Add dataset card with metadata, links, and sample usage for TT-SAX benchmark

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +26 -0
README.md CHANGED
@@ -1,3 +1,29 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ task_categories:
4
+ - other
5
+ tags:
6
+ - time-series
7
+ - natural-language-querying
8
+ - symbolization
9
  ---
10
+
11
+ # NLQTSBench – TT-SAX Benchmark
12
+
13
+ This dataset accompanies the paper [TT-SAX: A Two-Track Time Series Symbolization for Natural Language Shape Querying](https://huggingface.co/papers/2610.30593). It contains the NLQTSBench benchmark data, generated query sets, and per-query experiment records used to evaluate symbolization methods for natural language shape queries over time series.
14
+
15
+ The TT-SAX method represents each window as two aligned words – a trend word and a level word – enabling more effective natural language shape querying. For implementation details and reproduction instructions, see the [GitHub repository](https://github.com/Atlamtiz/tt-sax).
16
+
17
+ ### Sample Usage
18
+ The following example from the GitHub README demonstrates the TT-SAX encoding:
19
+
20
+ ```python
21
+ import numpy as np; from ttsax import ttsax
22
+ day = np.concatenate([np.zeros(96), np.linspace(0, -4, 48), -4 * np.ones(48), np.linspace(-4, 0, 48), np.zeros(48)])
23
+ print(ttsax(day))
24
+ ```
25
+
26
+ Output:
27
+ ```
28
+ sssssssfFFFFfssrRRRRrsss|dddddddddcbaaaaaabcddddd
29
+ ```