8
8e92d06
56e9ca8
8e92d06
56e9ca8
 
 
 
 
 
 
 
 
 
 
 
 
8e92d06
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
56e9ca8
8e92d06
56e9ca8
8e92d06
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
---
license: cc-by-nc-4.0
language:
- en
pretty_name: MemoryVCD — URL edition
size_categories:
- 100K<n<1M
task_categories:
- text-generation
tags:
- amazon-reviews
- recommendation
- memory
- multimodal
- evaluation
configs:
- config_name: interactions
  data_files:
  - split: test
    path: tables/interactions/*.parquet
  default: true
- config_name: products
  data_files:
  - split: test
    path: tables/products/*.parquet
- config_name: negative_items
  data_files:
  - split: test
    path: tables/negative_items/*.parquet
- config_name: users
  data_files:
  - split: test
    path: tables/users/*.parquet
- config_name: evaluation_windows
  data_files:
  - split: test
    path: tables/evaluation_windows/*.parquet
- config_name: evaluation_candidates
  data_files:
  - split: test
    path: tables/evaluation_candidates/*.parquet
---

# 🧠 MemoryVCD · URL Edition

**323 users · 12 domains · 535,698 interactions**

[📄 Paper](https://aclanthology.org/2026.findings-emnlp.1804/) · [📥 PDF](https://aclanthology.org/2026.findings-emnlp.1804.pdf) · [💻 MemoryCD Code](https://github.com/AgentMemoryWorld/MemoryCD) · [🔗 URL Edition](https://huggingface.co/datasets/WZDavid/MemoryVCD) · [🖼️ Full Edition](https://huggingface.co/datasets/WZDavid/MemoryVCD-Full)

Real user histories for evaluating long-term, cross-domain personalization. MemoryVCD extends the MemoryCD data release with product image references, complete metadata for the saved candidate pools, and explicit evaluation inputs. Full adds the available cached product images.

## 📄 Paper and release relationship

**MemoryCD: Benchmarking Long-Context User Memory of LLM Agents for Lifelong Cross-Domain Personalization** — Zhang et al., *Findings of EMNLP 2026*.

The paper studies memory grounded in real Amazon review histories, with four tasks: rating prediction, item ranking, review summarization, and review generation. It evaluates single-domain and cross-domain memory. See the [paper](https://aclanthology.org/2026.findings-emnlp.1804/) for methods and results.

This repository is the **MemoryVCD extended release**, not a claim that every added visual feature or candidate file was evaluated in the MemoryCD paper. Statistics below describe the released files.

| Repository | Purpose | Product images |
|---|---|---|
| [MemoryCD](https://huggingface.co/datasets/WZDavid/MemoryCD) | Original release; unchanged | Original item metadata |
| [MemoryVCD](https://huggingface.co/datasets/WZDavid/MemoryVCD) | Extended histories, candidate metadata and evaluation inputs | URL strings |
| [MemoryVCD-Full](https://huggingface.co/datasets/WZDavid/MemoryVCD-Full) | Same data plus existing image cache | 243,394 embedded image records |

## 📊 At a glance

| Statistic | Count |
|---|---:|
| Users | 323 |
| Domains | 12: 4 evaluation targets + 8 memory domains |
| Interaction records | 535,698 |
| History product/domain pairs | 276,418 |
| All product/domain metadata records | 481,333 |
| Unique sampled negative product/domain pairs | 211,364 |
| Negative items outside the history product set | 204,915 |
| Missing metadata for saved negatives | **0** |
| Product image URLs | 456,761 |
| Valid cached image records (Full) | 243,394 |

Counts use `(domain, parent_asin)` as the item key. A negative for one user can be a historical item for another; history and negative counts therefore overlap.

## 🗂️ Browse the data

| Subset | Rows | Contents |
|---|---:|---|
| [interactions](https://huggingface.co/datasets/WZDavid/MemoryVCD/viewer/interactions/test) | 535,698 | One review per row |
| [products](https://huggingface.co/datasets/WZDavid/MemoryVCD/viewer/products/test) | 481,333 | History and candidate item metadata |
| [negative_items](https://huggingface.co/datasets/WZDavid/MemoryVCD/viewer/negative_items/test) | 211,364 | Metadata for all saved sampled negatives |
| [users](https://huggingface.co/datasets/WZDavid/MemoryVCD/viewer/users/test) | 323 | User index and activity counts |
| [evaluation_windows](https://huggingface.co/datasets/WZDavid/MemoryVCD/viewer/evaluation_windows/test) | 15,504 | Memory boundaries and test targets |
| [evaluation_candidates](https://huggingface.co/datasets/WZDavid/MemoryVCD/viewer/evaluation_candidates/test) | 19,751 | Positive and negative item IDs with source provenance |

All subsets use `test` to label the evaluation corpus. This is **not** a random train/test split: `evaluation_windows` defines memory and test targets within each history.

