Dataset Viewer
Auto-converted to Parquet
text
stringlengths
2
512
labels
int64
1
5
source
stringclasses
2 values
validator_id
stringclasses
5 values
review_sentence_id
stringlengths
3
8
verification_id
stringlengths
1
4
冬物に使用していないので、その点は評価できませんが、最近洗濯している春もののマキシ丈シャツワンピースに使用しています。
3
historical
system
None
None
本体に英語でそのような表記がありました。
3
ai_feedback
ayumi
10009863
2497
袋が破れていて最悪でした。封を切ったあとに気付いたのでそのまま捨てました。
1
historical
system
None
None
いつまでも使い続けたくなります。
4
historical
system
None
None
とても使いやすくて軽くあの安さであの機能は得しました。
5
historical
system
None
None
この手のタイプで、持ち手が手の平に収まるくらい長くて、多少グリップが効いて滑りにくい商品なら文句なしだと思います。
4
historical
system
None
None
うちのものには合いませんでした。
1
historical
system
None
None
コスパがよい
4
historical
system
None
None
新しいものじゃない!そして、穏やかでもない!耐久性も最低と思います。
1
historical
system
None
None
あとはカバー無しで傷や凹みなど、どれだけ長持ちするのか?長く使えたら嬉しいです。
3
historical
system
None
None
その質感を希望する時は良いですが、しっとりと仕上げたい時は、これをしっかり乗せて、仕上げにはもう一つグレードの高いコートの塗料を使うと綺麗に仕上がります。
4
historical
system
None
None
1年半使用してホネが折れ壊れた。ただ軽さは普段の通勤で苦にならないし、これ以上軽くかつ丈夫な商品も無いため再度購入した。
3
historical
system
None
None
整骨院要らずになりにけり〜♪
5
historical
system
None
None
1年経ち、一部の部品が無くなっていることに気づいた。
2
historical
system
None
None
マイクロファイバーのふきんは最初あまり期待しておらず、ふけたらまぁいいか…程度で注文しましたが、フワフワで吸水性もよく、乾くのが早いので買ってよかったです。
5
historical
system
None
None
説明の内容とはかなり違います。
1
historical
system
None
None
取り付けも簡単に出来気に入ってます。
5
historical
system
None
None
一昔前までは明るいタイプのこの手のLED電球は意外と重く、熱を持つ感じでしたが、この電球は明るいですが本体が軽くてびっくりしました。
4
historical
system
None
None
私はある病気で、走ったりができません。
3
ai_feedback
lisa
894654
4900
研がなくて良いのと、味も良かったのでまた購入したいと思います。
5
historical
system
None
None
内側に少し厚みのあるビニール袋(10リットル用)を仕込んで水を注入しました。
3
historical
system
None
None
カロリーはないと思って飲んでる
3
historical
system
None
None
一つ言うなら、良く洗い流すです。
4
historical
system
None
None
黒の突っ張り棚探しててこちらを見つけて購入。
4
historical
system
None
None
どこにもぶつけていないのですが、いつの間にか端部が欠け、そこからひび割れていました。2枚目はどこまで持つか…。
1
historical
system
None
None
1コロで極密サボテンのように。
3
historical
system
None
None
消臭力についても他人からは言われたことがないのでなんとも。
3
historical
system
None
None
サイズはぴったり過ぎるくらいでしたが、はいているうちに次第に履き慣れてくることを期待しています。
3
historical
system
None
None
この商品で自分のサイズ測った後に宝石店に行き自分のサイズで指輪を作りましたが、宝石店で測ったサイズと全く同じでした。
4
ai_feedback
hal
2279699
312
仮にお願いしても、1週間弁当箱が使えず、一時凌ぎの替えの弁当箱を購入して(結局それを使い続けそう)、過失だと言われて有償交換となる可能性があるなら、新しいものを買った方がマシです。
1
historical
system
None
None
他のケースと比べて購入を決めました ソフトケースだと滑りやすかっりするのですがこの製品はそれはなくお値段も手が出しやすくてよかったです
5
historical
system
None
None
値段相応かな。
3
historical
system
None
None
2枚購入し、切り出して、アルミアングルで接続して真空管アンプ用のシャーシを製作しました。
4
historical
system
None
None
個包装になっているとより良いと思います
3
historical
system
None
None
クエン酸と混ぜて、重曹クエン酸水を毎日飲むようにしました。
3
historical
system
None
None
むしろ多少の音量カットでいいという人であれば向いていると思いますが、自分はもう少し遮音性があると良かったので、感想的には★2で価格の安さを加味して★3という評価でした。
2
historical
system
None
None
お店に問い合わせたところ、数回使用で剥がれるとのことでした。
1
historical
system
None
None
サイズ表を見てL-LLを購入しましたがアンダーがサイズが大きく、良くも悪くもつけている感じがしませんでした。
1
historical
system
None
None
店頭で購入するとけっこう重たくて大変なので、ネットで購入してみました。
2
historical
system
None
None
髪の毛をセットする液体を入れていますが頭全体に満遍なくかかってとても助かっています。買ってよかったです。
5
historical
system
None
None
壁紙用に購入 想定していたより切れ味が微妙で、端がうまく切れなかった 自分が下手なだけと言う可能性も捨てきれない
2
historical
system
None
None
とても良かったです。おわってしまいまさしたが、後からでも楽しめました。
5
historical
system
None
None
効果に期待しつつ、ぼちぼちやっていきたいです(˘ω&#728。
3
ai_feedback
hal
8567298
3748
パッケージの箱は、中国語で書かれてました。
3
ai_feedback
lisa
2357429
8790
又、本体だけだと風がある時等、火の安定性が弱いので五徳を合わせて購入すると良いと思います。
2
historical
system
None
None
やはり温度管理が難しい。
1
historical
system
None
None
汗をかくような暑い日やマスクをしていてもプリマヴィスタの下地をつけていると化粧が残っているので、もう手放せません。
