istella

  • বর্ণনা :

Istella ডেটাসেটগুলি হল Istella দ্বারা প্রকাশিত তিনটি বড় মাপের লার্নিং-টু-র্যাঙ্ক ডেটাসেট। প্রতিটি ডেটাসেটে ক্যোয়ারী-ডকুমেন্ট জোড়া থাকে যা বৈশিষ্ট্য ভেক্টর এবং সংশ্লিষ্ট প্রাসঙ্গিক বিচার লেবেল হিসাবে উপস্থাপিত হয়।

ডেটাসেটে তিনটি সংস্করণ রয়েছে:

  • main ("Istella LETOR"): 10,454,629টি ক্যোয়ারী-ডকুমেন্ট জোড়া রয়েছে৷
  • s ("Istella-S LETOR"): 3,408,630টি ক্যোয়ারী-ডকুমেন্ট জোড়া রয়েছে৷
  • x ("Istella-X LETOR"): 26,791,447 ক্যোয়ারী-ডকুমেন্ট জোড়া রয়েছে।

আপনি নিম্নলিখিত হিসাবে ডেটাসেটের main , s বা x সংস্করণ ব্যবহার করবেন কিনা তা নির্দিষ্ট করতে পারেন:

ds = tfds.load("istella/main")
ds = tfds.load("istella/s")
ds = tfds.load("istella/x")

যদি শুধুমাত্র istella নির্দিষ্ট করা হয়, istella/main বিকল্পটি ডিফল্টরূপে নির্বাচিত হয়:

# This is the same as `tfds.load("istella/main")`
ds = tfds.load("istella")
  • হোমপেজ : http://quickrank.isti.cnr.it/istella-dataset/

  • সোর্স কোড : tfds.ranking.istella.Istella

  • সংস্করণ :

    • 1.0.0 : প্রাথমিক প্রকাশ।
    • 1.0.1 : float64 সমর্থন করার জন্য সিরিয়ালাইজেশন ঠিক করুন।
    • 1.1.0 : বৈশিষ্ট্যগুলিকে একটি একক 'float_features' বৈশিষ্ট্যে বান্ডিল করুন৷
    • 1.2.0 (ডিফল্ট): ক্যোয়ারী এবং নথি শনাক্তকারী যোগ করুন।
  • স্বয়ংক্রিয় ক্যাশে ( ডকুমেন্টেশন ): না

  • বৈশিষ্ট্য গঠন :

FeaturesDict({
    'doc_id': Tensor(shape=(None,), dtype=int64),
    'float_features': Tensor(shape=(None, 220), dtype=float64),
    'label': Tensor(shape=(None,), dtype=float64),
    'query_id': Text(shape=(), dtype=string),
})
  • বৈশিষ্ট্য ডকুমেন্টেশন :
বৈশিষ্ট্য ক্লাস আকৃতি ডিটাইপ বর্ণনা
ফিচারসডিক্ট
doc_id টেনসর (কোনটিই নয়,) int64
float_features টেনসর (কোনটিই নয়, 220) float64
লেবেল টেনসর (কোনটিই নয়,) float64
query_id পাঠ্য স্ট্রিং
  • তত্ত্বাবধান করা কী (দেখুন as_supervised doc ): None

  • চিত্র ( tfds.show_examples ): সমর্থিত নয়।

  • উদ্ধৃতি :

@article{10.1145/2987380,
  author = {Dato, Domenico and Lucchese, Claudio and Nardini, Franco Maria and Orlando, Salvatore and Perego, Raffaele and Tonellotto, Nicola and Venturini, Rossano},
  title = {Fast Ranking with Additive Ensembles of Oblivious and Non-Oblivious Regression Trees},
  year = {2016},
  publisher = {ACM},
  address = {New York, NY, USA},
  volume = {35},
  number = {2},
  issn = {1046-8188},
  url = {https://doi.org/10.1145/2987380},
  doi = {10.1145/2987380},
  journal = {ACM Transactions on Information Systems},
  articleno = {15},
  numpages = {31},
}

istella/প্রধান (ডিফল্ট কনফিগারেশন)

  • ডাউনলোড আকার : 1.20 GiB

  • ডেটাসেটের আকার : 1.12 GiB

  • বিভাজন :

বিভক্ত উদাহরণ
'test' ৯,৭৯৯
'train' 23,219

istella/s

  • ডাউনলোড 450.26 MiB

  • ডেটাসেটের আকার : 421.88 MiB

  • বিভাজন :

বিভক্ত উদাহরণ
'test' ৬,৫৬২
'train' 19,245
'vali' 7,211

istella/x

  • ডাউনলোড সাইজ : 4.42 GiB

  • ডেটাসেটের আকার : 2.46 GiB

  • বিভাজন :

বিভক্ত উদাহরণ
'test' 2,000
'train' 6,000
'vali' 2,000
,

  • বর্ণনা :