## 🎯 Negative items and sampling

`products` includes both historical items and all items required by the saved ranking/choice candidates. `negative_items` is a convenient filtered view of those same metadata records; it does not add new interactions or change the candidate pools.

| Domain | Unique negative items | Outside history | Missing metadata |
|---|---:|---:|---:|
| Beauty_and_Personal_Care | 51,966 | 49,416 | 0 |
| Books | 53,433 | 52,571 | 0 |
| Electronics | 52,641 | 51,075 | 0 |
| Home_and_Kitchen | 53,324 | 51,853 | 0 |
| **Total** | **211,364** | **204,915** | **0** |

| Candidate task | Saved rows | Negatives per row | Source |
|---|---:|---:|---|
| Item ranking | 9,876 | 19 | Saved random candidate pools |
| Item preference choice | 9,875 | 3 | Saved hard-negative pools |

**No resampling was performed.** Candidate order and IDs are preserved. The saved hard-negative files contain 3 negatives; any description expecting 4 does not match these files. Base and time-cutoff sources are distinct: filter `task`, `source_file`, and `cutoff` before evaluation. Time-axis source files cover the original first 100 users, while evaluation windows cover all 323 users. Missing candidate rows are not fabricated.

Join `evaluation_candidates.neg_asins` to `negative_items.parent_asin` **within the same domain**. `candidate_reference_count` counts occurrences across all saved candidate rows; `candidate_tasks` lists the tasks using each item. The coverage audit is in `reports/negative_metadata_coverage.json`.

## 🚀 Load and join

```python
from datasets import load_dataset

repo = "WZDavid/MemoryVCD"
reviews = load_dataset(repo, "interactions", split="test")
candidates = load_dataset(repo, "evaluation_candidates", split="test")
negatives = load_dataset(repo, "negative_items", split="test")

meta = {(r["domain"], r["parent_asin"]): r for r in negatives}
example = candidates[0]  # Select your intended task/source/cutoff for evaluation.
negative_metadata = [meta[(example["domain"], a)] for a in example["neg_asins"]]
print([(r["parent_asin"], r["title"]) for r in negative_metadata])
```

For positive item metadata, use the `products` subset. To read cached images:

```python
images = load_dataset("WZDavid/MemoryVCD-Full", "product_images",
                      split="test", streaming=True)
example_image = next(iter(images))
print(example_image["domain"], example_image["parent_asin"])
image = example_image["image"]  # PIL image; cached bytes are embedded in Parquet.
```

## 🌍 Domain statistics

The first four domains are evaluation targets (314,116 interactions). The remaining eight provide additional memory (221,582 interactions). Users overlap across domains.

| Domain | Users | Interactions | History products |
|---|---:|---:|---:|
| Beauty_and_Personal_Care | 323 | 76,094 | 33,475 |
| Books | 323 | 79,874 | 48,054 |
| Electronics | 323 | 44,268 | 25,441 |
| Home_and_Kitchen | 323 | 113,880 | 60,900 |
| All_Beauty | 293 | 2,438 | 1,336 |
| Arts_Crafts_and_Sewing | 319 | 18,588 | 10,508 |
| Automotive | 321 | 6,638 | 4,348 |
| Grocery_and_Gourmet_Food | 323 | 48,960 | 19,154 |
| Health_and_Household | 323 | 80,391 | 35,543 |
| Movies_and_TV | 278 | 9,395 | 8,433 |
| Office_Products | 323 | 28,325 | 14,032 |
| Sports_and_Outdoors | 323 | 26,847 | 15,194 |
| **Total (unique users)** | **323** | **535,698** | **276,418** |