5
historical
system
None
None
とても簡単にトリマーでき、切れ感も満足です。
5
historical
system
None
None
使い勝手がよくしっかりとカメラを守ってくれてます
5
historical
system
None
None
アトピー体質な為、乾燥する身体に塗ってます。
3
ai_feedback
hal
5735611
1549
穴が開いてて使えない
1
historical
system
None
None
事前に日本で買って行けたので安心でした。 同じメーカーの販売ブースが空港内にもありましたが、アマゾンの方が安かった気がします。
5
historical
system
None
None
今回パッドが変わって、左右が繋がった!
3
historical
system
None
None
外気温氷点下で外の車に置きっぱなしで昼に食べてもご飯は湯気が出ます。
4
historical
system
None
None
肩こりも無くなり、肩周りの筋肉、筋が、物凄くほぐれますね。
5
historical
system
None
None
纏めると、パックアウトよりかなり安かったのと好みの色合いだったのが決め手となりタフシステムを揃えましたが、正直言って私の使用用途だとパックアウトの方が向いているようでした(向こうは種類、拡張性、応用性も高いので)これからシステムケースを買う方の参考になれば幸いです。
4
historical
system
None
None
愛猫の便が緩くて色々な餌を与えてきましたがこちらにしたところ安定しています。
4
historical
system
None
None
生地は薄めで柔らかいので、冬は寒いと思います。
2
historical
system
None
None
でも、欲を言えばもう少ししっとり系でも好きかなと。
2
ai_feedback
ayumi
2659953
5632
荷物が多くて、遅れている。と思いましたが。
4
historical
system
None
None
調整温度も沢山あるし、時間も長い設定まである。
4
historical
system
None
None
以前買った商品と同じとは思えません ノンオイルとは言え あまりにもくっつき過ぎて 味もとても美味しいとは言い難い
1
historical
system
None
None
確かに回転を制御することは出来ましたが、ボリュウーム調整位置を合わすのに少し手間が掛かります。
2
historical
system
None
None
よくよく計算したら、1個265円。 高すぎ笑 ちゃんと値段見れば良かった。 再購入はないですね。
1
historical
system
None
None
切れ味良くて、 1ヶ月ぐらい使えます。
5
historical
system
None
None
注文から発送までが早く大変助かっています。
5
historical
system
None
None
購入するなら普段の靴のサイズでいいと思います。
3
historical
system
None
None
アリメツ作戦開始からそろそろ4時間。
3
historical
system
None
None
また、品物はしっかりしていて長く使えそうです。
4
historical
system
None
None
皆さんすぐ慣れると言ってましたが、私は3の痛みに慣れるまで結構かかりました。
2
historical
system
None
None
袋を開けて、電子レンジで加熱するだけで、もちもちで美味しいごはんができあがります。
5
historical
system
None
None
寒い冬の時期は重宝するアイテムです。
4
historical
system
None
None
タイヤの慣らし運転が終わり本格的に走行です窒素入れてることもあり走行は静かドライ路面でのコーナーは安定してます
5
historical
system
None
None
色は地味めですが、飽きもこなさそうで、丈夫そうで、長く使えそうです。
4
historical
system
None
None
人によりますでしょうが、、、。
2
historical
system
None
None
色々とつけてみましたが、どれもガタガタして実用性皆無でした。飾る目的ならまぁ安いしいいんじゃないでしょうか
1
historical
system
None
None
お祝いの御礼に贈りました。
3
ai_feedback
lisa
5641983
7490
すぐに届きました!オイルクレンジングで使いやすくお風呂で使えるのもありがたい!続けて使ってみたいです。
4
historical
system
None
None
浴室が寒い方は水勢の良さから発生する風で寒く感じるかもしれません。
2
historical
system
None
None
し狭い部屋に置くにはちょうど良い大きさですが組み立てが1人では難しくて息子と2人で2時間弱かかりました。
2
historical
system
None
None
今の状態ではとても満足しているのですが、へたるのが早そう。
4
ai_feedback
lisa
3218239
4939
ロジクール M570のチャタリング修理の定番商品。注文後すぐに届きました。
5
historical
system
None
None
何より見た目が格好良いのでキレイに物が収納できる。
5
historical
system
None
None
商品はまだ使ってないので何とも言えないです。
3
historical
system
None
None
玄米と同じように炊くとべちゃべちゃして好きではありませんでした。
1
historical
system
None
None
改良点は、先端部を丸くして、安全面に配慮が必要。
2
historical
system
None
None
安物買いの銭失いとはこの事ですね。
1
historical
system
None
None
なので普段は小指の爪1個分と別途脂っ気のないローションなどでセットしています。
3
historical
system
None
None
安全靴っぽくないスニーカーな感じが気に入って買ってみました。
3
historical
system
None
None
犬のトイレの掃除や散歩後の足拭きに使用しています。
3
ai_feedback
kaori
5359414
899
6月1日に種まきをし、10日に発芽しました。
3
historical
system
None
None
サバゲーで使う場合の問題点は上手く合うフェイスガードを探せるかどうかかと思います。
2
historical
system
None
None
また違う色もほしいな。
4
ai_feedback
ayumi
3224778
2654
手放せなくなります。
5
historical
system
None
None
掃除機が無いと困るのでまだ修理にも出せません。
1
historical
system
None
None
使いやすさ。保冷剤などで冷やしつつ、少しずつズラしながら使用するのに慣れが必要。
3
historical
system
None
None
2が出てるのを知らず、1読み終わって気付いてすぐ読みました。 面白すぎます。 サクサク読めすぎてまた次が気になります。
5
historical
system
None
None
拭き取りやすい。
4
historical
system
None
None
お互いに邪魔はしていませんが。
3
historical
system
None
None
なので星4つです。
4
historical
system
None
None
End of preview. Expand in Data Studio

