Investigation into the speech rhythm of two Saudi dialects using the SAAVB corpus

The speech rhythm of Arabic has been traditionally classified as stress-timed. Speech rhythm can be measured acoustically in terms of durational characteristics of consonantal and vocalic intervals. This paper calculates and analyses speech rhythm using the Saudi corpus SAAVB. Some variables, such as ΔV, ΔC and %V, are calculated and measured in detail. The experiment and results show that the metric of consonantal interval standard deviation, ΔC, demonstrates the speech rate clearly. Also, the ΔC and %V metrics can discriminate between hypothesized rhythm classes. The %V metric detects a significant similarity between females with the Riyadh and Buraidah dialects using question sentences. The ΔC, %V and ΔV metrics have a significant difference between question sentences and statement sentences. Finally, conclusions and a general discussion on speech rhythm metrics are given for SAAVB.