## 🖼️ Images and metadata

| Domain | Required metadata | Image URLs | Cached image records |
|---|---:|---:|---:|
| Beauty_and_Personal_Care | 82,891 | 82,890 | 49,225 |
| Books | 100,625 | 78,730 | 49,539 |
| Electronics | 76,516 | 76,502 | 47,696 |
| Home_and_Kitchen | 112,753 | 112,753 | 72,262 |
| All_Beauty | 1,336 | 1,336 | 781 |
| Arts_Crafts_and_Sewing | 10,508 | 10,499 | 2,229 |
| Automotive | 4,348 | 4,348 | 1,397 |
| Grocery_and_Gourmet_Food | 19,154 | 19,149 | 2,911 |
| Health_and_Household | 35,543 | 35,543 | 7,689 |
| Movies_and_TV | 8,433 | 5,840 | 244 |
| Office_Products | 14,032 | 13,977 | 3,936 |
| Sports_and_Outdoors | 15,194 | 15,194 | 5,485 |
| **Total** | **481,333** | **456,761** | **243,394** |

Full includes all matched valid images in the release cache, **not 100% image coverage**: 50.57% of required metadata rows and 30.23% of history products have cached images. One invalid cached image was excluded. External URLs may expire. Product photos are separate from user-uploaded review photos; review image bytes are not bundled.

## 🧾 Key fields and files

| Table / file | Important fields or contents |
|---|---|
| `interactions` | `user_id`, `domain`, `parent_asin`, `rating`, `timestamp`, review title/text, `image_url` |
| `products` / `negative_items` | `title`, `description`, category, store, average rating, image URL; `in_user_history`, cache availability |
| `evaluation_candidates` | `task`, `source_file`, `cutoff`, `pos_asin`, `neg_asins` |
| `evaluation_windows` | Target indices, memory prefix/boundary, source domains |
| `product_images` (Full) | `(domain, parent_asin)`, embedded `image`, source URL and image hash |
| `users/users_12domains.jsonl.gz` | Original nested histories, one user per line |
| `metadata/*.jsonl.gz` | Per-domain metadata including required candidate items |
| `evaluation/source_candidates/` | Original saved hard-negative files |
| `tools/release_dataset_helpers.py` | Load nested histories and reconstruct evaluation inputs |
| `reports/`, `manifest.json` | Coverage checks, validation, file sizes and SHA-256 hashes |

Metadata uses the evaluator's slim fields: title up to 200 characters and description up to 400; price is not included. `image_shard` paths refer to the Full repository. Join images with `(domain, parent_asin)`. The `users` subset is an index; use the compressed histories for the complete nested records.

## ⚙️ Evaluation and limitations

| Setting | Released definition |
|---|---|
| Time axis | Five cutoffs: 0, .25, .5, .75, 1; a three-interaction target window |
| Domain axis | D0–D5 and Dtx; final three target-domain interactions |
| Chronology | Memory precedes the target boundary; see helper functions for exact reconstruction |
| Ordering | `source_user_index`, `source_interaction_index` and stable interaction IDs preserve provenance |
| Repeated keys | 10,404 excess records share user/domain/item/timestamp; retained to preserve timelines |
| Selection bias | Users have substantial cross-domain activity; not a representative random sample |
| Review images | Extra-eight-domain extraction did not retain review-image fields; absence is not proof of no original image |

The same histories/windows support multiple tasks; task counts must not be summed as independent users. Exact source ordering and candidate versions matter for reproducibility.

## 📚 Citation and source

Please cite the foundational MemoryCD paper and identify the MemoryVCD edition/revision used:

```bibtex
@inproceedings{zhang-etal-2026-memorycd,
  title = "{M}emory{CD}: Benchmarking Long-Context User Memory of {LLM} Agents for Lifelong Cross-Domain Personalization",
  author = "Zhang, Weizhi and Wei, Xiaokai and Huang, Wei-Chieh and Hui, Zheng and Wang, Chen and Gong, Michelle and Yu, Philip S.",
  booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2026",
  year = "2026",
  pages = "38554--38575",
  url = "https://aclanthology.org/2026.findings-emnlp.1804/",
  doi = "10.18653/2026.findings-emnlp.1804"
}
```

Derived from [Amazon Reviews 2023](https://huggingface.co/datasets/McAuley-Lab/Amazon-Reviews-2023). Original reviews and images retain applicable source terms and third-party rights. This packaging does not grant new rights to those materials.

Data release: **2026-11-12** · Negative-item view and documentation: **2026-11-18**. Shared tables are identical across URL and Full editions; Full additionally includes cached image bytes.