Dataset Card for bert-jp-sentiment-20250731-1035

Dataset Summary

This dataset contains Japanese e-commerce product review sentiment data for fine-tuning BERT models. The data combines historical training data with new human-verified sentiment labels from the RLHF feedback loop.

  • Created: 2025-07-31T10:35:07.149235
  • Version: 1.0
  • Total samples: 113529

Dataset Structure

Data Instances

Each instance contains:

  • text: Japanese review text
  • labels: Sentiment label (1-5, where 1=Very Negative, 5=Very Positive)
  • source: Data source (historical, user_feedback, ai_feedback, manual)
  • validator_id: ID of human validator (for human-verified data)
  • verification_id: ID of the verification record in the database

Data Splits

Split Size
train 100914
validation 12615

Label Distribution

Training Set

  • 1 (Very Negative): 18530
  • 2 (Negative): 17260
  • 3 (Neutral): 25542
  • 4 (Positive): 19104
  • 5 (Very Positive): 20478

Validation Set

  • 1 (Very Negative): 2316
  • 2 (Negative): 2158
  • 3 (Neutral): 3193
  • 4 (Positive): 2388
  • 5 (Very Positive): 2560

Human Verification

  • Total human-verified samples: 8755
  • Number of validators: 4
  • Verification sources: ai_feedback

Source Distribution

  • historical: 104774
  • ai_feedback: 8755

Preprocessing

The following preprocessing steps were applied:

  • Text stripping and empty removal
  • Encoding issue cleaning
  • Stratified splitting

Recommended Usage

Model Training

from datasets import load_dataset
from transformers import AutoTokenizer, AutoModelForSequenceClassification

# Load dataset
dataset = load_dataset("HALDATA/bert-jp-sentiment-20250731-1035" if self.organization else metadata['dataset_name'])

# The labels are 1-5. They need to be mapped to 0-4 for training.
dataset = dataset.map(lambda example: {"labels": example["labels"] - 1})

# Load recommended model
model_name = "tohoku-nlp/bert-large-japanese-v2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name, num_labels=5)

# Tokenize data
def tokenize_function(examples):
    return tokenizer(examples["text"], truncation=True, padding=True, max_length=128)

tokenized_datasets = dataset.map(tokenize_function, batched=True)

Training Configuration

Recommended hyperparameters:

  • Learning rate: 2e-5
  • Batch size: 16
  • Epochs: 3
  • Max sequence length: 128
  • Warmup ratio: 0.1

Dataset Creation

Source Data

  1. Historical Data: Previous training datasets from production models
  2. Human-Verified Data: New sentiment labels validated by human annotators through:
    • User feedback corrections
    • AI validation discrepancies
    • Manual annotation

Quality Control

  • Text encoding validation
  • Duplicate detection and removal
  • Class balance analysis
  • Validator consistency checks
  • Minimum samples per class verification

Ethical Considerations

  • All data has been anonymized and contains no personally identifiable information
  • Human validators followed consistent annotation guidelines
  • Data sources are clearly tracked for transparency

Limitations

  • Dataset is specific to Japanese e-commerce product reviews
  • Sentiment categories reflect the original annotation scheme (1-5 scale)
  • May contain domain-specific terminology and expressions

License

This dataset is intended for research and development purposes.

Citation

@dataset{japanese_sentiment_2025,
  title = {Japanese E-commerce Sentiment Dataset},
  author = {RLHF Training Pipeline},
  year = {2025},
  url = {https://huggingface.co/datasets/HALDATA/bert-jp-sentiment-20250731-1035}
}
Downloads last month
46