Istella ডেটাসেটগুলি হল Istella দ্বারা প্রকাশিত তিনটি বড় মাপের লার্নিং-টু-র্যাঙ্ক ডেটাসেট। প্রতিটি ডেটাসেটে ক্যোয়ারী-ডকুমেন্ট জোড়া থাকে যা বৈশিষ্ট্য ভেক্টর এবং সংশ্লিষ্ট প্রাসঙ্গিক বিচার লেবেল হিসাবে উপস্থাপিত হয়।

ডেটাসেটে তিনটি সংস্করণ রয়েছে:

  • main ("Istella LETOR"): 10,454,629টি ক্যোয়ারী-ডকুমেন্ট জোড়া রয়েছে৷
  • s ("Istella-S LETOR"): 3,408,630টি ক্যোয়ারী-ডকুমেন্ট জোড়া রয়েছে৷
  • x ("Istella-X LETOR"): 26,791,447 ক্যোয়ারী-ডকুমেন্ট জোড়া রয়েছে।

আপনি নিম্নলিখিত হিসাবে ডেটাসেটের main , s বা x সংস্করণ ব্যবহার করবেন কিনা তা নির্দিষ্ট করতে পারেন:

ds = tfds.load("istella/main")
ds = tfds.load("istella/s")
ds = tfds.load("istella/x")

যদি শুধুমাত্র istella নির্দিষ্ট করা হয়, istella/main বিকল্পটি ডিফল্টরূপে নির্বাচিত হয়:

# This is the same as `tfds.load("istella/main")`
ds = tfds.load("istella")
  • হোমপেজ : http://quickrank.isti.cnr.it/istella-dataset/

  • সোর্স কোড : tfds.ranking.istella.Istella

  • সংস্করণ :

    • 1.0.0 : প্রাথমিক প্রকাশ।
    • 1.0.1 : float64 সমর্থন করার জন্য সিরিয়ালাইজেশন ঠিক করুন।
    • 1.1.0 : বৈশিষ্ট্যগুলিকে একটি একক 'float_features' বৈশিষ্ট্যে বান্ডিল করুন৷
    • 1.2.0 (ডিফল্ট): ক্যোয়ারী এবং নথি শনাক্তকারী যোগ করুন।
  • স্বয়ংক্রিয় ক্যাশে ( ডকুমেন্টেশন ): না

  • বৈশিষ্ট্য গঠন :

FeaturesDict({
    'doc_id': Tensor(shape=(None,), dtype=int64),
    'float_features': Tensor(shape=(None, 220), dtype=float64),
    'label': Tensor(shape=(None,), dtype=float64),
    'query_id': Text(shape=(), dtype=string),
})
  • বৈশিষ্ট্য ডকুমেন্টেশন :
বৈশিষ্ট্য ক্লাস আকৃতি ডিটাইপ বর্ণনা
ফিচারসডিক্ট
doc_id টেনসর (কোনটিই নয়,) int64
float_features টেনসর (কোনটিই নয়, 220) float64
লেবেল টেনসর (কোনটিই নয়,) float64
query_id পাঠ্য স্ট্রিং
  • তত্ত্বাবধান করা কী (দেখুন as_supervised doc ): None

  • চিত্র ( tfds.show_examples ): সমর্থিত নয়।

  • উদ্ধৃতি :

@article{10.1145/2987380,
  author = {Dato, Domenico and Lucchese, Claudio and Nardini, Franco Maria and Orlando, Salvatore and Perego, Raffaele and Tonellotto, Nicola and Venturini, Rossano},
  title = {Fast Ranking with Additive Ensembles of Oblivious and Non-Oblivious Regression Trees},
  year = {2016},
  publisher = {ACM},
  address = {New York, NY, USA},
  volume = {35},
  number = {2},
  issn = {1046-8188},
  url = {https://doi.org/10.1145/2987380},
  doi = {10.1145/2987380},
  journal = {ACM Transactions on Information Systems},
  articleno = {15},
  numpages = {31},
}

istella/প্রধান (ডিফল্ট কনফিগারেশন)

  • ডাউনলোড আকার : 1.20 GiB

  • ডেটাসেটের আকার : 1.12 GiB

  • বিভাজন :

বিভক্ত উদাহরণ
'test' ৯,৭৯৯
'train' 23,219

istella/s

  • ডাউনলোড 450.26 MiB

  • ডেটাসেটের আকার : 421.88 MiB

  • বিভাজন :

বিভক্ত উদাহরণ
'test' ৬,৫৬২
'train' 19,245
'vali' 7,211

istella/x

  • ডাউনলোড সাইজ : 4.42 GiB

  • ডেটাসেটের আকার : 2.46 GiB

  • বিভাজন :

বিভক্ত উদাহরণ
'test' 2,000
'train' 6,000
'vali' 2,000